. ' • • ZOOLOGICA SCIENTIFIC CONTRIBUTIONS of the NEW YORK ZOOLOGICAL SOCIETY VOLUME XXVII 1942 Numbers 1-19 Published by the Society The Zoological Park, New York NEW YORK ZOOLOGICAL SOCIETY General Office: 630 Fifth Avenue, New York City OFFICERS Fairfield Osborn, President Alfred Ely, First Vice-president Laurance S. Rockefeller, Chairman, Executive Committee & Second Vice-president Harold J. O’Connell, Treasurer Cornelius R. Agnew, Treasurer SCIENTIFIC STAFF General William Bridges, Editor and Curator of Publications Jean Delacour, Technical Adviser John Tee-Yan, Executive Secretary Zoological Park Lee S. Crandall, Curator of Birds Leonard J. Goss, Veterinarian Claude W. Leister, Curator of Mammals John Tee-Van, Acting Curator of Reptiles Charles M. Breder, Jr., Director Christopher W. Coates, Aquarist Myron Gordon, Research Associate in Genetics Ross F. Nigrelli, Pathologist G. M. Smith, Research Associate in Pathology Homer W. Smith, Research Associate in Physiology Department of Tropical Research William Beebe, Director Jocelyn Crane, Research Zoologist Henry Fleming, Entomologist Aquarium William K. Gregory, Associate Gloria Hollister, Associate John Tee-Van, Associate Mary VanderPyl, Associate Editorial Committee Fairfield Osborn, Chairman William Beebe Charles M. Breder, Jr. William Bridges Jean Delacour John Tee-Van CONTENTS. Part 1. April 30, 1942. PAGE 1. Social and Respiratory Behavior of Large Tarpon. By C. M. Breder, Jr. (Text-figure 1) 1 2. Tetanus in an Elephant (Elephas maximus) . By Leonard J. Goss 5 3. Descriptive Ecology of La Cueva Chica, with Special Reference to the Blind Fish, Anoptichthys. By C. M. Breder, Jr. (Plates I-III; Text-figure 1) 7 4. Quantitative Serologic Relationships Within the Artiodactyla. By Joseph G. Baier, Jr. & Harold R. Wolfe. (Text-figure 1) 17 5. Observations on the Electric Discharge of Torpedo occidentalis. By C. W. Coates & R. T. Cox. (Text-figures 1-4) 25 6. The Ampulicidae and Sphecidae (Sphecinae) Taken at Kartabo and Other Localities in British Guiana. (Hymenoptera) . By H. T. Fernald 29 7. A Resume of Mexican Snakes of the Genus Tantilla. By Hobart M. Smith 33 8. Eastern Pacific Expeditions of the New York Zoological Society. XXX. Atlantic and Pacific Fishes of the Genus Dixonina. By William Beebe. (Plates I & II) 43 Part 2. July 20, 1942. 9. Notes on Tschudi’s Types of Peruvian Birds. By Herbert Friedmann & H. G. Deignan 49 10. A Revision of the Kingfishers, Ceyx erithacus and rufidorsus. By S. Dillon Ripley 55 11. On the Reproduction of Gobiosoma robustum Ginsburg. By C. M. Breder, Jr. (Plates I & II ; Text-figure 1) 61 PAGE 12. Trichodina spheroidesi and Trichodina halli spp. nov. Parasitic on the Gills and Skin of Marine Fishes, with Special Refer- ence to the Life-history of T. spheroidesi. By Morton Padnos & Ross F. Nigrelli. (Plates I-III; Text-figures 1-4) 65 13. Mortality of Albino Embryos and Aberrant Mendelian Ratios in Certain Broods of Xiphophorus hellerii. By Myron Gordon. (Text-figure 1) 73 14. The Schooling Behavior of Mackerel : A Preliminary Experi- mental Analysis. By Arthur Shlaifer. (Plate I) 75 15. Food, Eggs and Young of the Carnivorous Snail, Euglandina rosea (Ferussac). By William Marcus Ingram & Walter Edward Heming. (Plate I) 81 16. The Anatomy and Morphology of the Hypophysis of Several Species of Ovo-viviparous Poeciliids. By Hugh E. Potts. (Plates I & II) 85 Parts 3 & 4. October 23, 1942. 17. The Synonymy of the Garter Snakes (Thamnophis) , with Notes on Mexican and Central American Species. By Hobart M. Smith 97 18. Membracidae (Homoptera) from British Guiana. By W. D. Funkhouser 125 19. A Consideration of Evolutionary Hypotheses in Reference to the Origin of Life. By C. M. Breder, Jr. (Text-figure 1) 131 Index to Volume XXVII 145 ZOOLOGICA SCIENTIFIC CONTRIBUTIONS of the NEW YORK ZOOLOGICAL SOCIETY VOLUME XXVII Part 1 Numbers 1-8 Published by the Society The Zoological Park, New York CONTENTS PAGE 1. Social and Respiratory Behavior of Large Tarpon. By C. M. Breder, Jr. (Text-figure 1) 1 2. Tetanus in an Elephant ( Elephas maximus) . By Leonard J. Goss 5 3. Descriptive Ecology of La Cueva Chica, with Especial Reference to the Blind Fish, Anoptichthys. By C. M. Breder, Jr. (Plates I-III; Text-figure 1) 7 4. Quantitative Serologic Relationships Within the Artiodactyla. By Joseph G. Baier, Jr. & Harold R. Wolfe. (Text-figure 1) 17 5. Observations on the Electric Discharge of Torpedo occidentalis. By C. W. COATES & R. T. Cox. (Text-figures 1-4) 25 6. The Ampulicidae and Sphecidae (Sphecinae) Taken at Kartabo and Other Localities in British Guiana. (Hymenoptera) . By H. T. Fernald 29 7. A Resume of Mexican Snakes of the Genus Tantilla. By Hobart M. Smith 33 8. Eastern Pacific Expeditions of the New York Zoological Society. XXX. Atlantic and Pacific Fishes of the Genus Dixonina. By William Beebe. (Plates I & II) 43 SCIENTIFIC CONTRIBUTIONS of the NEW YORK ZOOLOGICAL SOCIETY 1. Social and Respiratory Behavior of Large Tarpon. C. M. Breder, Jr. New York Zoological Society (Text-figure 1). Introduction. This report is in the nature of a continu- ation of the studies of Shlaifer & Breder (1940) and was made incidental to other work to be reported elsewhere. For these lat- ter purposes five tarpon were confined in a pen built of slats 13'2" X 7'0" X 5'8" deep. This was submerged in the artificial channel serving as a means of egress from Palmetto Key, Florida. The five fish averaged about five feet in standard length. The smallest, estimated at four feet in length, was a male in a ripe condition ; the remaining four were females that were nearly ripe. They were caught and placed in the pen by Marshall B. Bishop late in May. The counts on their rises for respiratory purposes began on June 11 as given in Table I. In this the writer was assisted by Mr. Ben Dontzin who made readings Nos. 10 to 22. One reading was made simultaneously with a similar reading in the tarpon pool described in Shlaifer & Breder (1940). It is noted in Table I as PI. Since these authors have al- ready described the respiratory behavior of immature and small tarpon it was thought desirable that some comparable measure- ments be made on the adult, sexually mature individuals. The methods employed have already been set forth by Shlaifer & Breder (1940), and Shlaifer (1941) has shown that atmospheric respiration is obligatory on this species. Results. The results obtained are clearly compar- able with those gotten on the smaller fish. There are certain noteworthy differences, however, which include chiefly that the fishes of these larger sizes do not rise with quite the frequency of the smaller speci- mens. If the number of rises per hour is re- ferred to the temperature range, Table II, and compared with the earlier data on smaller fishes, it will be noted that in the 26°-30° range the figures are almost identi- cal but slightly lower for the larger fish, while in the 31°-35° range the larger fish are definitely much lower than the smaller. This may well have to do with the environ- ment rather than the size of fish. The large fish were held in a pen through which circu- lated tidal waters, whereas the small fish were confined to a stagnant pool in which gaseous exchange occurred only through the air-water interface as there was no con- tinuous replacement by flow. A comparison of these data are given in Text-figure 1. It is evident from observations in both Boca Grande and Captiva Passes that tarpon tend to rise for air in groups. This social aspect of the respiratory need for air was discussed at length for the smaller sizes by Shlaifer & Breder (1940). Similar data for large fish in a pen, handled in identical fash- ion, is given in Table III. It will be noted that here there is relatively little social in- fluence being shown. It is much less than the average shown for small fish in various- sized bodies of water. These authors found that the larger the body of water the less the imitation, presumably due to the greater chance separation of the fishes. On this basis the present large fish should show a marked amount of imitation. That they showed less suggests the presence of some other factor. It is to be noted that the social attitude of the small fish in the pool (PI) is comparable to readings taken in it a year earlier and not with the large penned fish. These fish were approaching the spawning condition which well may have an effect on their social attitude. Later it was found on examination 1 MAY I 1342 2 Zoologica : New York Zoological Society [XXVII: 1 26-30 31 -35 °CENTIGRADE Text-fig. 1. Comparison of respiratory activity of large and small tarpon showing both means and extremes. Data on small fish from Shlaifer & Breder (1940). that these fish were actually reabsorbing their eggs under the apparent influence of captivity. They were, for their size, rela- tively closely confined, but not any more so than the small fish studied in aquaria which showed the greatest amount of imitation. A more detailed comparison of this influ- ence in large and small tarpon is impossible for the present as the effects of temperature and oxygen content at least would have to be much better understood before an at- tempted explanation of the effect on respira- tion of absolute size of fish or its condi- tion in regard to spawning time could be ascertained. As may be noted from Table I, the varia- tion in respiratory activity was marked by abrupt changes in rises per hour that could not be associated with temperature, time of day or any other ascertainable influence. As a further continuation of the work of Shlaifer & Breder (1940) observations were made on a single tarpon of 35.6 cm. in stand- ard length in the laboratory pool. These data were taken by Mr. Ben Dontzin. Con- densed, they appear as follows : Rises Date Time °C per hour 12/26/41 5:00 p.m. — 2 12/27/41 11 :15 a.m. 26.5 2 12/27/41 5:15 p.m. 23.5 4 12/28/41 10:45 a.m. 27.0 1 12/29/41 11:10 a.m. 29.0 6 12/30/41 1:15 p.m. 26.0 5 Each period of observation continued for an hour beginning with the time indicated. 1942] Breder: Social and Respiratory Behavior in Tarpon 3 The data is presented as in Table VII of Shlaifer & Breder (1940). It agrees well with their work and in reference to their temperature comparisons stands as follows : Temperature in 5° intervals centigrade. Rises per fish hour 21-25 26-30 Present data Mean 4 3.5 Maximum — 6 Minimum — 1 Shlaifer & Breder (1940) Mean 4.8 5.7 Maximum 10.4 9.6 Minimum 1.6 3.3 It thus develops that this single fish rose to breathe with slightly less frequency than did those in the larger groups (3 to 5) previ- ously studied. This is in keeping with ex- pectation on a social basis. Although this data is scant it can, in reference to the earlier work, be considered as confirmatory. The pool contained a large number of small Mugil which were not present dui’ing the earlier work, but which apparently have lit- tle if any effect on respiratory behavior. References. Shlaifer, A. & Breder, C. M., Jr. 1940. Social and Respiratory Behavior of Small Tarpon. Zoologica, 25(30) : 493-512. Shlaifer, A. 1941. Additional Social and Physiological Aspects of Respiratory Behavior in Small Tarpon. Zoologica, 26(11): 55-60. Table I. Respiratory Activity of Adult Tarpon in Confinement. (5 mature fish in a live car). (One hour observation periods). % of Minutes Rises per Greatest Time with no Rises No. Date Time °C Rises Fish Hour Between Rises All Per Fish 1 6/11 1:45 33.8 18 3.6 5:33 70.0 14.0 2 6/12 2:00 33.8 16 3.2 6:03 73.3 14.6 3 6/13 6:15 32.9 42 8.4 4:40 48.3 9.6 4 6/14 1:45 33.6 30 6.0 4:02 58.3 11.6 5 6/14 7:00 32.2 29 5.8 6:26 61.6 12.3 6 6/15 2:30 33.6 36 7.2 3:23 48.3 9.6 7 6/17 1:30 33.8 23 4.6 6:29 65.0 13.0 8 6/18 1:30 31.1 18 3.6 5:13 71.6 14.3 9 6/24 1:45 31.7 20 4.0 8:30 73.3 14.6 10 6/25 10:00 31.1 40 8.0 4:02 53.3 10.6 11 6/25 2:00 32.9 12 2.4 8:39 81.6 16.3 12 6/25 6:30 31.7 24 4.8 6:50 68.3 13.6 13 6/26 10:30 31.7 47 9.4 4:42 51.6 10.3 14 6/26 2:00 33.8 18 3.6 6:44 70.0 14.0 15 6/26 6:45 32.2 18 3.6 7:29 63.3 12.6 16 6/27 11:30 31.1 46 9.2 4:37 50.0 10.0 17 6/27 7:00 30.0 21 4.2 5:47 68.3 15.6 18 6/29 10:00 30.0 41 8.2 5:25 53.3 10.6 19 6/29 6:00 31.1 20 4.0 4:43 71.6 14.3 20 6/30 6:30 30.5 11 2.2 19:35 81.6 16.3 21 7/2 9:30 30.0 26 5.2 7:13 71.6 14.3 22 7/2 7:00 29.4 12 2.4 9:06 81.6 16.3 Mean 31.9 + 26— 5.2- 6:36— 65.3— 13.1 + Maximum 33.8 47 9.4 19:35 73.3 16.3 Minimum 29.4 11 2.2 3:23 48.3 9.6 Simultaneous reading with No. 10 above of four • immature fish in a land-locked pool. PI 6/25 10:00 33.9 51 12.75 3:17 40.0 8.0 4 Zoologica : New York Zoological Society [XXVII : 1 Table II. Respiration in Regard to Temperature Compared with that of Small Fish. Rises per Fish Temperature in 5° intervals centigrade Hour 16-20 21-25 26-30 31-35 36-40 — 4.2 3.6 8.4 — 8.2 4.0 6.0 — — 5.2 8.0 5.8 2.4 4.8 7.2 — — — 9.4 4.6 Large fish in pen — — — 9.2 2.4 — — 4.0 3.6 — 2.2 3.6 — 3.6 — — — — 3.2 — Mean 5.0 5.2 Maximum — — 8.2 9.4 — Minimum — — 2.4 2.2 ■ Small fish ( From Shlaifer & Breder, 1940) Mean 0.8 4.8 5.7 14.5 10.0 Maximum 0.8 10.4 9.6 22.3 10.8 Minimum 0.8 1.6 3.3 8.6 9.0 Table III. Fishes Rising in Groups by Percent of Total. No. 1 2 3 4 5 Total 1 100 0 0 0 0 18 2 100 0 0 0 0 16 3 100 0 0 0 0 42 4 93.5 6.65 0 0 0 30 5 93.1 6.9 0 0 0 29 6 100 0 0 0 0 36 7 100 0 0 0 0 23 8 100 0 0 0 0 18 9 100 0 0 0 0 20 10 95 5.0 0 0 0 40 11 100 0 0 0 0 12 12 100 0 0 0 0 24 13 100 0 0 0 0 47 14 100 0 0 0 0 18 15 100 0 0 0 0 18 16 95.65 4.35 0 0 0 46 17 100 0 0 0 0 21 18 100 0 0 0 0 41 19 100 0 0 0 0 21 20 100 0 0 0 0 11 21 100 0 0 0 0 26 22 100 0 0 0 0 12 Average 99.0— 1.0 + 0 0 0 26- Maximum 100 6.9 0 0 0 47 Minimum 93.1 0 0 0 0 11 Simultaneous reading with No. 10 above of 4 immature fish in a landlocked pool. PI 80.5 7.85 11.65 0 — 51 1942] Goss: Tetanus in an Elephant 5 2. Tetanus in an Elephant ( Elephas maximus) . Leonard J. Goss Veterinarian, New York Zoological Park. An Indian elephant was received at the New York Zoological Park on October 8, 1940, at the approximate age of eight years. It weighed 2,100 pounds. On May 2, 1941, the elephant’s keeper noticed that the animal was unable to open its mouth. Two and one- half hours later, examination revealed the following: jaws tightly closed; temperature, 98.6°; hypersensitiveness to sound; pro- lapsing of the nictitating membranes; and erection of the tail (poker tail) when the animal was touched or excited. A diagnosis of tetanus was made. There were numerous small cracks on the feet around the toenails and healing superficial wounds in the skin at the base of the ears. None of these wounds was sensitive to palpation or was the type of wound usually associated with tetanus infection. After examination, 100,000 units of tet- anus antitoxin were given subcutaneously: 50,000 on each side of the neck. The next day, the animal seemed less sen- sitive to noise and palpation and was able to open its mouth just enough to admit a one-inch stomach tube. It drank water when the tube was placed in the mouth and water allowed to run slowly. The animal attempted to eat but the mouth could not be opened sufficiently to admit food. It was felt that improvement was sufficient and no antitoxin was given on this day. Considerable edema was present in the neck region whei'e the antitoxin had been administered. The morning of the third day, May 4, the animal was found broadside, in tetanic spasms. Two ounces of chloral hydrate as a 7.% solution were given per l'ectum after manual removal of the fecal material. The spasms were relieved in twenty minutes and deep sleep occurred which persisted for two hours. During this time, 120,000 units of antitoxin were given and the patient again was examined for wounds which might account for the infection. This examination was not revealing. As the sleep wore off mild spasms recurred and an additional 4 ounces of chloral hydrate were given per l’ectum. During the night the chloral hydrate anes- thesia wore off but no spasms followed. On May 5, the animal was unable to rise. Slings and a hoist were used to raise the patient to its feet; the mouth could be opened about one-third its normal opening. Sixty thousand units of antitoxin were given subcutaneously in the region of the flank and three loaves of bread made into small balls were placed in the mouth. These were swallowed with difficulty, followed by water from the hose held in the mouth. Stiffness of the legs was quite apparent, but the ani- mal was capable of moving about slowly and was permitted outdoors all day. Constant unsuccessful attempts were made to eat grass and hay. In the afternoon a bran mash consisting of six quarts of bran in a bucket of water was given through a stomach tube held in the mouth. This was followed by three loaves of bread given in small balls. On May 6, the condition of the animal was unchanged. She was kept quiet and fed in the manner previously described plus three dozen bananas and two pounds of sugar. This food was given daily and the condition remained the same until May 9 when the stiffness of the legs became more pronounced and marked edema developed in the forelegs from the feet to the shoulders. Eighty thou- sand units of antitoxin were given. The quantity of bran was doubled to twelve quarts per day in addition to three dozen bananas, two pounds of sugar and three loaves of bread. From May 9 to May 25, hand feeding was continued as usual and no change in condi- tion occurred. By May 25 the animal had lost considerable weight and was down and un- able to rise. She was assisted to her feet by the use of a hoist and was kept in slings and fed through a hose until June 4 when she began to eat hay and had complete use of her jaws. 6 Zoologica : New York Zoological Society [XXVII: 2 Throughout the illness urine and fecal material were passed but in diminished quantities. Several points of interest are apparent in this case : 1. No external wound could be found which might account for the infection. 2. Chloral hydrate is an effective anes- thetic for elephants when given in doses smaller than those required for horses of a similar weight. 3. 360,000 units of tetanus antitoxin were used in treating the patient. 4. It must be assumed that the infection was of alimentary origin, which is not improbable because of the habit of elephants of eating large quantities of dirt and refuse from the ground. 1942] Breder: Ecology of La Cueva Chica 7 3. Descriptive Ecology of La Cueva Chica, with Especial Reference to the Blind Fish, Anoptichthys. C. M. Breder, Jr. New York Zoological Society. (Plates I— III ; Text-figure 1). Introduction. A blind cave-dwelling characin was de- scribed from the state of San Luis Potosi, Mexico, by Hubbs & Innes (1936) under the name Anoptichthys jordani. The original material had been collected by Senor Salva- dor Coronado that year, and sent in a living condition to Mr. C. B. Jordan of Texas, who in turn transmitted the material to Dr. Hubbs. Since then the fish has become estab- lished in small aquaria as a novelty. Nothing was recorded concerning the habitat of the form other than that mentioned by Hubbs & Innes (1936) and Hubbs (1938). The New York Aquarium undertook the organization of a small expedition to study the nature of the environment of this cave. This trip, in March, 1940, described by Bridges (1940), occupied fifteen days actu- ally spent in the cave. The present report contains the observational data obtained and as much ecological data as the field work yielded. A fuller discussion of the biological implications must wait on further labora- tory work, which was in progress at the New York Aquarium, and has lately been transferred to the Department of Animal Behavior at the American Museum of Natural History. It had been the original plan to have this translated into Spanish, edited by Senor Coronado and published in Mexico as a joint contribution. Due to inability to main- tain satisfactory contact with Coronado, and rather than have the finished manu- script lie for an indeterminate period, it was decided to delay publication no longer. The author takes this opportunity to thank Senor Coronado for his able assist- ance in the field. His energy, indefatigable efforts and general help went far toward the successful prosecution of our field work and we are grateful to Senor Antonio G. Garcia, Jeffe del Departmento Technico de la Direccion de Pisca e Industrias Mariti- mas, for releasing him from his ordinary duties in our behalf. This paper was to have preceded the following documents : Hobbs (1941), Breder & Gresser (1941a, 1941b and 1941c). This proved impractical because of the delay alluded to above. The present contribution records the basic field data of the expedi- tion, including the climatic, geologic and faunistic conditions as encountered in La Cueva Chica. Geology. The accompanying map and vertical sec- tion of the cave, Text-figure 1, which has al- ready appeared in Bridges (1940), gives a general idea of the cave. An inset gives the geographical location of the cave which is readily accessible from the concrete highway that runs from Laredo, Texas to Mexico City. All rock specimens have been examined by Dr. Horace E. Wood II and prove to be limestone formations of various types. The only exception to this is some litter on the floor of the cave which extends back to Pool No. 3. Presumably it actually goes further but is either covered with water or bat guano beyond that point. The litter it- self consists of a wide variety of materials, mud, broken logs and water-worn stones and pebbles, some of which are conglom- ei'ates of reddish jade-like materials. These objects are carried in by means of rainy season torrents. The entire region is honeycombed with sink-holes and caves of various sizes pri- marily formed by the solvent and eroding action of water. This water, heavily charged with calcium, has re-deposited materials to form stalactites, stalagmites and flow-stones, making typical cave structures. Apparently in La Cueva Chica both activities are going on simultaneously or alternately in different 8 Zoologica: New York Zoological Society [XXVII: 3 places. The consequence is that there are many badly eroded structures, while a little distance from them are new ones in the early formative state. Broken chunks of rock clearly fall from the ceiling more or less regularly. Various types of stalactites, flow-stone and cup formations are all in evidence in the various photographs. There was also, in the higher places, a consider- able amount of crystalline calcite, rhombic crystals, mostly blackish in color, and more or less amorphous masses with partly formed irregular crystals varying from yellow to tan. Pool No. 1, well protected by rock walls and under a relatively low ceiling, was found to be covered with a line dust, Plate I, Fig. 1. This was of a calcareous nature, checked by Dr. R. T. Cox spectographically, perhaps crystalized on the walls as the water receded and then powdered off on the sur- face of the water. It was not found in the damper portions further from the cave mouth. The region is one of hot springs as well as normal and cooler surface water. One such hot spring, El Banito, is only three miles from the cave under consideration. These springs are heavily charged with sulphur and evidence of the complete lack of connection with such places by La Cueva Chica is the absence of sulphurous odors and the abundant life it contains. All other places examined showed no evidence of blind fishes, nor did the local people know of any other than in La Cueva Chica. Near the river just east of Pujal a deep hole has a number of lateral fissures. In places where light enters, normal Platypoecihis and Astyanax may be found but the cave waters proper were barren. This so-called “well” is believed to intercommunicate with La Cueva Chica. Various springs in the immediate vicinity show similar conditions, as does El Nilo, a cave from the mouth of which water flows, reversing sequences at La Cueva Chica. Because of certain geological features of the general region and the interest in the possible effects on evolution of radio- active materials, tests were conducted on the presence of such emanations. Dr. M. D. Whittaker of the Department of Physics of New York University kindly undertook to make such tests on a series of water samples from each pool including evaporated con- centrates. His findings were completely negative. Since, if radioactive material were present, they would surely appear in the ground waters, it is considered established that evolutionary activity in this cave pro- ceeds without any such acceleration. For a general discussion of the geology of the region, see Schuchert (1935) and Muir (1936). The accompanying photographs show clearly the general nature of the formations. Still other photographs of the cave are given in Bridges (1940) and Dunton (1940). Water. The chemical nature of the water of La Cueva Chica is shown in Table I. The analyses have been made by the Laboratory of the New York City Department of Water Supply through the courtesy of Mr. Herman Forster. These analyses show that the cave waters are high in nitrogen as compared with the river samples. The springs, as would be expected, are intermediate. The chlorides, while variable, show no distinct trend from one type of water to another. The hardness of the underground waters is clearly less than that of the surface streams, which in part at least probably accounts for the preference of the local people to caves for their water supply. On the other hand, the alkalinity tends to be higher in the caves than outside. Phosphates are practically absent, being reported as “considerably less than .05 ppm P04.” Sulphates, on the other hand, are high in the surface waters and in one of the hot springs. They are relatively low in the cave waters. The water apparently feeds into Pool No. 1 or 2 through subterranean springs. The former is probably connected with the latter as is indicated in the map, Text-fig. 1. Pool No. 2 spills over into a running brook which widens in the area of the series of cups from which it runs into Pool No. 3. It drops into this over a nearly vertical plunge of about 25 feet. This flow continues on until the major bat roost is reached. Here there is a small waterfall from an overhanging shelf and further on another, in two streams down a mud slope to Pool No. 4. The water draining into Pool No. 4 is clearly of considerably greater volume than that leaving Pool No. 2. Consequently, it is to be interpreted that there are a con- siderable number of tributary additions along this length of the gallery. In fact, in many places there are to be seen wet and dripping places on the walls, showing the influx of additional water. The water in all places is crystal clear, even in the far recesses of the cave where everything is floored with and the water surface covered by bat guano. The tempera- tures and pH readings are given in Table I. These both are remarkably uniform and no significant temperature differential could be noted between surface and bottom in depths up to twenty feet. An exception to this must be made in several readings of temperature in the small basins above Pool Table I. Temperatures, Humidities and Water Analyses. 1942] Breder : Ecology of La Caeva Chica « •2-S 05 w pi c 05 P5 H W o o ^ £ • CD •H |> P CO Sg ^ £ J ^ cfl C CU ri ^ 9i*P! CJ C 41 ^ J> j § .2 ft c$ £ CL, o>-= ca'3> i-3 cj co 00 P CD . .5 & _)Q oo 05 «! O ” [> ^ H 1C 50 *h 05 rf ^ ^ I> H (M 00 ^ 05 oj 05 i i e* « .2 a o s Ph £ 05 S3 ° 0 o £ c (M ^ c Ph oj _Jh in o> m K*l) ctf C < o P2 03 11 ^ “ 1 c3 ii £-PP £ rt c3 ro S ^ WPh ^ PP rt .£ a) 5 .2 <££ " oj cCZ o o CO o o CD O CO o 00 CO o 00 o CD 05 LO ic o o * Cl u *> o o oo $£ o o *•• s o o © - o o «5 g o o o co o ^ °co o o >A 05 o o 00 Ol o o _p h ^ ro CD 00 ~ 05 P 05 00 05 CO 05 Cl rpp © c co :pp ^ CP 05 © 00 *R u o W a 'a! k? W P <5 « £ a o o •H1 c <35 3 rH S .ft ctf S 73 '3 a .s ai 3 ■s £ Q <5 c a> bJD o £ .2. 2 oi S 1 4 s d a> 'w a> ' CP a» bX) o aJ CP o p a O 03 CP o *Q c3 O .2 O CO 13 ~ P c3 50 u g « $ g .tT g ftfi S ft ft ft o c ft£; ft _ _ U ft 5 ft * 2 O o O W M a) S -o « ^ OJ rt ^ >-. k P'S 3j:'3 » ft5 10 Zoologica : New York Zoological Society [XXVII: 3 No. 3. Here readings of 23.5, 24.0 and 25.0 were recorded. This seemed to be associated entirely with the speed of flow through a given cup, those receiving a good flow being substantially the temperature of the large pools, while those that were relatively stag- nant varied, generally on the high side, presumably increasing in temperature be- cause of the generally warmer cave at- mosphere. The entire temperature situation in this cave is apparently influenced by the proxim- ity of the underlying magma. We are as- sured by Dr. H. E. Wood II that such tem- peratures could not be maintained in such a cavern on a basis of surface air and water temperatures alone. During the rainy season it is impossible to enter the cave, according to local state- ments. When seen by us a completely dessi- cated stream bed led into the mouth of the cave. In addition to local statements there was much evidence that during the rainy season this stream becomes a roaring tor- rent, almost surely completely closing the cave mouth. The internal evidence of the cave supported this, and it would seem that the place fills with water with the possible exception of the high-vaulted chambers which may hold pockets of air at all times. Mud carried to high cavities in the wall gave evidence of this. A considerable forma- tion of flow-stone steps is reached before Pool No. 1 is found, and were perfectly dry at the time of our visit. Probably early in the season most of the action of the water is corrosive and at- tritional, while later with the water moving slowly with both solution and deposition going on, there is a tendency to build up more deposits of limestone, which process goes on throughout the dry season wdierever water remains. Locally the water is considered thorough- ly potable and is much used by a nearby Indian village, inhabitants of which draw their w^ater from Pools No. 1 and 2. Cave Climate. The climate of La Cueva Chica at the time of our visit was fairly static, but probably this cave varies considerably throughout the year in regard to tempera- ture and humidity, at least much more so than most caves that have been reported on, partly because of its small size. Since water enters the cave in great quantities for part of the year, it undoubt- edly influences the temperature to a con- siderable extent, very likely tending to re- duce it, since swollen, rainy-season streams are generally much cooler than other sur- face waters. The ventilation of this cave is extremely limited. The only openings to the outside that we could find were the entrance we used and a small crevice opening about two hundred feet away. This latter was detected only by reason of some smoke from photo- graphic flares, used for the taking of motion pictures, finding its way out this small opening too narrow to pass a man. The be- havior of smoke from these flares and the long time it hung in the chambers pre- cluded the existence of any hidden crevice of importance. Also, the behavior of the bat colony indicated that they used the one major entrance only. Although the air was oppressive and heavy, it was not unpleasant to smell until the area of the major bat roost was entered. Here it had the acrid, gagging character- istics generally associated with a sizeable bat colony. Apparently the only change of air is that induced by weather changes, which “pump” air in or out, depending on the behavior of the barometric pressure. Added to this would be whatever dissolved gases invade or evade through the water surface to accumulate or be carried along to greater depths by the water flow. The daily flight of bats in and out no doubt also contributes to the agita- tion of the air, preventing any stratification. Up to the first pool the direct effects of daily weather changes could be detected. That is, on dry days the walls would show condensed moisture where the damp air of the cave tended to cool when it made con- tact with the outer dry air. Here at such times the atmosphere of the entrance was refreshing, but beyond such a point, varying from day to day, it was consistently op- pressive. Data on temperature and humidity are given in Table I. Terrestrial Organisms. The only moderately large terrestrial animals that apparently dwell regularly in La Cueva Chica are the bats. These are exceedingly numerous, and while it was not the purpose of the expedition to study the bats, they were sufficiently conspicuous to warrant some remarks. When Dr. Myron Gordon entered the cave in 1939 as far as Pool No. 2, he found a good sized colony over that body of water. On our visit we did not encounter more than a few small groups of bats until the region marked on the map “minor bat colony” was reached. From here on bats were numerous, concen- trated into two colonies, the second of which was immense. The few specimens collected were kindly identified by Dr. J. E. Hill, of the American Museum of Natural History, as: Artibeus jamaicensis Leach, Natalus mexicanus Miller and Mormoops megalophylla senicula Rehn. Due to the inrush of water in the 1942] Breder: Ecology of La Cneva Chica 11 rainy season these bats almost surely must vacate the cave for part of the year. Micronycteris megalotis mexicanus Miller was taken in El Nilo. Mr. Marshall Bishop reported seeing some bats scamper high up the walls in true vampirine fashion. This, coupled with local accounts of vampire attacks on farm ani- mals, sounded convincing, and the finding of droppings in El Nilo that looked very typical of vampire droppings, leads us to believe that they are actually dwelling in this region. Aside from that of human beings, there was no other evidence of mammalian activ- ity within the cave. Birds, reptiles and amphibians seemed to be completely absent. Arthropods in the form of insects and arachnids were ubiquitous. A representa- tive collection was made by Mr. Bridges and has been deposited in the American Museum of Natural History. They will form the basis of a separate report by Dr. W. J. Gertsch. Probably the most conspicuous insects were small flies which flew about our lights in large numbers. These flies were kindly identified by Dr. C. If. Curran, of the American Museum of Natural His- tory, as Pholeomyia indecora. Lowe (Mili- chiidae) and some Psychoda sp. (Psycho- daidae). The former was the pi'edominant form. The most evident arachnids were large whip-scorpions which clung openly to the walls. The bat guano contained great quantities of a macroscopic free-living nematode. When this material is finally studied it will prob- ably be found to contain a fauna of some variety. This material is now in the hands of Dr. R. F. Nigrelli. Aquatic Organisms. The invertebrate aquatic organisms con- sisted of microscopic forms and two macro- scopic forms. One, a crayfish, Macrobrach- ium jamaicensis (Herbst), which was lighter in color than those outside but with functional vision, is not to be considered as a cave form proper. Another, and smaller form, has been described as a new sub- species by Dr. H. H. Hobbs, Jr., under the name Cambarus blandingii cuevachicae, (Hobbs 1941). This lack of optical differ- entiation is equally true of the microfauna which is being studied by Dr. Nigrelli. The only aquatic vertebrate encountered was the fish that the expedition set out to study. These fish had twice before been collected. Originally they were taken in 1936, as discussed in the introduction. Gordon and Coronado in 1939 made a hur- ried visit to the cave and took a second col- lection. In the first visit specimens were taken from Pools 1 and 2, while on the second they could be found only in Pool 2. As neither party was equipped to go further than Pool 2, it remained for the present and third to examine the fish fauna to the workable end of the cave. The most striking feature of this faunal unit was the discovery that these blind fish were not a uniform group. The fish pre- viously collected were all blind and their offspring likewise grew up to be sightless creatures. Mr. Albert Greenberg of Tampa, Florida, has been especially successful in breeding this fish in captivity and obtained uniform material to the fifth generation. Shortly after the return of the expedition, a visit to his establishment showed that he had thousands of specimens of various ages, and he had noted that although he had reared them through five generations in light, all were completely blind. It was quickly found in the cave that the fish ranged all the way from eyeless, pale creatures to fish that could not be dis- tinguished from the normal river Astyanax mexicanus (Filippi). The introduction of a light into the cave apparently causes those individuals endowed with eyes sufficiently perfect to recognize a light beam to retreat hastily, while the truly blind individuals seemed to give no atten- tion to the strongest beam of a flashlight (See Breder & Gresser, 1941a, 1941b and 1941c). It was only after we had become thoroughly familiar with the cave and had baited the fish to given spots that we ob- tained the eyed and partially eyed forms. Further, we found that there was a dis- tinct gradient in that the further we went into the cave the more numerous became the fully eyed forms, and those fish not at all distinguishable from the normal river fish were only obtained beyond the large falls at Pool 3. Table II gives a measure of this gradient based on the arbitrary division of the fishes in blind, sunken eyed, covered and uncovered, and “normal-eyed,” as based on our preserved collections. These rather arbitrary categories may be described as follows : Blind — Eye socket covered with tissue level with the cheek, no evident eye struc- ture. Sunken eye — Some evident eye structure, but sunken below rim of orbit. Covered — Evident sunken eye covered with tissue. Uncovered — Evident sunken eye ex- posed as in a pit. “Normal” — Eye convex and appearing as in a river fish, irrespective of its size, which was frequently very small. Pigmentation followed a similar course although not fully correlated with eye struc- ture. The pigmentation has been also arbi- trarily divided, the five categories of which 12 [XXVII: 3 Zoologica : New York Zoological Society Table II. Eye Condition and Pigmentation of Cave Characins. Eye Condition Expressed in % of catch. Based on 119 specimens. Sunken Eye Location1 Blind Covered Uncovered “Normal” ] Sta. 1 85 6 9 Sta. 2 162 8 45 31 Sta. 3 — 9 9 82 Pigmentation Location None Little Moderate Considerable Full Sta. 1 90 2 6 2 Sta. 2 34- 34 5 8 19 Sta. 3 3 29 32 24 12 1 Sta. 1 indicates Pool II on map ; Sta. 2, Pool III and Sta. 3, Pool IV. 2 Two specimens in this group blind on one side only. are given in Table II. The extent of correla- tion of these two features associated with cave life are indicated in Table IV. The sizes of the eyes of those fish with “normal” eyes are given in Table III, com- pared with river fish. Because of the varia- tion in relative eye size with absolute size in fishes, this table has been broken into three size groups for purposes of compari- son. From this treatment it is clear that the cave fishes extend from normal eye size to very small as compared with the river fish. Actually, the smaller eyed fishes taken in the river may represent a true genetic contaminant issuing from the cave or a gen- eral constitutional and initial eye variation in this group. Since the connection with the river is from the far end of the cave there may be a more or less continual interchange be- tween the river and cave fauna. The cave ends, as far as human entry is concerned, in an eliptical chamber, the form of which is well indicated in the map and section together with the presumed under- ground exit of the flowing water. Here the Rio Tampaon is about half a mile distant. A study of this variation in the eyes and pigmentation of these fishes must be re- served until an extended laboratory effort is made to obtain at least a basic under- standing of the genetic foundation of this population. Because of the bearing on laboratory work the data of Tables II and IV have been used by Breder & Gresser (1941a). During our visit to the cave, the fish were clearly in their reproductive season. Large females turgid with eggs were com- mon as were small specimens that could not have been more than a month old. Mr. Greenberg found that in captivity the addi- tion of some cold water to an aquarium would induce spawning, which led him to suppose that the rainy season might be the reproductive period. These two items taken together might be used to argue that spawn- ing occui’s throughout the year, the peak perhaps being reached when the rains come. Against this view would be the mechanical circumstances accompanying the torrential waters passing through the cave during that period, which would hardly seem friend- ly to spawning, especially of the type em- ployed by these blind fish. The sexes of the fish, together with their sizes, are given in Table V. An attempt to examine the scales of these fish led to the surprising finding that re- placement scales were the rule rather than the reverse. In fact, only thirteen of the twenty fish large enough to show markings were useful in this connection, as is in- dicated in Table V. Whether this is to be construed as evidence that due to swim- ming in the dark these fish continually knock scales off themselves or whether they take a serious battering when the torrents of the rainy season come, can only be specu- lated upon at this time. Although these fish under certain conditions will ram into ob- jects, Breder & Gresser (1941a and 1941c), Size Range S.L. in mm. 0-30 31-60 61-90 Table III. Eye Size of Cave and Surface Characins. Expressed as % of standard length. Based on 69 specimens with “normal” measureable eyes. Cave Specimens Maximum Mean Minimum 11 10 8 7 5 5 Surface Specimens Maximum Mean Minimum 12 11 10 10 8 6 8 8 8 7 6 1942] Breder: Ecology of La Cueva Chica 13 Table IV. Association of Eye Condition and Extent of Pigmentation. Expressed in % of extent of pigmentation. Based on 119 specimens. Eye Extent of Pigmentation in Per Cent Condition None Little Moderate Considerable Full Blind 98i — 2 Sunken Eye (covered) 56 33 11 — Sunken Eye (uncovered) — 62 15 15 8 “Normal” 2 23 28 23 24 1 Two specimens in this group blind on one side only. there is no observation noting that they dis- lodged scales by such accidents. It would be unwarranted to attempt to define the nature of the markings on the scales as to whether they are annual or otherwise. While they seem to occur in a reasonable sequence with size, we have no way of equating this with time. Although we know nothing of the various effects of their underground environment on the circulae of the scales, it must not be supposed that it is markedly uniform, for there is a clear annual cycle in the wet and dry season rhythm of this region. The fish are surely subjected to an annual change in temperature, quantity of water, rate of flow, food and perhaps in the chemical nature of the water itself. The food problem, as already suggested by the presence of other than cave creatures, is simple in this cave. The great abundance of food objects leading directly to the out- side renders the problem of primary diet quite simple. Organisms all the way from Cyclops and dipterous insects to the entire carcasses of bats are available to these fishes. Their stomach contents were found to consist of bat droppings and parts of other and smaller cave characins and their eggs. This would suggest that the only regular input of energy into the population for large parts of the year is bat dung. Their ability to thrive and reproduce on the ordin- ary foods supplied to small aquarium fishes also suggests the lack of any peculiar spe- cialization in dietary requirements. Discussion. The primary items of a cave fauna such as the present certainly center about the manner of establishment of a population of blind fishes and the nature of the asso- ciation of lack of light and congenital blind- ness. Superficially simple-looking, critical examination of the possible development of such a condition presents some distinctly puzzling phases. The finding of a long series of inter- mediates between the blind fish and the normal eyed river fish, which in itself is unique among cave fishes, gives hope that this material should prove of value in any attempt to understand this general associa- tion of blindness and darkness. Such a study, however, involves much further work. As the situation stands, the facts in hand are subject to various interpretations. It might be assumed that these fishes in their normal river environment carry the genetic factors for an eye defect. Dr. C. L. Hubbs informs me that large series of Mexican Astyanax which he has examined show a surprisingly large amount of individual variation in eye diameter. Blind larval fishes Table V. Size, Sex and Growth of Cave Characins. Based on 119 specimens. Standard Length in mm. Sex Maximum Mode Minimum Female 83 45-50 40 Male 62 50-55 38 Immature 50 35-40 16 Rings on Scales 6 * 5 4 ❖ ❖ -h 3 ❖ ■fi 2 ❖ ❖ ❖ 1 * 0 * * * * * * * * * * 16 21 26 31 36 41 46 51 56 61 66 71 76 20 25 30 35 40 45 50 55 50 65 70 75 80 Standard length in 5 mm. intervals. Of the 20 fish above 55 mm. only 13 had other than replacement scales. * 81 85 14 Zoologica : New York Zoological Society [XXVII: 3 in the open river could hardly be expected to survive. However, with these fishes find- ing the way into cave waters, the eyed offspring presumably would have no ad- vantage over the eyeless, resulting in the survival of some of each. Following this thought along, two pos- sible conditions suggest themselves. It could be that such an entry was made some time ago and the resultant population as found is cut off from the river fauna and the eyed fish go on living in the absence of any detriment to having eyes, even if they cannot be used. An alternative interpretation would be that there is a continual interchange of fishes between the cave and the river. It is not surprising that blind fish are not to be found in the river, for they clearly have strong disadvantages in such an en- vironment, principally their blindness and conspicuous light hue. The contrary would not hold, and there may a more or less continual penetration of the cave waters by normal river fish. If this latter view is correct, it would suggest that the develop- ment of such a population of blind fish took a much longer time than would the first alternative, because of the slowness of the spread of a character that is evidently re- cessive in a continually diluting population. The finding of a progressive series of eyed fishes as one moves toward the river en- courages such a view. A quite different interpretation would be to assume that at some distant time a group of fish became entrained in this cave in a state of complete isolation and became blind by whatever mechanism operates under such conditions. Then a further assumption would be made that this population of thoroughly blind fish again came in contact with the normal river fish from which they were originally derived. The resultant stock as found would then be the hybrid mixture of these two groups. The increasingly higher number of eyed fishes as one nears the river would be compatible with this view. Other views concerning the possible direct effect of environment on vision and pigmen- tation find little support in the present mate- rial. There are eyed and intermediates liv- ing in the cave successfully with the eye- less for an unknown number of generations. Five generations of the entirely blind stock l’eared in brilliant light show no suggestion of returning vision or pigmentation. Experimental studies are here called for and in the words of Gresser & Breder (1940), “Until at least some of these are undertaken, it would seem to be pointless to attempt further speculation.” Progress in this direction has already been made and is reported in Breder & Gresser (1941a, 1941b and 1941c). Summary. 1. La Cueva Chica is able to support a population of temperature limited characins by virtue of nearby thermal waters which prevent the subterranean waters from fall- ing below a relatively high value. 2. The cave characins are supported by a large variety of food items which trace directly or indirectly to the outside by way of bat droppings. 3. The cave characins ( Anoptichthys jordani ) show complete intergradation with the river characins ( Astyanax mexicanus ) through a long series of individuals with intermediate eyes and pigmentation and surely represent a single population. There is a pronounced gradient in these features from one end of the cave to the other. 4. Other animal organisms found living in the cave, bats, crustaceans, insects, spid- ers and related forms and a considerable microfauna, are not modified in any way comparable to that of the fishes and are not confined exclusively to a cave habitat. 5. The cave itself, small in extent, so far as human entry is concerned, except for its high temperature is typical of limestone formations and shows no other exceptional features. Bibliography. Breder, C. M., Jr. & Gresser, E.B. 1941a. Correlations between structural eye defects and behavior in the Mexican blind characin. Zoologica. 26(16) : 123-132, pi. I-IV, 2 text-figs. 1941b. Behavior of Mexican cave characins in reference to light and cave entry. Anat. Rec. 81(4) : 112 [Abstract 216]. 1941c. Further studies on the light sensitivity of the Mexican blind characin. Zoo- logica. 26(28) : 289-296, pi. I, 2 text- figs. Bridges, W. 1940. The blind fish of La Cueva Chica. Bull. N. Y. Zool. Soc. 43(3): 74-97, 24 illustrations. Dunton, S. C. 1940. Factors in cave photography. Journ. Biol. Photo. Assoc. 9(2): 50-58, 6 illustrations. Gresser, E. B. & Breder, C. M., Jr. 1940. The histology of the eye of the cave characin, Anoptichthys. Zoologica. 25 (10) : 113-116, pi. I— III. Hobbs, H. H., Jr. 1941. A new crayfish from San Luis Potosi, Mexico. (Decapoda, Astacidae). Zoo- logica. 46(1) : 1-4, 1 text-fig. Hubbs, C. L. 1938. Fishes from the caves of Yucatan. Carnegie Instit. Wash. Pub. (491) : 261-295. 1942] Breder: Ecology of La Cueva Chica 15 Hubbs, C. L. & Innes, W. T. 1936. The first known blind fish of the fam- ily Characidae: a new genus from Mexico. Occ. Pay. Mas. Zool. Univ. Michigan. (342) : 1-7, pi. I. Muir, J. L. 1936. Geology of the Tampico Region, EXPLANATION Plate I. Fig. 1. Calcareous scum on the surface of Pool No. 1. Photo by E. B. Gresser. Fig. 2. Blind fish as found in Pool No. 2. Here the individuals are nearly all of the fully blind type. This is the place from which the original collection was made, representing at once the exact type locality as well as the source of the parent stock of this form now to be obtained from dealers. Photo by S. C. Dunton. Fig. 3. Cup-like basins below Pool No. 2. Photo by S. C. Dunton. Mexico. Amer. Assoc. Pet. Geol., Tulsa, Oklahoma, xix+208 pp. Schuchert, C. 1935c. Historical geology of the Antillean- Caribbean Region or the lands bor- Caribbean Sea. Wiley, N. Y. xxvi + 811 pp. Oklahoma, xix+208 pp. OF THE PLATES. Plate II. Fig. 4. Pool No. 4, showing the low arch across its middle. The material floating on the water is caked bat guano. Photo by S. C. Dunton. Plate III. Fig. 5. Cave fish in various stages of eye de- generation. Reading from the top down: Fully blind and pigmentless type; Somewhat pigmented with a minature eye; “normal” fish from cave; “normal” fish from the Rio Tampaon. Photo by S. C. Dunton. ^ V ..- W' ■ I ■ ■' ■ pn :;,".v - .■ ■ m y '.■■;• . •; •• ' • ' • •,■' - ■ • . '■ ’ ■• ',• '. >■ ' fe>£ > t; m $ v ' ftp- '> ~-K‘ ?W\WS 0%''V> Sx> 1 :^®Svfe h'- ■ ;ft .- -V :.,t V 1:$%^ l :.;: ' - ■• ..;• ../..■• ■•':-• V. \Y •}'< >• J. ' l(. 1 ‘ 1 •' , ft " . " : iiSKY ' t£, ' : ■ : : m$; i YY Y Y?Y ;g ■ .y ,:. ' 1 ■ > Very low arch ■£& : : : - ■' ft ' , v u ft ft. . ftftftftftftft; :00 /r> ;yy.Y: ’ i \ ir-V:' ■ '• yy ■ .. ■■■■- . mmm Y" ;Y - Soft mud slope;;.;1 , y;-' "v:;- :: ^ :r - v: YvY'V"' ;-■■■■■•• ;.v- - >.< Y.V-Y Y ■ ■■ . Yv"Y. • ■ ''VV-: Y 'Y. . YY.y> : - ; ■ ;y . ■ ■" ■■ 1 ■ . Y. ,, ®r.c : teii-jo : wtensspS n . 'j ( *!*& ,'4-' , teu$m mmmmm : ■ .i ■ ■ ■ ■ ' ■ -Y' V;Y. V'/ ; :*s ' ■ : ' ■ ; ■■■■'. ■ : '0: v ":0 l \ . . ■ ; ‘ w arch gpgsst v .' ■ ■ : ', ' /'S:,: .1 Bat guano covers all ! from here to end #1 -Y'Jv fcv m,;,' ..v :,yvY;YY ' '!.> :: . ’Y; ySSt PLAN OF CAVE Bm§ mmmm ■ ■'■' 1 ’ ;' \ , > . v Y ft* 0/0 '•>;■•••' ’ ‘. V:/ '•'•••’»'»' • YY: yY,Y :.,Y. 0m00tl ... . . ' . '.'V. . ■■.■'''■■ : ; - ..... ; '■■ r.; ■-•-■ ; ■'■■■■ YY ; .;; y;y :y ..,, ■ ■ YY/ : y 0 ■; ' 1 I ' ,Y, / , , ' Y - ; •; • *.!.;'<■ jY:. ’ 'Y Y Y; . ':■:■■■:•■ ' ' ''Y ’ YYYYY.; .■ Y : ; Y ■■ • yv;.'.y;y7vV",y:." ■■■'-' mm'£ ';;;y ‘•v, •. y*-. * . f- -v * ^ 'iT'uivM, ' \shf. ' i o'j'--, >. ^ m yy.Y' ,<; Y, Y.YYY YY l: Y; ’ Y': ■:•, ') . Y Slflpt V^‘Y. ■y, ;; y y; vy''yy: '■.'■■ .YY.' - ■ ; v . ‘0000 \ yy.1 : i v.y ; ■ ” s ;® V?>i,iYy-Y ItSilJ '.yy-;y.;Yv; ' YYY:y-'y;Y/,.-;Y:,/;.Y; I;y £y 'y -.y ■■ y,;.. y;;y; U0 0£sm M AAmMMimMi MAJOR BAT ROOST \ • ' YY >Y:/ V; t -a 5f?e -- fo.’ji'S jU « %Y ■YfY/® " laaMSY IsSteg^ill Soft mud slope gM§§ '■'V .-■:. .:" - ■■ if..' ' s vv':i. ■ :• YYY Y-YiY ..■• .■/'.. YV Y / : Y-' 'M . ,Y Y". •. . , C.-r-Y/JYS a ' : , yftl 0MJ) A ■ , ssfil . LOCATION OF LA CUEVA CHICA fit :A<*S'3 iifelS Scale of Miles 10 . . Limit of direct daylight U ■ !\ ' , 4* i ' Very low arch fp AJ P / ( ' ' f LaCueva f Chica DIBanito Axtla Slope up ’ll Nilo J Isig^p f-. VALLES >■ Soft mud slope . ■■ r.y.-./ ■■ . .About 9 feet higher ; than main floor / A '• ' ’oo'S-^^T^-' Sm°oth fi00r Low arch mg \ . ■ ; Bat guano covers all '/ •\i from here to end m.,1 * \ , / / jj \ '■', , Much recently ■ ; , -A Undercut - connects . ■ . ■ , with Pool Low arch// /A/A! fAS! / PLAN OF CAVE SHiligs* fS}“ tJMiSfl ■HSTju i»n ^ fis "- v * - ?*':U ■ ssK&dHHMr fttfl MINOR BAT ROOST ■■ ^AA ' f V.. ../ ■'A i ®§ .. .: * V&-S&SS& .V MAJOR BAT ROOST Connects with ' - Pool 1 7 C - ■‘ mm vAa/A;: ' SB& La Cueva Chica Pujal, S.L.P. MEXICO Limestone Water PROJECTION OF CENTER LINE OF CAVE ON THE NORTH AND SOUTH AXIS Scale of Feet C Small crevice opening to surface about ^;v’; 200 feet away V; ;AV'; Text-fig. 1. Plan and elevation of habitat. BREDER. PLATE I. Fig. 1. Fig. 2. DESCRIPTIVE ECOLOGY OF LA CUEVA CHICA, WITH ESPECIAL REFERENCE TO THE BLIND FISH, ANOPT1CHTHYS. BREDER. PLATE II. Fig. 3. Fig. 4. DESCRIPTIVE ECOLOGY OF LA CUEVA CHICA, WITH ESPECIAL REFERENCE TO THE BLIND FISH, ANOPTICHTHYS. BREDER. PLATE III. Fig. 5. DESCRIPTIVE ECOLOGY OF LA CUEVA CHICA, WITH ESPECIAL REFERENCE TO THE BLIND FISH, ANOPTICHTHYS. 1942] Barer & Wolfe: Serological Artiodactyla Relationships 17 4. Quantitative Serologic Relationships Within the Artiodactyla.1 Joseph G. Baier, Jr. and Harold R. Wolfe Department of Zoology, University of Wisconsin (Text-figure 1). Introduction. Any study of animal relationships justi- fies the use of a serological method capable of stating in exact mathematical terms the degree of relationship of the animals being studied. In the present study a volumetric quantitative precipitin technique having these qualifications was employed in a study of the blood sera of some Bovidae and Cervidae. Although the precipitin reaction has been used in the study of plant and animal rela- tionships since 1900, only a few investi- gators have thought of its use in exact quan- titative terms. Boyden (1926) expressed its quantitative nature and stated that measure- ments of degree of relationship so obtained are independent of interpretation. Further elaborating on the idea, Boyden & Baier (1929) devised an exact quantitative volu- metric precipitin technique that is “simpler and more rapid than any other which has been used in the quantitative study of blood relationships, and that through it highly significant measurements of biological rela- tionships may be made.” Their technique of measuring volumetrically the amount of pre- cipitate formed in the reaction was a de- cided improvement over the methods of Nuttall (1904), Schur (1904), Hamburger (1905), and Mollison (1924), in that of all these workers, they were the only ones who gave an adequate statement of the reliability of their technique. In one series of 36 de- terminations the average error of the indi- vidual readings was 5 per cent, when com- pared with the mean of the series. The aver- age deviation of the means of successive pairs of readings was the same as the devi- ation of the whole series, while the means 1 The Works Progress Administration furnished personnel assistance under the University Natural Science Project, Work Project No. 10324 for this research. The authors wish to thank the various contributors listed in Table I for furnishing the samples of blood sera used in this study. of the values taken in quartets dropped to 3 per cent. An error in technique of this value, when supported by statistical analy- sis, is indeed very significant. According to Boyden (1934), “the results of the applica- tion of such a technique to the study of serologic relationships should be of great interest. It is likely that this technique will succeed in distinguishing closely related species, which have heretofore been indis- tinguishable by the precipitin test.” So far, this is the only volumetric test which has been used in serological relationships. Using this improved volumetric tech- nique, Baier (1933) established the con- stancy of in-vitro factors for proper execu- tion of the tests. Wolfe & Baier (1938) by using the ring test and the volumetric pre- cipitate measurement procedure showed that the in vivo injection procedure may influ- ence the “type” of precipitin that may be produced. They found that high-titered (ring test) antisera were produced by one or two series of injections of undiluted antigen while continued re-injections resulted in an increase in the precipitate forming power of an antiserum without causing an increase in the ring test “titer” of the antiserum. They indicated the presence of (1) a “titer”-pro- ducing antibody and (2) a “precipitate”- forming antibody. It is important when at- tempting volumetric relationship studies that an antiserum be employed having high precipitate forming powers. A more recent technique of precipitate measurement which should parallel the volumetric technique is that of the Libby Photronreflectometer (Libby, 1938) which measures the amount of precipitate formed in the precipitin reaction by nephelometric methods. So far two papers have appeared (Boyden, 1938, and DeFalco, 1941) indicat- ing its possible use in relationship studies. With the reliability of the volumetric technique well established, an investigation 18 Zoologica: New York Zoological Society [XXVII: 4 of the serological relationships of some Bovidae and Cervidae was attempted to parallel a similar study of Wolfe (1939) who reported on some of these same sam- ples of blood. In his paper, however, only the “ring” test was employed. The earliest work of actual precipitate measurement in a problem of animal rela- tionship studies is that of Nuttall (1904), who with Strangeways reported, among others, some studies made with the bloods of some Cervidae and Bovidae. Since then only the paper of Boyden (1934) reported one brief study of an ox-sheep reaction as a suggestion of the possible use of a volu- metric precipitin technique. Materials and Methods Antibodies were produced in healthy adult male and female rabbits of various breeds. The undiluted serum antigens were injected intravenously at intervals of a month or more; the initial series of injections con- sisted of three injections given on alternate days and the total quantity of antigen in- jected was 3 ml. Subsequent series of injec- tions consisted of two injections on alter- nate days of a total of 1.5 ml. This method of antigen injection was shown by the au- thors (1938) to be conducive to the produc- tion of good precipitating antisera. The animals were bled from the heart with sterile syringes and needles, the blood allowed to clot, and the exuded serum fil- tered through Seitz filters. The serum was transferred to sterile ampoules and stored in the refrigerator until used. The serum antigens used for antibody production (Table 1) were deer, ox, buffalo, sheep, and goat. Test antigens consisted of various species of Artiodactyla (Table II). The tests, in vitro, were carried out fol- lowing the method of Boyden & Baier (1929 ) using standardized Van Allen thrombocy- tocrits to measure volumetrically the pre- cipitate obtained by incubating known amounts of antigen and antiserum. Mixtures of 0.5 ml. of antiserum and 0.5 ml. of antigen (either homologous or heterolog- ous), diluted to give a protein content so as to stay in the range of relative antibody ex- cess, were made directly into the thrombocy- tocrits. These were then placed in a water bath at 37.5° ± 0.5° C for one hour and centrifuged for two fifteen-minute periods in a tachometer controlled centrifuge at the rate of 2,400 R.P.M. Readings were made in duplicate and sometimes in quadruplicate. The protein content of all antigens was based on total nitrogen which was deter- mined by a modified macro-Kjeldahl method. The test antigens were standardized to give similar protein content by diluting with buffered saline (Evans, 1922). It was neces- sary at times to dilute the antiserum with buffered saline in order to avoid an exces- sive amount of precipitate. Tests with any one antiserum were always made at a con- stant dilution of antiserum. Any variations in the measurable precipitate were due then to differences in the proteins of the various blood sera used in these relationship studies. Relationships were recorded in terms of the percentage of volume given by a heterolog- 053 .04 .026 .02 CONCENFRA T/ON OF An T/GEN .16 .08 0/3 1942] Baier & Wolfe: Serological Artiodactyla Relationships 19 ous determination in comparison to the homologous precipitate volume taken as 100 per cent. The average of at least six read- ings on the curve of reaction, of three dif- ferent antigen concentrations in the area of relative antibody excess was taken for any one relationship value. By using six read- ings as a minimum the statistical reliability of the test could be established (Boyden & Baier, 1929) and by taking these readings from the area of relative antibody excess the resolution of the precipitate by excess antigen could be avoided as well as obtain- ing a greater constancy in the readings (Baier, 1933). He also demonstrated that it is not advisable to use the entire curve of reaction since readings taken in the area of antigen-antibody equilibrium are unreliable. Text-fig. 1, taken at random from Table V, illustrates these points from the reaction of anti-deer serum with the homologous and heterologous antigens used in this study. The region of relative antigen-antibody equilibrium is shown as the discontinuous peak of the curve. To the left is the area of antigen excess, while to the right is shown the area of antibody excess where the heterologous readings were made for rela- tionship studies. Results. In Table III are presented the data ob- tained from two anti-ox and one anti-buffalo sera. These antisera were reacted with their homologous antigens and with several heterologous antigens. The data of 21-3 and 22-2 show that the buffalo and eland sera are more closely re- lated to ox than are the sera of the other Bovidae or the Cervidae. Furthermore, the per cent, values indicate that buffalo is more closely related to ox than is the eland. This essentially verifies the results of Wolfe (1939), but it should be emphasized that the technique employed in the present paper enables a distinction between ox and buf- falo antigens which he could not show using the ring test with unabsorbed sera. The two ox antisera did not give similar degrees of reaction in per cent, with the heterologous antigens but the relative posi- tions of the animals was constant. Similar results will be noted throughout this paper. It is advisable, therefore, to emphasize Table I. Test antigens. Family Scientific Name Common Name Source of Material Bovidae Bos taurus Ox (2 samples) Mayer Packing Company Biso?i bison American buffalo Yellowstone National Park Taurotragus oryx Eland New York Zoo Poephagus grunniens Yak San Diego Zoo Anoa depressicornis Anoa New York Zoo Ovis aries Sheep (3 samples) Mayer Packing Company Capra hircus Goat University Farm Ovis tragelaphus Aoudad New York Zoo Ovis canadensis by Ovis Mountain sheep hybrid San Diego Zoo musimon Antilope cervicapra Black buck antelope New York Zoo Cervidae Odocoileus virginianus White-tailed or Virginia deer Madison Zoo Cervus axis Axis deer New York Zoo Cervus ca?iadensis Wapiti (2 samples) New York Zoo Table II. Quantity Inj< ?cted First Additional Bled Homologous Rabbit Antigen Series of Series Series (days following Titer Number Injected Injections (undiluted) (undiluted) last injection) (ring-test) 21-3 Ox 4 3 ml. 1.5 ml. 10 1,024,000 22-2 Ox 3 3 ml. 1.5 ml. 10 256,000 54-2 Buffalo 3 3 ml. 1.5 ml. 10 512,000 30-2 Sheep 3 3 ml. 1.5 ml. 10 512,000 50-3 Goat 4 1.5 ml. 10 512,000 51-3 Goat 4 1.5 ml. 10 256,000 62-1 Virginia deer 2 1.5 ml. 7 512,000 62-2 Virginia deer 3 1.5 ml. 7 512,000 62-3 Virginia deer 4 1.5 ml. 8 512,000 62-4 Virginia deer 5 1.5 ml. 8 512,000 62-5 Virginia deer 6 1.5 ml. 8 512,000 * Between .75 to 1.5 mg. total protein per kg. of body weight. This is actually the second series of injections as the animals were previously injected with minute quantities and reported by Wolfe (1939). 20 Zoologica : New York Zoological Society [XXVII: 4 Table III. Antigen 21-3 anti-beef Antisera anti-beef 22-2 anti-buffalo 54-2 Ox-1 100.0 100.0 78.8 Buffalo 94.3 84.9 100.0 Eland 84.8 64.17 Yak — — 77.7 Anoa — — - 74.4 Sheep - 5 44.9 52.90 — Sheep - 6 — 54.53 — Sheep - 6W — — 38.4 Goat 36.0 53.03 Black buck antelope 42.2 32.47 22.7 Wapiti 43.71 42.8 30.7 Axis deer 50.13 39.23 42.3 Virginia deer 45.8 53.93 44.8 *Data may not be reliable due to excessive hemoglobin in test samples. phylogenetic position rather than actual per cent, relationship. The percentages in the reactions of the Bovidae sera, other than the buffalo and eland, and of the Cervidae sera, were, on the whole, quite similar. Thus the wapiti, Vir- ginia deer and axis deer seem to be as closely related to the ox as are the sheep, goat and black buck antelope. Since such closely re- lated forms as the goat and sheep or the axis deer and wapiti did not, as would be ex- pected, give similar percentage reactions, it seems necessary to treat the more distantly related forms as a group, rather than to attempt to give each animal a definite posi- tion in the table. The buffalo antiserum (54-2) was reacted with yak and anoa bloods as well as with some of the antigens tested with ox antisera. The ox, yak, and anoa bloods all gave simi- lar percentages and showed a much closer relationship to buffalo than did the other Bovidae and Cervidae. This result was to be expected. Again the more distantly related forms gave inconsistent results and must be treated as a group. Table IV presents the data obtained from anti-sheep and anti-goat sera reacted against their homologous antigens as well as representative heterologous antigens. The anti-sheep serum (30-2) was able to dis- tinguish between sheep and goat sera, and their high percentage values indicate a closeness in the relationships of these forms. That these percentage values are statisti- cally reliable is indicated by the ratios of the means to their respective probable er- rors. For sheep-6 serum the ratio was 55:1, while for goat serum the ratio of the mean to the probable error of the mean was 140 :1. These figures, in being well above the 4:1 ratio generally accepted as indicating sta- tistical reliability, are highly reliable in stating that by means of this volumetric test it was possible to distinguish sheep serum from goat serum which had not been hitherto usually possible using the ring test with unabsorbed sera. The order of rela- tionship for the other animals indicated that ox and buffalo were more closely related to the sheep than were the Virginia deer, black buck antelope, eland, axis deer and wapiti. The reactions of the sera of these distantly related forms did not give the consistent re- Table IV. Antigens 30-2 Antisera 50-3 51-3 anti-sheep anti-goat anti-goat Sheep - 6 100.0 — — Sheep - 5 99.60 91.9 63.8 Goat 96.30 100.0 100.0 Aoudad — 84.3 — Mountain sheep hybrid — 81.2 — Ox-1 87.40 42.8 27.6 Buffalo 85.57 44.7 24.8 Eland 70.40 — — *Black buck antelope 71.57 — — Virginia deer 77.27 51.3 30.7 *Axis deer 68.53 40.9 20.5 *Wapiti - 1 61.7 — 23.1 *Wapiti - 2 — 54.8 22.8 Data may not be reliable due to excessive hemoglobin in test antigens. 1942] Baier & Wolfe : Serological Artiodactyla Relationships 21 suits expected of closely related forms and these species should be regarded as a group rather than individually. The two anti-goat sera (50-3, 51-3) were more specific than the anti-sheep serum. The sheep serum again could be distinguished from the goat serum, and the high per- centage reaction of the aoudad and moun- tain sheep hybrid show their closeness to the goat. On the other hand, the remaining Bovidae and Cervidae reactions were lower and it cannot be said which ones are more closely related to the goat. This result is consistent with the data for anti-ox and anti-buffalo sera presented in Table III. Antigoat serum 51-3 gave much lower percentages with the heterologous antigens than did 50-3, indicating that the former serum can be considered to be more specific. Such differences in serum specificity are also known to occur with the ring test method. Table V illustrates the data of five anti- Virginia deer sera produced in a single rabbit. This rabbit was given several series of injections and bled after each series. Animals injected by this method were shown by Wolfe & Baier (1938) to produce high precipitating antisera useful in quan- titative volumetric precipitin studies but giving very aspecific reactions if the ring test technique is used. The reactions of all five antisera gave larger amounts of precipitate with members of the deer family than with the Bovidae. In every case the axis deer and wapiti could be distinguished readily from the Virginia deer, while the Bovidae gave considerably lower percentage values. The relative close- ness of the axis deer and wapiti to the Vir- ginia deer is not definite from the data pre- sented. In two out of three instances where both tests were made the axis deer showed the closer relationship, but in the other just the opposite condition was observed. Then again, the two wapiti bloods do not give the same degree of reaction. The only explana- tion the authors wish to offer is that these samples of sera contained a large amount of hemoglobin and possibly the per cent, of protein, obtained on the basis of total nitro- gen, was inaccurate. The authors feel that exact protein content of the active antigens is essential and disagreements of the type illustrated can be avoided only if better methods can be devised for measuring only the reactable antigens. Discussion. The data presented in this paper confirm and extend the evidences for facts concern- ing the blood relationships of some species of Bovidae and Cervidae previously shown by morphological and by other serological techniques. That morphology has its place in phylo- geny is not disputed; that it has its limita- tions is evident. The serologist can apply his studies to a more exact and quantitative estimation of present relationships with a technique which is independent of mor- phology. This is the aim of the present paper. The volumetric method of measuring the amount of precipitate formed in the pre- cipitin test has enabled a distinction be- tween some very closely related forms. Thus ox and buffalo, and sheep and goat could be distinguished from each other. This is usually not possible with unabsorbed anti- sera using the ring test. The advantages of the volumetric test over the ring test is of this nature. Its disadvantages are that the test is time-consuming in its operation, protein contents of all test antigens must be very accurately determined, and as to the data, it has not been possible so far to in- dicate the degree of relationship of more distantly related forms. It is hard to ex- plain why the more distantly related forms reported in this paper show inconsistent de- grees of relationship to a test antiserum and why all of these distantly related forms, re- gardless of their phylogenetic position, show about the same per cent, of distant relationship. No attempt was made to study Table V. Anti- Virginia Deer Sera Antigen 62-1 62-2 62-3 62-4 62-5 Virginia deer 100.0 100.0 100.0 100.0 100.0 *Axis deer 83.7 — 96.4 72.4 59.6 *Wapiti - 1 76.2 66.3 88.6 — 66.4 * Wapiti - 2 — 77.0 — — 78.1 Goat 69.1 43.7 76.8 56.8 — Sheep - 5 — 45.9 77.2 69.6 50.6 Ox - 1 — 44.7 70.9 — 48.8 Ox -5 57.1 — — — — Buffalo — 41.6 76.8 63.5 53.2 Eland — 40.6 — 50.9 — *Black buck antelope 60.9 37.3 61.9 59.0 — Data may not be reliable due to excessive hemoglobin in the test antigens. 22 Zoologica : New York Zoological Society this problem at the present time; distantly related forms were merely spoken of as a group rather than as individuals. A possible explanation may be that the relative position of the region of antibody- antigen equilibrium in shifting toward the left (refer to Text-fig. 1), as reactable anti- gen decreases and toward the right as re- actable antigen increases will alter the na- ture of the curves being studied, and only by studying antigens having approximately the same reactable antigen content will the curves of reaction be reliable. No degree of controlling total antigen content by means of Kjeldahl determinations can alleviate this difficulty as the Kjeldahl determinations measure the total protein content of test antigens and not the per cent, of reactable or specific protein antigens. The ring test titers on the other hand, are not influenced either by slight variations in protein concentration or by the proportions of antigen and antibody in the reaction mix- tures since the end point used (titer) is sim- ply the maximum dilution of an antigen that will form a ring of precipitate at the junc- tion of antigen and antibody. Comparisons of the data obtained by the ring test and by the present volumetric technique can be made by referring to the paper of Wolfe (1939), who reported on three of the rabbits used to produce anti- sera in the present paper. These rabbits (numbers 50, 51, and 62) were given one series of minute injections of antigen to produce the specific antisera required to distinguish closely related forms using the ring test, and were then given additional series of larger injections to produce the high precipitate forming antisera for use in the volumetric precipitin test. The results are in general agreement as has been stated previously. The technique employed has consistently enabled a distinction of such closely related forms as ox from buffalo, sheep from goat, and Virginia deer from axis deer and wapiti. Were the ring test to be employed, such dis- tinctions could not be consistently made with undiluted sera. The volumetric method used in this paper seems to be a very reliable one in showing the differences of very closely related forms. On the other hand, the more distantly related forms used in this work can be classified together only as a group rather than as individuals. A more or less rough grouping is possible from the data presented. The more closely related forms could be placed in one group and sub- divided according to their closeness of rela- tionship while the more distantly related forms were placed in a second group and subdivision was not possible except in one instance (anti-sheep serum). In tabular form, this data can be pre- [XXVII : 4 sented as a brief serological classification in the following fashion. A. Classification based on anti-beef sera Group I A. Ox B. Buffalo C. Eland Group II Other Bovidae and Cervi- dae tested B. Classification based on anti-buffalo serum Group I A. Buffalo B. Ox, yak, anoa Group II Other Bovidae and Cervidae C. Classification based on anti-sheep serum Group I A. Sheep B. Goat Group II A. Ox, buffalo B. Other Bovidae and Cervidae D. Classification based on anti-goat sera Group I A. Goat B. Sheep C. Aoudad, mountain sheep hybrid Group II Other Bovidae and Cervidae E. Classification based on anti-deer sera Group I A. Virginia deer B. Axis deer and wapiti Group II Bovidae Summary. 1. The sera of thirteen representative species of Bovidae and Cervidae were used as test antigens. 2. Eleven antisera were produced against five of these thirteen species. 3. Antisera were produced having high precipitate forming powers. 4. The per cent, of relationship is reported on the basis of the volume of precipi- tate formed in the reaction mixture when compared with the homologous re- action taken as 100%. 5. Ox, buffalo and eland are related to each other in the order named, and could be distinguished from each other. 6. Virginia deer could be distinguished from axis deer and wapiti. 7. Ox, yak and anoa are closely related to buffalo. 8. Sheep and goat could be distinguished from each other. 9. Aoudad and mountain sheep hybrid could be distinguished from goat and these forms were more closely related than were the other Bovidae. 10. This work confirms ring test studies, but enables a finer distinction of closely related forms. 1942] Baier & Wolfe: Serological Artiodactyla Relationships 23 Bibliography Baier, Joseph G., Jr. 1933. Quantitative studies on precipitins. Physiol. Zool., 6:91. Boyden, Alan A. 1926. The precipitin reaction in the study of animal relationships. Biol. Bull., 50:73. Boyden, A. A. 1934. Precipitins and phylogeny in animals. Am. Nat., 68:516. Boyden, A. A. 1938-39. Serological study of the relation- ships of some common invertebrata. Carnegie Inst. Wash. Year Book No. 38:219. Boyden, A. A., & Baier, J. G., Jr. 1929. A rapid quantitative precipitin tech- nique. Jour. Immun., 17:29. DeFalco, R. J. 1941. A comparison of antigens by inter- facial and nephelometric methods. Proc. Soc. Exp. Biol, and Med., 46:500. Evans, A. C. 1922. A buffered physiological salt solution. Jour. Inf. Dis., 30:95. Hamburger, H. J. 1905. Zur Untersuchung der quantitativen Verhaltnisse bei der Prazipitinreak- tion. Folia haemotologia, 2:539. Libby, R. L. 1938. The photronreflectometer — an instru- ment for the measurement of turbid systems. Jour. Immun., 34:71. Mollison, T. H. 1924. Serodiagnostik als Methode der Tier- systematik und Anthropologie. Abder- halden’s Handbuch der biologischen Arbeits-Methoden, Abt. 9, Teil 1, Halfte 1, p. 554. Nuttall, G. H. F. 1904. Blood immunity and blood relation- ship. Cambridge, England: Cambridge University Press. Schur, H. 1904. In: Kolle and Wasserman. Handbuch der pathogenen Micro-organismen, 4:631. Wolfe, H. R. 1939. Serologic relationships among Bovidae and Cervidae. Zoologica, 24:309. Wolfe, H. R., & Baier, J. G., Jr. 1938. Comparison of quantitative precipitin techniques as influenced by injection procedures. Physiol. Zool., 11:63. 1942] Coates & Cox: Discharge of Torpedo occidentalis 25 5. Observations on the Electric Discharge of Torpedo occidentalis. C. W. Coates Aquarist, Neiv York Zoological Society and R. T. Cox Department of Physics, New York University. (Text-figures 1-4). In June of 1941 a specimen of Torpedo occidentalis was brought to the New York Aquarium and several more were brought in during the next few weeks. These were all large, in excess of 15 kilograms in weight, and one was very large, 61 kilo- grams, probably a record size. A few electrical measurements were made on the first of these fish within a few hours after its arrival at the Aquarium. Circum- stances unfortunately delayed further ob- servations, and before they could be resumed all these specimens died. Two others, how- ever, remained available at Point Lookout, Long Island, at the fish pier of Mr. Robert Doxsee, from whom the former specimens had been obtained. To avoid possible injury to these fish by the handling necessary in transporting them, it was thought advisable to observe them without bringing them to the Aquarium. Through the courtesy of Mr. Doxsee a cathode-ray oscillograph and accessory equipment were set up at Point Lookout, partly on the pier and partly on the live well floating alongside, in which the fish were kept. Observations under these conditions were somewhat difficult and, what was more unfortunate, the two fish, after having been for several weeks in the live well, were evidently in very poor condition. Although for these reasons our observa- tions are necessarily rather fragmentary, they appear to contain some information not recorded elsewhere, and a brief report on them would thus seem to be worth while. The first specimen showed a peak voltage of 220 volts when the dorsal and ventral surfaces of its electric organ on one side were connected to the oscillograph on “open circuit,” so that no appreciable current was drawn except what flowed in the circuit made through the body of the fish. Except for the voltage drop caused by this current, the full electromotive force would have been measured in this way. Probably the electro- motive force was not much higher than the measured peak voltage, and may therefore be taken as approximately 220 volts. In this species, however, the electromotive force appears to vary widely with the condi- tion of the fish. The measurements made on this specimen were so arranged as to keep the fish out of water as short a time as pos- sible. Actually they were all completed in about one minute. We can not say how many discharges were made in this time; there were certainly several hundred, possibly a thousand. Whether from being out of water, from fatigue, or from the handling inciden- tal to the measurements, the peak voltage on open circuit dropped 60 volts during this time. More striking still was the difference between this specimen and the two observed at Point Lookout. Although these were of about the same size as the first specimen, their peak voltage on open circuit was only about 25 volts. The oscillographic traces produced by the first specimen were not recorded photo- graphically but only noted visually. Even so, certain differences between the traces made by the torpedo and those made by the electric eel and described elsewhere1 were evident. The discharge of the torpedo did not show so sudden a rise in voltage or so abrupt a transition from a rapidly rising to a gently falling voltage. A few photographic 1 Coates, C. W., R. T. Cox, and L. P. Granath. The Electric Discharge of the Electric Eel, Electrophnrtis elec - tricus (Linnaeus). Zoologica, Vol. XXII (Part 1), No. 1, April 5, 1937. 26 Zoologica : New York Zoological Society [XXVII: 5 TIME M/LL/ SECONDS- (a) TIME MILLISECONDS (b) Text-fig. 1. a. Oscillographic trace of the elec- tric discharge of Torpedo occidentalis. b. Oscil- lographic trace of the discharge of Electro- phorus electricus. traces were obtained of the discharge of one of the two specimens obsei’ved at Point Lookout. Text-fig. 1(a) was drawn from one of these: 1(b), shown for comparison, was made from a photographic trace pro- duced by the electric eel. In copying from the photographs, the time scale was made the same for the two figures. The scales of voltage are different between the two, hav- ing been so chosen as to make both peaks of the same height. It is apparent that the duration of the discharge of the torpedo was longer than that of the discharge of the electric eel. But it should be repeated that the discharge shown for Torpedo is that of a specimen in poor condition. The electric organs of Torpedo occiden- talis, like the large organs of the electric eel, throw off discharges in trains, the sepa- rate discharges following one another at an interval of a millisecond or so. With the specimen first observed there were more discharges in one train than the three, four, or five commonly observed with the electric eel. Without a photographic record, it can not be said just how many there were. Prob- ably there were at least a dozen in the aver- age train. The photographic traces obtained at Point Lookout, perhaps because of the poor condition of the specimen, showed fewer discharges to the train than were observed with the first specimen. The observed regularity of the discharges within a train and in successive trains sug- gested that the same quantity of tissue was active in each discharge, and it is natural to suppose that the entire organ connected to the oscillograph was active together. Also it was possible to show plainly that the right and left organs discharged sim- ultaneously. The method used was that employed before to measure the time lag between the discharge in anterior and pos- terior portions of the large organs of the electric eel, and it had been described else- where.2 One of the vertically deflecting plates in the oscillograph tube being joined to one of the horizontally deflecting plates, their junction was connected to a large sheet of metal on which the torpedo rested, the ventral surfaces of both organs being over the plate. Two smaller sheets, some distance apart on the dorsal surface of the fish, cov- ered the two organs on that surface. These two sheets were connected respectively to the other two deflecting plates of the oscillo- graph tube, so that the discharge of the organ on one side would produce a vertical deflection and that on the other side a hori- zontal deflection. ( Strictly speaking, the dis- charge on one side would produce a nearly vertical, that on the other a nearly horizon- tal, deflection, the complete separation of the two deflections being prevented by leak- age of the current across the body of the fish from one side to the other. An auxiliary observation, in which one sheet was over the organ on one side while the other was over non-electric tissue, was used to estimate the possible effect of this leakage) . Both organs discharging exactly together would produce equal vertical and horizontal deflections, and thus would combine to make an oscillo- graphic trace along a straight line at 45 degrees with either the vertical or the hori- zontal direction. A small time lag between the discharges of the two organs would change the trace into a loop, the width of the loop being greater the greater the time lag. Actually the trace appeared simply as a straight line. Had there been a time lag as great as .0001 sec. it should have been de- tectible by this method. Some muscular activity accompanies the discharge of Torpedo occidentalis. The fish, which is flat and roughly disk-like in shape, showed distinct tremors around the periph- 2 Coates, C. W., R. T. Cox, W. A. Rosenblith, and M. Vertner Brown, Propagation of the Electric Impulse Along the Organs of the Electric Eel, Electrophorus elec- tricus (Linnaeus), Zooloqica, Vol. XXV (Part 2), No. 14, July 3, 1940. 1942] Coates & Cox: Discharge of Torpedo occidentalis 27 Text-fig. 2. Diagram of the electric circuit for the measurement of the peak voltage of the discharge. ery of the body when the discharge occurred, although it is not certain whether these were directly before or during the train of discharges. This activity caused the margins of the wings to curl upward, sometimes more than 10 cm. from the surface on which the fish was lying. During a long series of discharges these margins were raised throughout the period with accompanying tremors starting from the thick central region of the body and spreading outward to the periphery. With the electric eel, no muscular activity, or at least none comparable to this in extent, accompanies the discharge. The circuit used in the measurements made on the first specimen is shown in Text-fig. 2. The fish rested on a metal sheet Slf which was at ground potential. The wet ventral surface made good electric contact with this sheet. The smaller sheet S2 was shaped to cover the dorsal surface of the electric organ on one side of the fish, firm pressure and moisture making good electric contact here also. These electrodes were con- nected to the oscillograph, so that the volt- age between them could be measured. By closing the switch K the variable resistance R could be connected across the organ so as to draw a current during its discharge. The resistance R being known and the potential difference V at the peak of the discharge being measured, then by Ohm’s law the cur- Text-fig. 3. Graph of peak voltage V against peak current I. 28 Zoologica: New York Zoological Society [XXVII: 5 rent I at the peak of the discharge is given by the equation V = IR Also the peak power P supplied by the elec- tric organ to the resistance is given by P = VI One set of measurements completed be- fore the fish showed signs of fatigue, is given in Table I below. Table I. Peak Voltage, Current and Power with Various Resistances. R V I P ohms volts amperes watts co * 220 0 0 370 230 0.6 140 100 210 2.1 440 50 205 4.1 840 25 200 8.0 1600 12.5 170 13.6 2310 * Open circuit. That these data are not precise is clear enough from the fact that the peak voltage on open circuit was observed as less than that across a resistance of 370 ohms. Such irregularities are caused both by the in- accuracy of the observation and the vari- ability of the fish. It will be noticed that the power rises with falling resistance over all the range of the observations. The reason for this is that without prior experience with the species we failed to provide calibrated resistances low enough to develop the maximum power Text-fig. 4. Graph of peak power P against peak current I. of the discharge. But some inference beyond the observations may be made, as is illus- trated in Text-figs. 3 and 4. In Text-fig. 3 the values of V and I from Table I have been plotted and a straight line has been drawn among the plotted points. This construction may be justified by Ohm’s law, and it is supported by many observations we have made on elec- tric fish of other species. Values of V and I taken from this graph were used to compute the power P for the graph shown in 4, in which the plotted points again represent values from Table I. The greatest current of which the organ is capable would be obtained by a complete short circuit, that is, by having R, and hence also V and P, practically zero. According to the graphs this current is 60 amperes. It is clear from the figures that the extrapolation is doubtful, since this value of the current is more than four times the highest reached in the observations. Probably, however, the organ is capable of a current of at least this order of magnitude. This does not mean that so great a current is normal to the fish. It is probable that the organ discharging in sea water carries something like half of its maximum current, since that would be the condition for the greatest release of power in the water. For the maximum power Text-fig. 4 gives a value a little more than 3 kilowatts. The extrapolation here is more trustworthy than that for the maximum cui’rent, since this power is only 30 per cent, higher than the highest observed. Since the two electric organs discharge together, we have 6 kilo- watts or about 8 horsepower as the electric power of both organs at the peak of the dis- charge. Of course it must not be overlooked that this is the peak power, and the average power, even during the discharge, would probably be some 25 or 30 per cent, of this. The average during a train of discharges would again be less than the average during one discharge, and would probably be 500 to 1,000 watts for both organs. Finally it should be noted that the trains of discharges do not appear to be given often. The cross-sectional area of the organ on either side, in a plane parallel to the ventral surface, was about 250 square centimeters. The current was in a direction perpendicu- lar to this plane, and its value for maximum power was, as shown above, about 30 am- peres. Hence the current density at maxi- mum power was about 0.12 ampere per square centimeter. The mass of the organ on each side was about 2,000 grams. The maximum power released outside the organ being 3,000 watts, the maximum power per gram of the organ must have had an instantaneous value around 1.5 watts or one-third of a calorie per second. 1942] Fernald: Ampulicidae and Spliecidae of Kartabo 29 6. The Ampulicidae and Sphecidae (Sphecinae) Taken at Kartabo and Other Localities in British Guiana. (Hymenoptera) d H. T. Fernald. (During the eight years of occupancy, by the Department of Tropical Research, of the New York Zoological Society’s Station at Kartabo, British Guiana, extensive collec- tions of insects were made. Most of these were taken in the quarter-mile area under intensive study — what may be called the Guiana Junglezone. Details and a general summary of this area may be found in “Studies of a Tropical Jungle,” Zoologica, Vol. VI, No. 1.) Introduction. The insects treated of in this paper were collected in British Guiana by members of the Tropical Research Station, New York Zoological Society, mainly at or near Kar- tabo. A few were taken at Arakaka on the Barima River in northern British Guiana; others at Mt. Everard on the Waini River; at Wismar on the west bank of the Deme- rara River; at Waratuk on the Potaro River below Kaieteur Falls ; and at Lama Stopoff, just east of Georgetown. While the Ampulicidae are few in number it is interesting to note that both species, described by Kohl, do not appear to have been recorded since, so far as literature available to the writer shows. In the Sphecinae all four tribes are repre- sented and as usual for tropical regions, the Chlorionini are the most abundant. That only four specimens of Sphecini are present is a little surprising as several others are liable to occur in this region. Only one of three species of Sceliphron from this part of South America is repre- sented, but twelve specimens indicate its abundance. No doubt further collecting in the northern and central portions of British Guiana will show the presence of other species of all the groups included in this paper. 1 Contribution No. 635, Department of Tropical Re- search, New York Zoological Society. Family Ampulicidae. Ampulex Jurine. Jurine, 1807. Nouv. meth. de class, les Hymen., PI. 14, p. 132. Ampulex sagax Kohl. Kohl, 1893. Ann. cles k.k. naturhist. Hof- mus. Wien, Vol. VIII, Hefts 3 & 4, pp. 477, 480. Two female specimens, one from Kartabo, 28-V-1924; the other 24978 only. The only reference to this species is the original de- scription where no locality is given. Kohl writes that he considers it a tropical species. Ampulex minor Kohl. Kohl, 1893. Ann. des k.k. naturhist. Hof- mus. Wien, Vol. VIII, Hefts 3 & 4, p. 480. One female from Kartabo, 21-V-1924. The only other reference is by Kohl whose specimen came from Brazil. Family Sphecidae. Subfamily Sphecinae. For many years this group was rated as a family but it has now been reduced to subfamily rank along with other former families. The Sphecinae are divided into four tribes, the Podiini, Sceliphronini, Chlorio- nini and Sphecini. All four of these tribes are represented in this collection. Tribe Podiini. Podium Fabricius. Fabricius, 1804. Syst. Piez., p. 183. Podium (Dynatus) nigripes Westwood. Westwood, 1832. Griff. Anim. Kingd., Vol. XV, p. 516, fig. 3. {Dynatus) . One female, Kartabo, 29-III-1924. This species occurs from Mexico to Argentina. 30 Zoologica : New York Zoological Society [XXVII: 6 Podium (Trigotiopsis) abdominale Perty. Perty, 1833. Delect, anim. artic. Brasil, p. 142, PI. 27, fig. 18. (Trigonopsis) . Three specimens : one male from Kartabo, 9-VI-1924; a female, also from Kartabo, 5-IV-1922 and a female variety from Bartica District, 6-V-1922. This species is found from Guatemala to Brazil and Peru. Podium brevicolle Kohl? Kohl, 1902. Abhandl. d. k.k. zool-botan. Ges. Wien, Vol. I, Heft 4, p. 62. One male specimen, probably of this spe- cies, taken at Kartabo and labelled “Peach killer.” This species is found from Mexico to Brazil. Podium gorianum Lepeletier. Lepeletier, 1845. Hist. nat. Ins., Hym., Vol. Ill, p. 324. Sixteen specimens, all females; eleven from Kartabo, 23-VIII-1920; 24-VII-1922; 29-V, 3- VI, 11-VI, 15-VI, 28-VI, 1924; No. 22414; and two undated; Bartica Dis- trict, 1, 2, and 3-VI-1924; 174; two without data. This species has been taken from Surinam to Brazil (Para). Tribe Sceliphronini. In this tribe only the genus Sceliphron is represented and that by only one species. Sceliphron Klug. King, 1801. Neu Schrift. Ges. naturf. Fr. Berlin, Vol. Ill, p. 561. Sceliphron fistulare Dahlbom. Dahlbom, 1843. Hym. Eur., Vol. I, p. 22, no. 8. (Pelopoeus) . Twelve specimens, all females; Kartabo, 1919, 63, 20-VII-1922 ; 8, 4; No. 20859; Kalacoon, 1916 (2) ; Arakaka, 21-XII-1922 (2); Lama Stopoff, 14-X-1917; Kartabo, Hym. ; no data, one specimen. This species is present in the neotropical region from Mexico to Paraguay (Argentina?). Tribe Chlorionini. This is the most abundantly represented tribe of the subfamily. Five divisions of the group, rated by some workers as genera, by others as subgenera, are recognized. The writer treats them here as subgenera, three of which are represented in this collection. Chlorion Latreille. Latreille, 1802. Hist. nat. Crust. & Ins., Vol. Ill, p. 333. Subgenus Priononyx Dahlbom. Dahlbom, 1845. Hym. Eur., Vol. I, p. 439, No. 14. Chlorion ( Priononyx ) striatum Smith. Smith, 1856. Cat. Hym. Brit. Mus., Vol. IV, p. 266. ( Priononyx ) . One female specimen taken at Kartabo, IV-3-1926. Distribution — Nevada?; Ari- zona to Argentina. Subgenus Isodontia Patton. Patton, 1880. Proc. Bost. Soc. Nat. Hist., Vol. XX, p. 380. Chlorion (Isodontia ) costipennis Spinola. Spinola, 1851. Mem. Acad. Sc. Torino, Vol. XIII, p. 54, No. 35. (Sphex) . Four males and eight females. Males: Kartabo, 6-X-1921, 6-VI-1924; Mt. Ever- ard, 15-XI-1922; Arakaka, 2-XII-1922; Females : Kartabo, 2-IV-1924, 3-VI-1924, 3-VII-1924; Bartica District, 20-IX-1917 (2) ; Penal Settlement; W. Bank Dem. R. 5-1-1923; Mt. Everard, 15-XI-1922. Distri- bution— Santo Domingo; Mexico to Brazil. Chlorion (Isodontia) dolosum Kohl. Kohl, 1895. Ann. des. k.k. naturh. Hof- mus., Vol. X, Heft 1, p. 49. Two female specimens taken at Kartabo, 21-V and l-VI-1924. From the literature available these speci- mens seem to be the only examples of this species captured since the one described by Kohl, recorded as from French Guiana, and they vary somewhat from his description. In one specimen the legs and petiole are entirely black, while in the other they are partly red. The abdomen, though partly red, has dark shades here and there, and the black hairs on the clypeus are large and almost bristle-like; elsewhere they are yel- lowish, almost golden in places. These insects, while for the most part Isodontian in character, also possess some features like Ammobia. In one specimen there is an evident stigmatal groove and the petiole is shorter and straighten than in most Isodontias. There seems to be what may be considered a rudimentary tarsal comb of seven short, stout spines much like those present in Ch. ( Isodontia ) aztecum to which this species also seems to be most closely related in general by its stout struc- ture and petiole length. The form of the second cubital cell and the distance apart on the radial vein of the second and third transverse cubital veins compared with that between the second transverse cubital and the second recurrent veins on the cubital vein are typically Isodontian. Also the jaws do not each reach to the base of the other when closed, the two teeth being short. This species may fairly be placed at the end of the series of Isodontia species, near- est to Ch. ( Isodontia ) aztecum on the one 1942] Fernald: Arnpulicidae and Sphecidae of Kartabo 31 hand and to Ch. { Ammobia ) species on the other, much as Ch. { Ammobia ) lucae is re- lated to the other Ammobias. Kohl’s speci- men was recorded as from French Guiana. Subgenus Ammobia Billberg. Chlorion (Ammobia) brasilianum Sauss. Saussure, 1867. Reise d. Novara, Zool., Vol. II, p. 39. Four female specimens: Kalacoon, 1916, Hym. 25; W. Bank Dem. R., 9-II-1923; Waratuk, 16-11-1921; 29-IV-1924. This species occurs from Guatemala to Brazil. Chlorion ( Ammobia ) melanopum Dahlbom. Dahlbom, 1843. Hym. Eur., Vol. I, p. 27. Eleven male specimens, taken at Kartabo, 29-11-1924; 1-III-1924; 10-V-1924; 24- V-1924; Trop. Research Station, New York Zool. Soc., No. 20900; No. 20550; Bartica, 27-V-1924 (2) ; Arakaka, 21-XII-1922 (2) ; one with no data. Reported from Brazil, only, heretofore. Chlorion (Ammobia ) funestum Kohl. Kohl, 1890. Ann. des k.k. naturhist. Hof- mus. Wien, Vol. 5, Heft 3, p. 397. Nine females represent this species in the collection. They were taken at Arakaka, 20- X1I-1922 ; 21-XII-1922 (2); Trop. Re- search Station, New York Zool. Soc., No. 20960; Kartabo, 5-XI-1920; W. Bank Dem. R, 9-II-1923 and 29-III-1924. It has been reported from Brazil. Chlorion (Ammobia) neotropicum Kohl.? Kohl, 1890, Ann. des k.k. naturhist. Hof- mus. Wien, Vol. V, Heft 3, p. 222. I have doubtfully referred four female specimens in this collection to this species as they more nearly agree with it than with any other. Kalacoon, 1916, Bartica Dist., Br. Guiana, Hym. 71 : Trop. Research Station, New York Zool. Soc., No. 21109; two without data. This species has been captured in Brazil. Chlorion (Ammobia) singnlaris Smith. Smith, 1856. Cat. Hym. Ins. Brit. Mus., Vol. IV, p. 261. {Sphex). Two males and three females. Males : Kartabo, 1922; 1917 Penal Settlement Hym. 222; Females: Kartabo, 1922; Arakaka, 21- XII-1922 (2). Distribution — Southern U. S. ; West In- dies; Mexico to Brazil. In this interesting species the males may be entirely black with no ferruginous mark- ings or they have such markings. Their presence with females of Cresson’s Ch. {Ammobia) dubitatum of which no certain males have been found, together with other facts of distribution, etc., makes it almost sure that these are the two sexes of singu- laris. In this lot the males have the fer- ruginous markings. Chlorion (Ammobia) ichneumoneum Linnaeus. Linnaeus, 1758. Syst. Nat., Ed. X, Vol. 1, p. 578, No. 27. {Apis). The thirteen specimens in this collection are as follows. Male: Kartabo, 23-V-1924; Females: Kartabo, 201266, 201285, 21106, 12-V, 18-IV-1919; Bartica District, 28-V- 1924 (2), 30-V-1924, 4-VI-1924, Hym. 125; two with no dates. This widely dispersed and variable species is found in its typical form in the United States. Farther south, increase of the fer- ruginous of the petiole and abdomen, a darkening of the wings and variation of the amount of ferruginous on the legs have been recognized as marking varietal forms which at their extremes have been named. The specimens in this collection are all varieties, but none of them has varied to such an extent as to become a named variety. Distribution — North, Central and South America, at least as far as Brazil, with varieties in the warmer regions. Tribe Sphecini. Sphex Linnaeus. Linnaeus, 1758. Syst. Nat., Ed. X, Vol. 1, p. 569. The insects of this genus were for many years called Ammophila. Sphex abbreviatns Fab. Fabricius, 1804. Syst. Piez., p. 204. (Pelopoeus) . Five specimens of this species are in the collection. Males: Kartabo, No. 20700, 201283, W. Bank Dem. R„ 5-1-1923, 1917 Penal Settlement, Hym. 28; Female: karta- bo, No. 20699. This species occurs from Central Amer- ica to Brazil. Sphex melanarins Dahlb. Dahlbom, 1843. Hym. Eur., Vol. I, p. 15. {Ammophila) . One male, Kartabo, ll-VI-1924. Distri- bution— South America; Brazil. Sphex muticus Dahlb. Dahlbom, 1845. Hym. Eur., Vol. I, p. 431. {Ammophila) . Two specimens. Male: Kalacoon, 1916, Hym. 300; Female: Kartabo, 29-III-1924. This species is reported from Brazil. 32 Zoologica : New York Zoological Society [XXVII: 6 Sphex opulentus Guerin. Guerin, Duperry, 1830. Voy. Coquille, Zool., Vol. II, P. 2, p. 261. ( Ammophila ). This large species, widely distributed in South America, is represented in this collec- tion by five specimens. Male: Kartabo, No. 201135, 23-XI-20; Females: Kartabo, 1922; No. 201135, 23-XI-1929 ; 120. It has been recorded from Colombia to Paraguay, at least. 1942] Smith : Resume of Mexican Tantilla 33 7. A Resume of Mexican Snakes of the Genus Tantilla. Hobart M. Smith Thirty-six species of Tantilla have been described or recorded (including the present article) from North America (with Mex- ico). Five of these have been synonymized with others ( bimaculata with calamarina, boulengeri with miniata, hallowelli with gracilis, kirnia and praeocula with fumi- ceps ). Of the remaining thirty-one, four may be eliminated as not belonging to Tan- tilla. T. lintoni and brevissima have been separated from this genus elsewhere (placed in Tantillita ) . Tantilla depressa is a syno- nym of Geagras redimitus. Tantilla nelsoni Slevin (Proc. Calif. Acad. Sci., ser. 4, vol. 15, 1926, pp. 200-201, Maria Madre Island, Nayarit, Mexico) is certainly not of this genus, differing from all other Tantilla in the possession of an undivided anal, entire nasal, 1-2 temporals, and white bands completely encircling a black body and tail. This combination of characters, together with absence of a loreal, and smooth scales in 15 rows, is unique in snakes, so far as I can determine. Accord- ingly, for this species I propose the generic name Exelencophis.1 The twenty-seven remaining North Ameri- can species belong to five major species groups. These may be further subdivided, forming nine groups, some of which may again be subdivided. The ultimate process of subdivision into the smallest practical super- specific groups reveals fourteen divisions, as shown in the accompanying diagram. The characters of the groups and subdivisions follow. I / i II < 1 { moesta 2 { deppei 3 { mexicana 4 f jani \ striata 1 An effort was made to investigate more fully the char- acters of this curious species, but Dr. Slevin informs me that the type and only example known is lost. Published by permission of the Secretary of the Smith- sonian Institution. r J cuniculator ( phrenitica cleviatrix miniata 6 - rubra coronata wagneri „ J wilcoxi \ rubricata 8 J bocourti 9 [ canula IQ I calamarina ( martindelcampoi 11 [ gracilis 42 < nigriceps \ fumiceps f atriceps 13 4 utahensis l hobartsmithi f planiceps 14 j eiseni ( yaquia I. Belly black; collar very broad. II. Large bodied; ventrals not reduced; head not conical ; lined. b. Dark middorsal line; secondary tempo- ral scale-like. 2. Nuchal collar near parietal. 3. Collar if present on posterior part of neck. c. Light lined; secondary temporal elon- gate. III. Collared; body form as in II; not lined. d. Secondary temporal elongate. 5. Back dark, sharply differentiated from ventral color. 6. Back dark or light but not sharply differentiated from ventral color. 7. Posterior border of collar very nar- row; black of head not emarginate behind eye. e. Second temporal scale-like, broad or long. IV. Head conical; no collar; ventrals reduced; body short; temporal single or secondary temporal scale-like. f. Supralabials normal, seven. g. Supralabials reduced, six. 10. Lined (three dark). 11. Not lined. 34 Zoologica : New York Zoological Society [XXVII: 7 V. Light above ; collar absent or poorly defined. h. Black head cap pointed behind; no evi- dence of a collar. i. Black head cap straight edged posteri- orly, a collar. 13. Head cap extending 2-4 scales be- hind head, also below angle of mouth. 14. Head cap extending 1-2 scales be- hind head, not to angle of mouth. I am indebted to Dr. Thomas Barbour, Mr. Benjamin Shreve, Dr. Joseph R. Slevin and Dr. E. H. Taylor for loan of specimens and other assistance. A number of specimens were secured (including types of wilcoxi rubricates), and the entire study was com- pleted, during tenure of the Walter Rath- bone Bacon Traveling Scholarship. Only diagnoses of Mexican species are given in the following. For diagnoses of United States species not here included, see Blanchard’s synopsis ( Zool . Ser. Field Mus. Nat, Hist,, vol. 20, 1938, pp. 369-376). Tantilla atriceps (Gunther). Homalocranium atriceps Gunther, Biol. Centr. Amer., Rept., 1895, p. 146, pi. 52, fig. B. Tantilla atriceps Amaral, Mem. Inst. Butantan, vol. 4, 1929, p. 218. Type Locality. Nuevo Leon, Mexico. Diagnosis. Body generally light above and below; a dark brown head cap, truncate pos- teriorly, extending one or one and a half scale lengths posterior to parietals ; cap not extending below angle of mouth; a faint light collar covering about one scale length, not bordered posteriorly by a distinct black line ; seven upper and lower labials ; one preocular; two or rarely one postoculars; two elongate temporals, separating labials and parietal ; prefrontals and labials sepa- rated ; mental in contact with chinshields, rarely not (type) ; ventrals 123 to 158; caudals 55 to 70 (males 54 to 70, females 51 to 64). Mexican Localities. Nuevo Leon (types, Brit. Mus.) ; 4 miles west of Saltillo, Coa- huila (EHT-HMS 4555) ; 102 kilometers north of San Luis Potosi (EHT-HMS 23474-5). Specimens Examined. Eight. Remarks. Through the courtesy of Dr. Howard K. Gloyd I have been enabled to ex- amine data on 50 specimens not seen by me, data for which were recorded by Dr. F. N. Blanchard. These show the existence of a great amount of variation in number of postoculars, ventrals and caudals. WThile variation in number of postoculars does not seem to have geographic correlation, the variations in ventral and caudal counts do have. The range of variation in ventral counts in males is 123 to 149; in females, 138 to 158. A comparison with the table of variation given by Blanchard (op. cit., p. 376) yields interesting results. In the nine forms (other than atriceps) listed by him, the maximum range of variation in males is 14, in females 23 (by error?). Since errors in determination of sex frequently occur in these small snakes, more significant is the range of both sexes combined ; in atriceps it is 36, but in other forms it does not exceed 27. Most closely comparable to the range of variation in atriceps is that of coronata (in- cluding wagneri) , which has range of 30. T. atriceps, having higher average counts than coronata, could be expected to have a greater range of variation, and does. Ac- cordingly, the existence of two forms in atriceps is very strongly indicated. Unfor- tunately differentiation in ventral counts in atriceps is not accompanied by differences in color, as in coronata, nor is the area of inter- gradation between the two extremes of ven- tral counts as well defined in atriceps as in the latter. The species rather obviously is in the process of either differentiation of two in- cipient subspecies (species), or of assimila- tion of two previously distinct species. It is difficult to know into which category atriceps should be placed. While the specimens from extremes of the range are easily distin- guished, the character of the species in the broad area between (including northern Coahuila, western Texas) is not known ex- cept from a series from a single locality (Chisos Mts.). These exhibit no great range of variation (20, both sexes), but unfor- tunately straddle the middle of the range of counts for the whole species, so that some are very typical of Mexican specimens, others (the majority) typical of Arizona specimens (range 132 to 151). While it ad- mittedly remains possible that two sub- species may be satisfactorily distinguished when larger series from more numerous lo- calities are available, it does not seem prac- tical to attempt subdivision of the species at present. Tantilla bocourti (Gunther). Homalocranium bocourti Gunther, Biol. Centr. Amer., Rept., 1895, p. 149. Type Locality. Guanajuato. Diagnosis. Light above and below (young may be darker above, but this color shading into white of ventral surface) ; head black or dark brown above ; first and fifth labials, and lower parts of other labials (except seventh) white; internasals white; a white nuchal collar covering one and one half scale lengths on nape, usually not involving tips of parietals, but bordering them; nuchal col- lar divided on midline or not; collar dark- bordered posteriorly by an area one scale in length or less ; seven upper and lower labials; one preocular; two postoculars; two 1942] Smith: Resume of Mexican Tantilla 35 temporals, the anterior elongate and in con- tact with postoculars, the posterior about as broad as long, scale-like; prefrontal sepa- rated from labials ; first lower labials in con- tact medially, occasionally not; ventrals 164 to 185 (195?) ; caudals 46 to 61. Mexican Localities. Known from the states of Jalisco (Magdalena, Guadalajara), Morelos (Cuernavaca, 8 kilometers north- east of Cuernavaca), Guanajuato (Guana- juato), Distrito Federal, Veracruz (Mira- dor), Guerrero (Omilteme), Puebla (22 kilometers north of Tehuacan, 10 miles northeast of Tehuacan), Michoacan (4 miles east of Tuxpam, between Zitacuaro and Rio Tuxpam). Specimens Examined. Thirty-one. Remarks. As pointed out by Taylor & Smith ( Univ . Kans. Sci. Bull., vol. 25, 1939, p. 254) , there is a discernible difference be- tween eastern and western specimens in ventral, caudal and total counts. These indi- cated that two races are recognizable. How- ever, further material has not borne out these supposed differences, but has shown that too great an overlap exists to permit separation. The counts are compared below. Specimens Examined. Six. Remarks. This is related to martindel- campoi, which also has a dark-lined pattern, one temporal, degenerate head scales (fu- sion), and few ventrals and caudals. On the basis of lined pattern, cleppei belongs with this group, but it has normal head scales. Tantilla canula Cope. Tantilla canula Cope, Journ. Acad. Nat. Sci. Phila., ser. 2, vol. 8, 1876, p. 144. Type Locality. Yucatan. Diagnosis. Body light gray above, stip- pled, white below; a middorsal light streak, not well defined; head reticulated, top lighter than sides; seven upper and seven lower labials; one preocular, two postocu- lars; two temporals, the anterior broad and scale-like (nearly as broad as long) , the sec- ondary more elongate ; labials separated from parietals or not; prefrontals and la- bial separated; mental and chinshields in contact or not; ventrals 105 to 114, caudals 36 to 43. Mexican Localities. Several specimens bear the locality datum “Yucatan.” The only definite record is from Chichen Itza. Ventrals Caudals Totals I western 165-180(11) 52-65(11) 219-244(11) males 1 [ eastern 164-176(4) 48-57(4) 221-228(4) { western 169-186(12) 46-59(12) 225-239(12) females I [ eastern 166-173(4) 46-52(4) 216-221(4) Tantilla calamarina Cope. Tantilla calamarina Cope, Proc. Acad. Nat. Sci. Phila., 1866, p. 320. Tantilla bimacidata Cope, Journ. Acad. Nat. Sci. Phila., ser. 2, vol. 7, 1875, p. 143 (Mazat- lan, Sinaloa). Type Locality. Guadalajara, Jalisco. Diagnosis. Body light gray above, white below; three dark dorsal lines, the median spreading over most of top of head; lateral stripes on adjacent halves of third and fourth scale rows, extending onto head and not uniting with median dark area on head ; six upper, six or seven lower labials; one preocular (rarely extremely minute and pre- frontal entering eye), one postocular; one elongate temporal, in contact or not with postocular; prefrontals and labials sepa- rated or not; mental and chinshields in con- tact; ventrals 119 to 132; caudals 27 to 35. Mexican Localities. Known from the states of Colima (Colima, Queseria, Teco- man), Distrito Federal (Santa Fe), Jalisco (Guadalajara), Morelos (4 miles south and 8 kilometers east of Cuernavaca), Nayarit (Sierra de Nayarit, San Bias, El Ocotillo), Puebla (Teziutlan) and Sinaloa (Mazatlan). Specimens Examined. Four. Remarks. As indicated by body form, shape of head, and low number of ventrals and caudals, the species appears to be re- lated to calamarina and its allies. It is con- siderably different from the latter group in having the normal quota of head shields. It shows a great similarity to vermif ormis of Central America. Tantilla cuniculator Smith. Tantilla moesta cuniculator Smith, Zool. Ser. Field Mus. Nat. Hist., vol. 24, 1939, pp. 32-34. Type Locality. Merida, Yucatan. Diagnosis. Dark above, white below; a dark lateral area sharply differentiated from a somewhat lighter middorsal color and from white ventral surface; nuchal collar involving posterior tip of parietals; a very dim dorsolateral light stripe between lateral and dorsal areas ; seven upper and lower la- bials; one preocular; two postoculars; two temporals, elongate, the anterior in contact with postoculars ; prefrontals separated from labials; mental in contact with chin- shields or not; ventrals 140 to 154; caudals 49 to 53. 36 Zoologica : New York Zoological Society [XXVII: 7 Mexican Localities. Known only from the type locality. Specimens Examined. One. Remarks. This species may be related to phrenitica, with which its counts, and some details of coloration, compare well. Tantilla deppei (Bocourt). H omalocranion deppei Bocourt, Miss. Sci. Mex., Rept., 1883, pp. 584-585, pi. 36, fig. 11. Homalocranium deppei Gunther, Biol. Centr. Amer., Rept., 1895, p. 151. Homalocranium miniatum Boulenger (nec Cope), Cat. Snakes Brit. Mus., vol. 3, 1896, p. 222. Type Locality. Mexico. Diagnosis. Pale brown, with a median and two lateral dark lines ; a nuchal collar, com- plete except medially; ventral surface white ; seven upper and lower labials ; one preocular; two postoculars; two temporals separating labials from parietals, the sec- ondary perhaps scale-like (see Bocourt’s fig.) ; prefrontals and labials separated; mental in contact with chinshields ; ventrals 147 to 152, caudals 55. Mexican Localities. Known only from “southern Mexico.” Specimens Examined. None. Remarks. The relationships of this form are difficult to determine. The lined pattern is suggestive of the calamarina group, but the scutellation and body form are different in all important respects. It most closely agrees with the melanocephala group, in- cluding mexicana and armillata. The small posterior temporal corresponds with this ar- rangement, but there is a nuchal collar, which does not occur in other members of the group (with a posterior neck collar) . Tantilla deviatrix Barbour. Tantilla deviatrix Barbour, Proc. Biol. Soc. Wash., vol. 29, 1916, p. 94. Type Locality. San Luis Potosi, San Luis Potosi. Diagnosis. Body light above and below, stippled above (said to be reddish in life) ; top of head light brown anteriorly, becom- ing black posteriorly; a white nuchal collar, involving tips of parietals, covering about two scale lengths, followed by a black border of about equal size; head cap extending laterally only to upper part of last labial, not reaching labial border posteriorly; upper la- bials white, sharply differentiated from dorsal color; an indentation of head cap be- hind eye, reaching to lower edge of upper preocular; seven upper and lower labials; one preocular; two postoculars; temporals elongate, the anterior in contact with post- oculars ; prefrontals and labials separated ; mental and chinshields separated; ventrals 154 to 160; caudals 63 to 66. Mexican Localities. The type locality and Alvarez, San Luis Potosi. Specimens Examined. One. Remarks. This species is well differen- tiated from bocourti by the elongate sec- ondary temporal. From wilcoxi it differs in the mental character (touching chinshields in wilcoxi), greater width of posterior black border of nuchal collar (one scale length or less in ivilcoxi ) ; and in various features of the coloration and the sides of the head. Tantilla hohartsmithi Taylor. Tantilla hohartsmithi Taylor, Trans. Kans. Acad. Sci., vol. 39, 1937, pp. 340-342, fig. 2. Type Locality. La Posa, ten miles north- west of Guaymas, Sonora. Diagnosis. Body light above and below; a black cap on head, truncate and straight- edged posteriorly, extending about one scale length on nape ; black of head not extending below angle of mouth onto gular scales ; a very dim, narrow light collar one scale wide, or less, not dark-bordered posteriorly ; a fine, but very distinct, middorsal black line ; seven upper and lower labials ; one preocu- lar; one postocular (probably varies, one or two) ; anterior temporal relatively broad, but longer than broad, in contact with post- ocular; a secondary and also tertiary tem- poral ; prefrontals separated from labials ; mental separated from chinshields; ventrals 129, subcaudals 53. Mexican Localities. Known only from the type locality. Specimens Examined. One. Remarks. Closely related to atriceps, which agrees in most characters of scutel- lation, and pattern. The latter also occasion- ally shows a faint middorsal dark line, which might appear more obvious (or not) in specimens dried as is the type of hobart- smithi. The latter is tentatively regarded as distinct through its possession of three temporals and the distinct middorsal black line, correlated with the fact that no a triceps have been collected near its type locality. Only further specimens will show whether these characters are invariable. Tantilla jani (Gunther). Homalocranium jani Gunther, Biol. Centr. Amer., Rept., 1895, p. 148, pi. 52, fig. D. Type Locality. Guatemala. Diagnosis. Ground color brown ; three narrow, black-edged white lines, the median involving the vertebral scale row, the lateral stripes involving adjacent halves of the third and fourth scale rows; median stripes 1942] Smith: Resume of Mexican Tantilla 37 sometimes not white, but brown, as ground color, but its dark edges evident ; a complete nuchal collar, about one and one half scale rows wide, involving tips of parietals ; seven labials ; one preocular ; two postoculars ; two elongate temporals, anterior broader than posterior; ventrals 139 to 154, caudals 40 to 50 (in 22 specimens). Mexican Localities. La Esperanza, Chi- apas, and Tapanatepec, Oaxaca. Specimens Examined. Twenty-two. Remarks. The cotypes are two, one from Guatemala, the other from “Hacienda Rosa de Jericho, Nicaragua” {fide Boulenger, Cat. Snakes, vol. 3, 1896, p. 221 ; Gunther says Matagalpa). The Nicaragua specimen has 136 ventrals, and no lateral light stripe, and accordingly I do not believe it is the same as the Guatemala specimen (the one figured by Gunther) which I designate lectotype. The specimens from Finca El Cipres, Vol- can Zunil, Guatemala, reported by Slevin ( Proc . Calif. Acad. Sci., ser. 4, vol. 23, 1939, p. 411) as fusca I have seen and cannot dif- ferentiate from the Chiapas and Oaxaca specimens above mentioned. Comparisons. The present species may be compared with those included by Boulenger in trilineata and fusca. Since the original descriptions indicate that his synonymies of these two are perhaps not wholly correct, jani is compared directly with the original descriptions. T. fusca Bocourt, 1883 (Guatemala) — uniform light brown above, 137 ventrals, more than 34 caudals; Guatemala. Absence of all markings, including collar and longi- tudinal lines, does not describe jani. Bou- lenger’s description of fusca says “Brown above, with or without a light, dark-edged lateral line, with or without a blackish ver- tebral line; head dark brown or blackish above, with yellow markings as in H. melanocephalum.” This remarkable com- posite description is based upon the original description of fusca, two cotypes of jani (which apparently represent two species), and upon a specimen from Duenas, Guate- mala (this is the one with a dark vertebral stripe) referred by Gunther to armillatum, but which is not that at all but possibly is the same as Gunther’s mexicana (or is an unnamed species). The specimen from Car- tago, Costa Rica, referred by Boulenger to melanocephala, was apparently correctly al- located with armillata by Gunther. The lat- ter differs from South American melano- cephala at least in ventral counts. T. trilineata Peters, 1880 (Guatemala, Honduras) — lateral light stripe “running on the third scale row”; whole gular region apparently pigmented, at least the lower labials not heavily nor distinctively pig- mented; no light internasal spot; ventrals 145 ; caudals 41. T. taeniata Bocourt, 1883 (Guatemala) — median stripe continuous, covering at least a full scale width ; dark borders not evident, nor dotted lines on fifth, sixth and first scale rows; lower labials not pigmented; stripes said to be half width of interspace (much less in jani ) ; 149 ventrals ; 68 caudals. T. trivittata Muller, 1885 (Guatemala, Honduras) — said to have six upper labials.-' Median stripe one and one-half scale rows wide, extending to end of tail ; lateral stripes to middle of tail. The species synonymized by Boulenger with virgata appear more distantly related, having the light lines and interspaces of about equal width. Tantilla martindelcampoi Taylor. Tantilla martindelcampoi Taylor, Trans. Kans. Acad. Sci., vol. 39, 1937, pp. 347-348, fig. 6. Type Locality. El Treinta, Guerrero. Diagnosis. Gray, with three broad, dark stripes on body and tail, the median one and two half scale rows wide, spatulate on head; lateral stripes on adjacent halves of fourth and fifth rows, connecting above eye with spatulate median stripe; six upper and seven lower labials; no preocular; one post- ocular ; one elongate temporal, in contact with postocular; prefrontals broadly in con- tact with labials ; mental in contact with chinshields; ventrals 114, subcaudals 39. Mexican Localities. Known only from the type locality. Specimens Examined. Two. Remarks. Related to calamarina (see dis- cussion). Tantilla mexicana (Gtinther). Elapomorphus mexicanum Giinther, Ann. Mag. Nat. Hist., ser. 3, vol. 9, 1862, p. 57, pi. 9, fig. 1. Type Locality. Mexico. Diagnosis. Black head cap extending onto neck several scale lengths, followed by a light ring or a pair of spots; a middorsal dark line, and perhaps a lateral also ; dorsal surface generally brown ; dark area on head reaching to labial border below eye and at the seventh labial; seven upper and lower labials; one preocular; two postoculars; two temporals, the posterior scale-like (but 2 Boulenger keys “trilineata” with the character of eight supralabials, and includes in his synonymy trilineata (de- scribed with seven), taeniata (described with seven) and trivittata (described with six). His only specimen (Bonacca Is., Honduras) was described by Gunther as having seven supralabials. Specimens I have seen do not indicate the basis for such confusion. 38 Zoologica : New York Zoological Society [XXVII: 7 slightly longer than broad) ; mental sepa- rated from chinshields; ventrals 158 to 159; caudals 44 to 52. Mexican Localities. “Mexico.” Specimens Examined. None. Remarks. Boulenger (Cat. Snakes, vol. 3, 1896, p. 216) concludes that the type (la- belled Salle’s collection) was not collected by Salle and is not from Mexico. There are facts, however, which indicate that it may actually have come from Mexico. Although the ventral count is matched by some South American specimens, it is lower than any Central American specimen related to me- lanocephala (i.e., armillata) . Another speci- men from Duehas, Guatemala, referred by Boulenger (op. cit., p. 221) to fusca and by Gunther to armillata very closely matches the characters of mexicana, to which I have referred it. There is little doubt that these are distinct from armillata, and it is scarcely likely that they can be termed the same as the geographically distant melanocephala. In the absence of specimens in American mu- seums it is at present impossible to deter- mine its differential characters. Tantilla miniata Cope. Tantilla miniator Cope, Proc. Acad. Nat. Sci. Phila., 1863, p. 100. Tantilla miniata Cope, Bull. U. S. Nat. Mus., no. 32, 1887, p. 84 (corrects spelling of spe- cific name, which he says was printed minia- tor by “errore typogr.”) . H omalocranium boulengeri Gunther, Biol. Centr. Amer., 1895, pp. 148-149, pi. 52, fig. F (Hua- tusco, Veracruz) . Type Locality. Mirador, Veracruz. Diagnosis. Dark reddish-brown above, the color shaded into that of white ventral sur- face on the first scale row; head black above; a nuchal collar involving tips of parietals, bordered behind by black, cover- ing one to two scale lengths; seven upper and lower labials; one preocular; two post- oculars ; two elongate temporals, the an- terior in contact with postoculars; pre- frontals separated from labials; mental and chinshields in contact; ventrals 159 to 161; caudals 46 to 49 (plus a few) . Mexican Localities. Huatusco and Mira- dor, Veracruz. Specimens Examined. One. Remarks. Related to phrenitica. Gadow’s miniata from Tezonapan (Teconapan, Tex- onapan), north of Ayutla, Guerrero, cannot be the same, but its identity is difficult to guess (Proc. Zool. Soc. London, 1905, p. 196). Boulengers’ incorrect allocation of Cope’s name (to deppei ) accounts for the recognition of boulengeri. Tantilla moesta (Gunther). Homalocranium moesta Gunther, Ann. Mag. Nat. Hist., ser. 3, vol. 12, 1863, p. 352. Tantilla moesta Cope, Proc. Acad. Nat. Sci. Phila., 1866, p. 126. Type Locality. Peten, Guatemala. Diagnosis. Dark brown above, somewhat lighter but still heavily pigmented below; anterior part of head of same color as body ; a very extensive nuchal collar involving over half of the parietals and laterally extending very nearly to the eye; seven upper and lower labials; one preocular; two postocu- lars; two temporals, not strongly elongate, the anterior in contact with postoculars ; prefrontal in contact with labials; mental in contact with chinshields; ventrals 140 to 154; caudals 55 to 63. Mexican Localities. Known from “Yu- catan.” Schmidt & Andrews’ record for Merida, Yucatan, refers to quite a different species ( cuniculator ) . Specimens Examined. Two. Remarks. The species seems to have no very close relatives. Tantilla nigriceps nigriceps Kennicott. Tantilla nigriceps Kennicott, Proc. Acad. Nat. Sci. Phila., 1860, p. 328. Tantilla nigriceps nigriceps Smith, Copeia, 1938, no. 3, pp. 149-150. Type Locality. Fort Bliss, New Mexico. Diagnosis. Body light above and below; a black head cap V-shaped posteriorly, the apex extending over three to five scale lengths on nape; seven upper and lower la- bials; one preocular; two postoculars; two temporals, both considerably longer than broad, neither extremely elongate; primary temporal in contact with postoculars or not ; prefrontals separated from labials or not; mental usually separated from chinshields, occasionally not; ventrals 146 to 161, caudals 35 to 62. Mexican Localities. Known only from Rio Santa Maria, near Progreso, Chihuahua. Specimens Examined. Sixteen. Remarks. The present form, with fumi- ceps, is very distinct from any other species. Tantilla nigriceps fumiceps (Cope). Scolecophis fumiceps Cope, Proc. Acad. Nat. Sci. Phila., 1860, p. 371. H omalocranion praeoculum Bocourt, Miss. Sci. Mpx., Rept., 1883, pp. 582-583, pi. 36, fig. 8 (Colorado) . Tantilla kirnia Blanchard, Zool. Ser. Field Mus. Nat. Hist., vol. 20, 1938, pp. 373-374 (9 miles east of Pleasanton, Atascosa Co., Texas). Type Locality. Unknown, restricted to that of kirnia, 9 miles east of Pleasanton, Atascosa Co., Texas. 1942] Smith: Resume of Mexican Tantilla 39 Diagyiosis. Body light above and below ; a black head cap V-shaped posteriorly, the apex extending over three to five scale lengths on nape; seven upper and lower la- bials; one preocular (rarely two) ; two post- oculars; two temporals, both considerably longer than broad, neither extremely elon- gate; primary temporal in contact with postoculars or not; prefrontals separated from labials or not; mental usually sepa- rated from chinshields, occasionally not; ventrals 130 to 150 (males 130 to 138, fe- males 141 to 150) ; caudals 39 to 48 (males 43 to 48, females 39 to 44). Mexican Localities. Recorded only from Mier, Tamaulipas. Specimens Examined. Seven. Tantilla phrenitica sp. nov. Holotype. U.S. National Museum No. 110381, female, collected at Cuautlapan, Veracruz, August, 1940. Paratypes. Twenty- five, including U.S.N.M. 110379-80, 110382-5, topotypes; U.S.N.M. 38134, Semacock (col- lected by Geo. P. Goll) , Guatemala ; U.S.N.M. 20835, Totontepec, Oaxaca; EHT-HMS 22367, El Limon Totalco, Veracruz; EHT- HMS 23561-70, 23572-76, 23582, topotypes. Diagnosis. Dorsal surface dark, its color sharply differentiated from white or orange ventral surface; centers of lateral scales lighter than their edges ; a nuchal collar, usually involving tips of parietals; snout dark; large postorbital light spot bordering edge of upper lip usually present; four an- terior infralabials dark; seven upper and lower labials ; one preocular ; two postoculars ; two elongate temporals, the posterior more elongate than the anterior, the latter in con- tact with postoculars ; prefrontals separated from labials ; mental and chinshields in con- tact; ventrals 135 to 147; caudals 36 to 42. Description of holotype. Rostral about as high as broad, portion visible from above a little less than half its distance from frontal; internasals about half as long as prefrontals; length of frontal (1.9 mm.) greater than its width (1.5 mm.) or dis- tance from tip of snout (1.8 mm.) , or length of interparietal suture (1.7 mm.), less than maximum length of parietals (2.9 mm.) ; width of frontal about twice width of a supraocular (0.75 mm.) ; nasal completely divided, strongly indented below and some- what less strongly above, anterior section broader than posterior section; preocular large, in contact with nasal; diameter of orbit (0.8 mm.) less than half its distance from tip of snout (2.0 mm.) ; two temporals, the secondary a little longer and narrower than primary; latter in contact with two postoculars (upper fused with supraocular on one side, abnormal) ; seven supralabials, last much the largest, others not much dif- ferent from each other in height; seven in- fralabials, fourth much the largest, four in contact with anterior chinshields ; mental in contact with chinshields ; posterior chin- shields narrower than and about three-fifths the length of anterior pair. Dorsals in 15 rows, smooth, not pitted; 137 ventrals ; anal divided ; 40 caudals ; total length 175 mm. ; tail 33 mm. Color. Dorsal surface of head and body dark brown or black; color of head extend- ing to labial border, except for a light spot (bordering lip) behind the eye, including the fifth supralabial and edges of adjacent scales; a light nuchal collar, involving tips of parietals and of secondary temporals, covering two scale lengths on the nape, somewhat broader laterally and including posterior half of seventh labial; body uni- form dark, except that the centers of the scales are stippled lighter, visible only mic- roscopically on most scales, but easily visible to the naked eye on first row of dorsals; dorsal pigment terminating abruptly at ends of ventrals; anterior portion of mental, and all of four anterior infralabials dark; re- mainder of infralabial border dark ; re- mainder of ventral surfaces of body and tail unmarked; tail and posterior portion of body (ventrally) orange, brightest pos- teriorly, anteriorly becoming white. Variation. In the twenty-five paratypes, one postocular occurs in three; 6-7 infra- labials in three; 6-7 supralabials in one; ventrals 137 to 142, caudals 37 to 44 in twelve males, 135 to 147, and 36 to 42 in thirteen females. The nuchal collar involves the tips of the parietals in fifteen (barely in three), borders the parietals in ten; mental in contact with chinshields in all. Discussion. This species is the same (ap- parently) as Boulenger’s (Cat. Snakes, vol. 3, 1896, p. 222) Guatemala specimen re- ferred to sehistosa. It is not the same as Bocourt’s sehistosa (Miss. Sci. Mex., Rept., 1883, p. 585, pi. 36, fig. 10), however, de- scribed from Alta Vera Paz and Mexico. This species is described as having 121 ventrals (lowest phrenitica count is 135). This difference might be attributed to error in counting, except that sehistosa is figured with a scale-like secondary temporal (al- ways elongate in phrenitica) , and completely without chin markings (mental and four anterior labials always dark in phrenitica) . The Mexican cotype of sehistosa may be- long to phrenitica (as I provisionally con- clude), but if so it certainly is not the fig- ured and described specimen, which must be considered the type and which probably is the Alta Vera Paz specimen. 40 Zoologica : New York Zoological Society [XXVII: 7 Tantilla rubra Cope. Tantilla rubra Cope, Journ. Acad. Nat. Sci. Phila., ser. 2, vol. 8, 1876, p. 144. Type Locality. “Japana,” Oaxaca (Tapana) . Diagnosis. Red (pink) above and below; snout, including internasals and half of pre- frontals, white; rest of sides and top of head black, except for area involved in nuchal collar and a w'hite circular area in- cluding the fifth labial and edges of adjacent scales; collar broad, including tips of pari- etals, extending anterolaterally to middle of seventh labial, extending posteriorly over two or three scale lengths ; posterior border of collar black, extending over two or three scale lengths ; seven upper and lower labials ; one preocular; two postoculars; two elon- gate temporals, the anterior in contact with postoculars; prefrontals separated from labials, occasionally not; mental and chin- shields in contact; ventrals 148 to 162, caudals 59 to 68. Mexican Localities. Recorded from Santa Efigenia, Tapana, and various localities in the vicinity of Tehuantepec, Oaxaca; and 22 kilometers north of Tehuacan, Puebla. The British Museum specimen from Orizaba (Boulenger, Cat. Snakes Brit. Mus., vol. 3, 1896, p. 219) does not seem to be rubra, nor can I allocate it with any other. It may be bocourti. Guanajuato records, of course, are incorrect. Specimens Examined. Eleven. Tantilla striata Dunn. Tantilla striata Dunn, Amer. Mus. Nov., no. 314, 1928, p. 3. Type Locality. Mixtequilla, Oaxaca. Diagnosis. Three light lines on body, the median involving one and two half scale rows, the lateral involving adjacent halves of third and fourth rows; no nuchal collar, replaced by two rounded light spots involv- ing tips of parietals; belly light; seven upper and six or seven lower labials; one preocular; two postoculars; two elongate temporals, the anterior broader than the posterior and in contact with postoculars; prefrontal separated from labials ; mental in contact with chinshields; ventrals 157 to 165; caudals 34 to 42. Mexican Localities. Known only from the vicinity of the type locality. Obviously re- lated to jani. Specimens Examined. Three. Tantilla wilcoxi wilcoxi Stejneger. Tantilla wilcoxi Stejneger, Proc. U. S. Nat. Mus., vol. 25, 1902, p. 156. Type Locality. Ft. Huachuca, Arizona. Diagnosis. Body light, above and below; head dark gray above and on sides as far as lower border of eye; a lateral extension of head cap to labial border, including parts of sixth and seventh labials ; black cap scarcely indented behind eye; snout pigmented; col- lar involving posterior tips of parietals, covering one to one and one-half scale lengths; a narrow black posterior border covering one scale length or less; seven up- per and lower labials; one preocular; two postoculars ; temporals elongate, the an- terior in contact with postoculars; pre- frontals and labials separated; mental and chinshields usually in contact, occasionally not; ventrals 149 to 164; caudals 62 to 69. Mexican Localities. Mojarachic, Chi- huahua. Specimens Examined. Six. Remarks. This species is well differenti- ated from bocourti, which it resembles in color, by the shape of the secondary tem- poral. The narrow posterior border of the light collar, lack of sharp differentiation be- tween white labial area and dark coloration of top of head, and absence of an indentation in the dark area behind eye differentiate this species from others of the group. Tantilla wilcoxi rubricata subsp. nov. Holotype. U. S. National Museum No. 110399, male, fifteen miles southeast of Ga- leana, Nuevo Leon. Paratypes. U.S.N.M. No. 110398, topotype; E. H. Taylor - H. M. Smith No. 23473, Mt. Zapaliname, Saltillo, Coahuila. Diagnosis. Top of head dark gray, the color extending a little below level of eye, not very sharply differentiated from white of supralabial region ; dark head color ex- tending laterally to labial border on fifth and sixth supralabials ; a narrow nuchal collar, covering little more than one scale length, involving tips of parietals (barely in one case) ; black border of nuchal collar very narrow, less than one scale length ; head scales normal; ventrals 140 to 146; caudals 51 to 56 ; mental in contact with chinshields. Description of holotype. Portion of rostral visible from above equal to length of inter- nasals; latter little more than one-third length of prefrontals, about a fourth their size; frontal with four sides, two meeting in an obtuse anterior angle, sides curved, posterior angle acute; length of frontal (2 mm.) greater than its distance from tip of snout (1.8 mm.) or length of interparietal suture (1.6 mm.), much less than greatest length of parietal (2.9 mm.) ; maximum width of supraocular (0.9 mm.) more than half width of frontal (1.5 mm.) ; nasal ap- parently divided both above and below, naris pierced nearer upper edge than lower; an- terior section of nasal larger and higher than posterior, which is elongate and in con- tact with preocular; two postoculars; two 1942] Smith: Resume of Mexican Tantilla 41 elongate temporals, anterior in contact with postoculars; seven supralabials, last largest; seven infralabials, the first in contact with midventral line on one side; four labials in contact with anterior chinshields, one with posterior. Scales in 15 rows throughout, smooth, not pitted; ventrals 140; anal divided; cau- dals 54. Color. Dorsal surface brownish gray, stippled lightly; ventral surface white; head dark gray above, lighter on snout, the color extending on sides to below eye; head cap black around edges, extended laterally to labial border on portions of sixth and sev- enth supralabials; labial border white, not sharply differentiated from dark color; a small indentation of head cap behind eye, not reaching upper postocular; a narrow white nuchal collar of about one and one- half scale lengths, involving tips of pari- etals ; collar bordered posteriorly by a nar- row black band covering less than one scale length ; lower labial border stippled. Variation. The topotypic female is very similar in color, markings and scutellation. The mental is in contact with both anterior chinshields. Ventrals 146, caudals 51. The specimen from Mt. Zapaliname has six supralabials, the normal sixth and seventh fused; six infralabials on one side (second and third fused) ; nasal not divided above naris ; nuchal collar barely involving ex- treme tips of parietals; some black stippling on some of the median belly scales; ventrals 144, caudals 56. The black posterior border of the nuchal collar is very narrow, nearly obsolete. Despite these numerous differ- ences I believe this is the same as the Ga- leana specimens. It is obviously anomalous in some characters, and in addition has several of the posterior ventrals divided medially. Tantilla yaquia sp. nov. Holotype. Museum of Comparative Zoology No. 43274, female, collected at Guasaremos, Rio Mayo, Chihuahua, by H. S. Gentry, in August, 1936. Diagnosis. Dorsal surface light brown, light below; head dark brown, black on sides and posteriorly, the cap extending about three scale lengths posterior to parietals; a faint nuchal collar, about one scale length, without a dark posterior border; a white area bordering lip behind eye, extending to middle of primary temporal; cap extending below angle of mouth; naris nearer upper than lower edge of nasal; latter divided above and below; ventrals 150, caudals 66. Description. Length of portion of rostral visible from above about equal to length of internasals ; latter a little less than half size of prefrontals ; frontal hexagonal, anterior angle obtuse, posterior angle a little less than right angle, sides slightly convergent posteriorly; frontal a little longer (2.2 mm.) than broad (1.8 mm.), longer than its dis- tance from tip of snout (1.8 mm.), subequal to length of interparietal suture, consider- ably less than maximum length of parietal (3.2 mm.) ; maximum width of supraocular (0.9 mm.) no more than half width of frontal; naris pierced somewhat nearer up- per than lower edge of nasal ; posterior sec- tion of nasal about two-thirds size of an- terior section, subequal to single preocular; two postoculars; two elongate temporals, the primary broader and in contact with post- oculars; seven infralabials, fourth and fifth entering eye, 1-2-3-5-4-6-7 in order of in- creasing size, seventh much the largest; seven infralabials, four in contact with an- terior chinshields; mental in contact with chinshields, the anterior pair of which is longer than the posterior. Scales in 15 rows, smooth, not pitted; ventrals 150; caudals 66. Total length 240 mm.; tail 61 mm. General ground color light brown above, clear below; top of head dark brown, tip of snout a little lighter; sides of head black, except lower parts of subocular and second labial, all of first, fifth and sixth labials, anterior half of seventh labial and lower half of primary temporal, all of which are cream-color; black on sides of head continu- ous with a black area extending posteriorly to cover about three scale lengths of neck; latter area extending laterally below corner of mouth onto first row of gulars, involving posterior half of seventh labial; posterior border of head cap straight on dorsal sur- face, curving forward on sides of head ; a light collar about one scale wide, poorly de- fined, bordering head cap posteriorly, fol- lowed by brown stippling of the back, its posterior edge not otherwise marked ; stip- pling on back light, most distinct at edges of scales; some stippling at ends of ventrals, and in gular region; ventral surface other- wise unmarked. Comparisons. The present form is closely related to eiseni and planiceps. T. eiseni has the naris equidistant from upper and lower edges of nasal ; ventrals more numerous (165 to 190) ; no light area behind eye, ex- tending from labial border to middle of pri- mary temporal ; head flatter, snout more protruding. This and yaquia seem to be more closely related to each other than either is to planiceps, which has low ventral counts and nasal not divided above naris. Key to North American Tantilla. 1. Ventral surface heavily pigmented .moesta Ventral surface light, only extreme lateral tips of ventrals sometimes pigmented . . 2 2. A lateral light stripe (a median also, but 42 Zoologica : New York Zoological Society [XXVII: 7 sometimes dim) at least anteriorly; no median middorsal dai'k stripe 3 No lateral light stripes; or, if present, a middorsal dark stripe also present. . . .4 3. Two light spots on nuchal region, involv- ing tips of parietals; ventrals 157 to 165, caudals 34 to 42 striata A light nuchal collar, involving tips of parietals; ventrals 139 to 154, caudals 40 to 50 jani 4. A lateral light stripe and a middorsal dark stripe; neck as well as head dark, collar (if present) crossing neck several scales back of head mexicana Not as described 5 5. Body with three or five distinct dark stripes 6 Body not striped ; or, if so, with only a median stripe 8 6. A light nuchal collar (may be broken me- dially) ; seven supralabials; five dark lines on body deppei No light collar; six supralabials; three dark lines 7 7. Lateral stripe on adjacent halves of fourth and fifth rows; no preocular; ventrals 114; caudals 39 martindelcampoi Lateral stripe on adjacent halves of third and fourth rows; a preocular (sometimes very minute) ; ventrals 119 to 132; cau- dals 27 to 35 calamarina 8. Dorsal surface of body dark and the dark color sharply differentiated from lighter ventral color at ends of ventrals 9 Dorsal surface of body dark or light, but if dark, the color not sharply differentiated from lighter ventral color at ends of ventrals 10 9. Sides of body darker than middorsal area, the two areas rather sharply differen- tiated and separated by a dim, narrow, light line; ventrals 140 to 154; caudals 49 to 53 cuniculator Sides of body not darker than nor dis- tinctly differentiated from color of mid- dorsum; ventrals 121 to 147; caudals 32 to 43 phrenitica 10. Head of about same color as back or, if darker, its color not sharply differen- tiated from that of back; and no light nuchal collar 11 Head black or dai'k brown, sharply differ- entiated from color of back, or, a light nuchal collar present 12 11. Sides of head of same color as top; a dor- sal light stripe (in fresh material) ; seven supralabials ; two postoculars ; ventrals 105 to 114 canula Labial border white, rather sharply differ- entiated from dorsal color of head; no dorsal light stripe; usually six supra- labials; usually one postocular; ventrals 115 to 138 gracilis 12. Black head cap V-shaped, its apex extend- ing three to five scale lengths posterior to parietals on middorsal line; no nuchal collar; body light 13 Black head cap with a straight or slightly convex posterior margin ; a nuchal collar at least faintly visible, not necessarily dark-edged behind ; body dark or light . 14 13. Ventrals in males 146 to 159, in females 150 to 161 nigriceps nigriceps Ventrals in males 130 to 138, in females 141 to 150 nigriceps fumiceps 14. Nuchal collar indistinct, of about one scale length, not dark-edged posteriorly (some- times a few black dots) 15 Nuchal collar very distinct, usually of more than one scale length, dark-edged posteriorly 20 15. Black of head extending below angle of mouth onto gular scales, and posteriorly two or more scale lengths 18 Black of head not extending below angle of mouth onto gular scales, and posteri- orly two or less scale lengths 16 16. A fine, distinct, black, middorsal line; three temporals; ventrals 129 in type; usually (?) one postocular hobartsmithi No middorsal dark line; two elongate tem- porals; ventrals 127 to 160; usually two postoculars 17 17. Ventrals in males 153 to 160, in females 163 to 172 utahensis Ventrals fewer atriceps 18. Ventrals 134 to 141; nasal not divided above naris planiceps Ventrals more numerous; nasal divided above naris as well as below 19 19. Ventrals 165 to 190 ; naris equidistant from upper and lower edge of nasal ; no light area bordering lip behind eye and ex- tending to middle of primary temporal eiseni Ventrals 150 in female type; naris much nearer upper than lower border of nasal ; a light area bordering lip behind eye, extending to middle of primary tem- poral yaquia 20. Anterior temporal about as broad as long, scale-like bocourti Anterior temporal elongate, much longer than broad 21 21. Posterior black border of nuchal collar covering one scale length or less 22 Posterior black border of nuchal collar covering two or three scale lengths ... 23 22. Ventrals 149 to 164, caudals 62 to 69 wilcoxi wilcoxi Ventrals 140 to 146, caudals 51 and 56 wilcoxi rubricata 23. Black of dorsal surface of head not reach- ing labial border deviatrix Black of dorsal surface of head reaching labial border 24 24. Nuchal collar covering two or three scale lengths, about as wide as its posterior border; snout white; dorsal and ventral surfaces pink rubra Nuchal collar much narrower, narrower than its posterior border; snout not white 25 25. Ventrals in males 131 to 141, average 135; in females 139 to 148, average 143; cau- dals in males 42 to 51, average 46; in females 41 to 46, average 44 ; light band on back of head well defined although often interrupted on the midline coronata coronata Ventrals in males 119 to 129, average 127 ; in females 123 to 145, average 131 ; cau- dals in males 50 to 67, average 57, in females 41 to 59, average 51 ; light band on back of head usually more or less obliterated coronata wagneri 1942] Beebe: Atlantic and Pacific Dixonina 43 8. Eastern Pacific Expeditions of the New York Zoological Society. XXX. Atlantic and Pacific Fishes of the Genus Dixonina.1 William Beebe. (Plates I & II) . [This is the thirtieth of a series of papers dealing with collections of the Eastern Pa- cific Expeditions of the New York Zoological Society made under the direction of Wil- liam Beebe. The present paper is concerned with specimens taken on the Eastern Pa- cific Zaca Expedition (1937-1938). For data on localities and dates of this expedition, refer to Zoologica, Vol. XXIII, No. 14, pp. 287-298.] In connection with studies of differentia- tion or of identity in species of fish from opposite sides of North and Central Amer- ica, I have reviewed the data on Dixonina. This consists of the published account by Fowler, of the type of Dixonina nemoptera, taken many years ago at Santo Domingo, West Indies, and of the second known At- lantic specimen by Metzelaar, from Puerto Cabello, Venezuela ( not Curacoa, as re- corded by Dr. Myers.). As regards Pacific records of Dixonina, Dr. George S. Myers, in 1936, searched the collections of the United States National Museum. He found a single individual of fifteen inches standard length, which long before had been acquired by the Museum, with no data except the locality Acapulco, Mexico. No description was published. In the Japanese publication “Marine Fishes of the Pacific Coast of Mexico,” 1937, with text by Yosio Hiyama, we are told in the preface that all the fish mentioned were collected in 1935-1936 on the trial fishing trip of a Japanese boat, and all taken by trawling. There are four lines of text on page 27 which refer very evidently to Albula, with the note that this species was “abundant in the Gulf of California.” We 1 Contribution No. 636, Department of Tropical Re- search, New York Zoological Society. find, however, the single sentence, “in some specimens the last rays of dorsal elongated.” In agreement with this, Plate Five presents an excellent colored figure of Dixonina, both as to correct relative length of the maxil- lary, number of lateral line scales and in the two elongate fin rays. So we must recog- nize this as the second Pacific published rec- ord of this species. The length of this speci- men seems to have been 270 mm. Along the northern part of the Pacific coast of Costa Rica, on the Eastern Pacific Zaca Expedition of the Department of Tropical Research, we rediscovered, in 1938, this interesting relation of the bonefish. At three places along a stretch of shore of about seventy-five miles we took a total of 19 Dixonina, both in seines and by hand line from the Zaca, the fish measuring from 80 to 365 mm. standard lengths. The two Atlantic specimens measure 381 and 260 mm. standard lengths respectively. With the published data of these I have com- pared two of my Pacific Dixonina, one of 352 and the other 365 mm., and the accompany- ing table gives the comparative results. In the same table are corresponding characters of two additional Pacific Dixonina, of 185 and 80 mm. standard lengths. Dixonina pacifica sp. nov. Type: Holotype, No. 26,131, Eastern Pa- cific Zaca Expedition of the Department of Tropical Research, New York Zoological So- ciety; Port Culebra, Costa Rica (No. Lat. 10°31'; West Long. 85°40') : caught on hand line from the Zaca, January 24, 1938; stand- ard length 352 mm. ; adult female in full breeding condition. Type in the collections of the Department of Tropical Research, New York Zoological Society. 44 Zoologica : New York Zoological Society [XXVII: 8 Diagnosis: The distinction between the Pacific Dixonina pacifica and the Atlantic Dixonina nemoptera is as follows. In the for- mer there is a greater number of lateral line scales (81 to 84, as compared with 76) ; in- creased number of vertical lines of scales on dorsal half (11, not 9), and on ventral half (9, not 7) ; decrease in predorsal scales (22 instead of 30) ; smaller eye (6.8 to 7, not 4.15) ; longer pelvic fins (2.2 and not 2.7) ; greater number of gill-rakers (7 11 in- stead of 4 9) . Identical characters in Dixonina from the two oceans are, relative head length, depth, snout, maxillary, lengths of the first and last dorsal rays and the last anal ray, pec- toral length, branchiostegals, and appar- ently fin counts, although as regards those of the dorsal and anal there is considerable confusion. Measurements and Counts: Standard length 352 mm. ; depth 80 ; head 120 ; eye 17; snout 44; maxillary 54; interorbital 18; pectoral length 54; pelvic length 54; snout to dorsal 195; snout to anal 330; snout to pectoral 117; snout to pelvic 247; dorsal height 60; last dorsal ray length 62; anal height 32; last anal ray length 40 mm.; dorsal fin count 11,20; anal fin count 1,9; pectoral count I, 16; pelvic count I, 10; gill- rakers 7-11; branchiostegals 14; lateral line scales 81 ; scale rows dorsal to lateral line 11; lateral line to anal 9. Range: Northeastern Pacific coast Mex- ico: (Gulf of California and Acapulco); Costa Rica: (Potrero Grande, Port Culebra and Piedra Blanca Bays) . Field Characters: An elongate, spindle- shaped fish of the eastern Pacific coast, with conical snout overhanging the mouth; maxil- lary longer than snout; last ray of dorsal and of anal fin produced into a long fila- ment ; shining silver with dark scale lines along upper half of body. More than 80 lateral line scales. Color: Brilliant silver, appearing dark above in certain lights with dark green on the head ; this dorsal pigmentation resolves into eight very dark lines along each side of the upper half of the body, covering about one-fourth of each scale nearest to the ad- joining line. Dorsal fin greenish, caudal dusky, anal silvery at base; a dusky spot at pectoral base, with yellow spot behind ; basal membrane of pectoral apple green, rest of fin dusky. These colors fade at death. Iris silvery. In the young fish, at least up to 90 mm. in length, two rows of dark spots extend along the side of the body, the upper close to the mid-back. In a 115 mm. fish the spots are less conspicuous and the dorsal dark lines begin to be distinct. The spots persist after death. In the full-grown fish the dark scale lines are fainter than in fish of medium size. Size: The largest recorded fish is that in the U. S. National Museum collection, “a fine 15-inch adult,” (381 mm.). Local Distribution: Wherever we found these fish, they inhabited the same coastal shallows, off sand or muddy shores, as Albula. Abundance: Common wherever found; twelve taken in one seine haul, and three and two on successive days with hand lines. Food: A 90 mm. fish (28,051 b) taken at Potrero Grande, Costa Rica, had in its stom- ach 1 mysid, 1 shrimp and 1 euphausid, all small. The food of another fish of 179 mm. from the same locality was an Atherina sp. of 40 mm. and an 80 mm. Anchoviella sp. The fully adult female (26,131) had, in its intestine, remains of a small fish and a small anomuran crustacean. In the stomach proper was a freshly swallowed and quite un- damaged Squilla hancocki, the fifth known specimen. Breeding: The type, specimen Number 26,131, standard length 352 mm. was taken by hand line from the Zaca in Port Culebra, Costa Rica, January 24, 1938. It proved to be a full-grown female with ovaries well de- veloped. The eggs seem almost ready for deposition. The two ovaries are large, wrapped about the stomach and intestine, fairly thick down the mid-dorsal line of the coelom, and thinning out into sheets or wings around the caeca and stomach, almost meeting ventrally. They are 170 mm. in length and 30 mm. at their widest extent around the caeca. They weigh together 45 grams, and a carefully estimated count, based on divisional weight, works out at about 70,000 eggs. Individual eggs measure .35 to .45 mm. in diameter. Study Material: 19 specimens. Costa Rica: Potrero Grande, 11 transitional ad- olescents (28,051, a, b, c, d, e; 26,046, a, b, c, d, e, f), 80-200 mm., 3 of which were in the immature spotted stage, seine; and 3 ad- ditional adolescents (26,123, 26,124 and 26,125), 220-270 mm., taken on hand line from Zaca, January 23, 1938. Culebra Bay, 2 adults (26,131, and one 26,131 a, lost after description), 352 and 365 mm., January 24, 1938, on hand line from Zaca; Piedra Blanca, 3 transitional adolescents (28,746 a, b, c), 80, 110 and 115 mm., all in spotted phase, February 2, 1938, seine. References : Dixonina nemoptera, Myers, G. S. Copeia, 1936: 83-85 (Announcement of discovery of a Pacific Dixonina in the col- lections of the U. S. National Museum, labelled Acapulco). Albida vulpes (in part) Kumada & Hiy- ama, Marine Fishes Pacific Coast of Mex- ico, 1937: 27, Plate 5 (short description, referring in part to Dixonina; Gulf of Cali- fornia, plate of Dixonina ) . 1942] Beebe: Atlantic and Pacific Dixonina 45 Table 1. Comparison of proportions and counts of Atlantic Dixonina nemoptera and Pacific Dixonina pacifica. Atlantic Standard length 381 260 Head length 109 108 Head (in length) 3.5 2.4 Depth (in length) 4.6 Dorsal fin V, 16, I 19, 1 Anal fin 111, 6, I 7, 1 Pectoral fin 1, 16 Pelvic fin I, 8 L.l. scales 76 76 Dorsal to L.l. 9 9 Anal to L.l. 7 7 Predorsal scales 30 Snout (in head) 2.6 2.5 Eye 5.6 5 Maxillary 2 2.1 Interorbital 4.15 1st dorsal ray 1.8 1.9 Last dorsal ray 1.7 Last anal ray 3.25 Pectoral length 2 Pelvic length 2.7 Gill-rakers 4 + 9 Branchiostegals Vertebrae 14 Ontogenetic Changes. The following are the ontogenetic changes apparent in this species, based on three in- dividuals, a small transitional adolescent in the immature spotted phase of 80 mm., a late striped adolescent of 185 mm., and a fully breeding adult measuring 352 mm. in standard length. The head increases slightly in comparison with the length, from young to adult, 3.9 times to 2.9; a corresponding change in depth is more marked, the increase being from 6.6 times to 4.4; The median fin count remains the same at all ages, although hints of two additional anterior spines in the anal were detected in a cleared 110 mm. adoles- cent; also in the same fish the pectoral count was distinctly increased to 1,18 in- stead of the usual 1,16, and the pelvic showed 1,11 elements instead of the more typical 1,10. The eye decreases relatively with age, measuring 4.6 times in the head in the youngest, as compared with 7 times in the oldest. The elongate last dorsal and anal rays increase markedly with age, the former changing from 5.3 to 1.9 times in the small- est and largest fish, and the anal filament from 6 to 3 times in the head. Ontogenetic changes in other characters are negligible. Median Fins. In regard to the moot question of whether the last two, closely associated rays of the dorsal and the anal fins should be counted as one, as one and a half, or as two seems to me to reduce itself automatically to a ques- Pacific 352 365 185 80 mm. 120 115 52 23mm. 2.9 3.1 3.5 3.9 4.4 5 5 6.6 11,20 II, 20 II, 20 1,9 I, 9 1,9 I, 16 I, 16 I, 16 I, 10 I, 10 I, 10 81 84 84 80 11 11 11 10 9 9 9 9 22 22 23 23 2.7 2.6 2.6 2.7 7 6.8 5.7 4.6 2.2 2.1 2.2 2.1 6.6 5.1 4.7 2 1.8 1.8 1.9 1.5 2.6 5.3 3 3.2 6 2.2 2 2.1 2.2 2.3 2.6 7 + 11 7 + 11 7 + 11 7+11 14 12 77 tion either of the recognition of natural evolution or of personal preference. My choice is to consider them as two rays. In addition to any phylogenetic interest ex- pressed in the elongated posterior ray of the median fins, it is an important differential generic character. It is also singularly per- sistent, although appearing sporadically in several genera of Isospondyli, such as Tar- pon, Dixonina, Dorosoma, Signalosa, Opis- thonema and Chanos. Thus if we should con- sider the posterior, closely-associated rays as one, we should have to describe the elon- gated character as formed by one-half of the posterior ray. Judging by many metameral characters of these and other fish, and by general evo- lutionary knowledge, we know that phvlo- genetically, and often ontogenetically, such structures usually decrease in serial num- ber. In the present case, we can be quite sure that the close association of the two pos- terior rays of the median fins is due to a comparatively recent reduction, and that at one time both were wholly separate as to subdermal elements. The same thing holds in the case of the present, intimately asso- ciated, but clearly distinguishable anterior spine or spines. As these spines, present in both dorsal and anal fins, are extremely hard, and wholly lack the nodes or segments characterizing all rays, we must omit the conventional “Isospondyli — Dorsal and anal without true spines.” In my discussion of the supporting bones of the median fins, instead of using the 46 Zoologica : New York Zoological Society [XXVII: 8 cumbrous “proximal or distal interspinous bones,” or “Pterygiophores”, or “inter- neural and interhaemal bones”, I am mak- ing use of Jordan’s terms, baseosts and axonosts, which seem simple and self- explanatory. Anal Fin. Based chiefly on a 185 mm. specimen, No. 28,046 b. With a few references to a 170 mm. fish (28,046 a), and to a 110 mm. fish (28,746). The normal count of the anal fin in this species, both in fully adult breeding indi- viduals, and in transitional adolescents from 200 down to 80 mm. standard lengths, is 1-9. In one specimen of 110 mm., however, I have found distinct traces of two minute spines anterior to the definite one which we count as I. No separate axonosts exist for the anterior four elements (the spine and three rays), although the anterior one flares out in a curved, double wing, and a distinct seam indicates that they, the next two, are fused together and to the distal head of the compound baseost. The first three baseosts (belonging to the spine and two rays) are fused at their distal ends, the second and third fanning out and forming the anterior and posterior edges of a large, thin, longi- tudinally oriented, bony wing or fan. The 1st baseost (serving the spine) is a minute, bony nubbin (.8 mm.), attached to the anterior, distal portion of the 2nd baseost. The small hard spine (2.3 mm.) is closely applied to the anterior surface of the 1st ray. The 2nd baseost (2.5 mm.) forms the anterior border of the above-mentioned wing or fan ; its ray is 6.5 mm. long. The 3rd baseost (6.5 mm.) forms the posterior bor- der of the wing, and its ray which is the first branched one, is 14 mm. The 4th baseost is like all the succeeding ones, but its axon- ost is absent, the 4th ray articulating di- rectly with the end of the baseost. From here back, we find five baseosts, long, slender, but rather blunt, each fringed, fore and aft, by a fin or wing of thin bone. Each fin almost touches the one in front and that behind, so there can be extremely little inter-baseost movement, even if the muscu- lar tissues permitted. Each baseost points sharply forward, the two last being even more acutely oblique, thus bringing the three posterior proximal tips close together. Each baseost articulates closely with its axonost. These are saddle-shaped, and each is partly covered with the pair of large, blunt spines which extends down and out from near the base of the rays. The 8th ray, while furnished with normal axonost and baseost, is placed somewhat above the level of the rest of the rays, and its articular surface is at a greater dis- tance from the corresponding surface of the axonost. The 9th or last ray, the elongated one, ends basally in two, long, slender, curved spines which extend down toward, but do not quite bestride, the last axonost. This is almost double the size of the rest, and in all nearly grown or adult fish completely lacks a baseost. In a 110 mm. Dixonina, how- ever, a faint, cartilaginous one is visible, one-third as long as the others. In this in- dividual, too, both of the last two rays are elongate, and share equally in this generic character. They measure 7 mm. in length, compared with 4.3 mm. of the 7th ray. The inter-rayal webbing is extremely deli- cate and is easily destroyed. From each side of each ray there arises a flap or narrow, vertical curtain of scales, attached only along the side, and hanging freely behind, about twice or three times as wide as the ray stem itself. Thus each ray possesses a web connecting it with the succeeding ray, and in addition, two longitudinal scale flaps which lie flat against the web, but can be lifted up at right angles and replaced again. The two posterior rays are wholly free from these scales, but the ray in front (7th in the anal) has three scales in a vertical line, forming its flap, and this number increases to a dozen or more along the long, anterior rays. Dorsal Fin. The count of the dorsal fin is consistent in both adult and young Dixonina pacifica, 11-20. Except for its extra spine and nine additional rays, the dorsal fin is a replica of the description I have given of the anal fin. In a 185 mm. fish the lengths of the ele- ments are as follows : First spine, very small (1.5 mm.) ; 2nd spine 2.3 mm.; 1st ray, 6 mm. ; 2nd ray, 12.3 mm. ; 3rd ray, 21.3 mm. ; 4th, or first branched ray, 28.6 mm.; 19th ray, 6.7 mm. ; and 20th or elongate ray, 19.3 mm. in length. Gill-arches. For gill-arch comparison I have chosen three individual fish, of 80, 185 and 352 mm. standard lengths, or approximately 25, 50 and 100 per cent. In the 80 mm. adolescent, the conventional gill-raker count of the first arch is 7-11, and in the 352 mm. fully adult individual it is the same, the anterior two or three on the hypobranchial being some- what indistinct from a concentration of surrounding spinous areas. Of the rakers on the lower arch, the ceratobranchial bears 7, the hypobranchial the remaining 4. Most of the following notes have to do with these 7, as they are typical of those on all other segments. In the smallest fish the 7 ceratobranchial rakers are the only armature on the outer side of this element, and are relatively long, straight-sided and covered solidly with long, sharp spines. Along the summit of the ridge 1942] Beebe: Atlantic and Pacific Dixonina 47 of this arch is a series of 6 narrow groups of spines, the profile view of the ridge re- sembling a jaw with wavy line of teeth. The inner side of the arch is a replica of the outer, 7-11, but with the individual rakers much smaller, and alternating with those on the outer side. In the 185 mm. fish the rakers have be- come somewhat shorter and less erect, with the spines short, except those at the summit which are long and curved inward. The line of ridge spines has grown down, sending triangular areas over each side, alternating with the rakers proper. This stage is very similar to that found in the adults of some of the pike perches (Stizostedion) . Along the hypobranchial these form a closely interlocked pattern of alternating triangles. These almost fuse with the dental armature of the basibranchial. Along the periphery of this latter area the teeth are thick, sharp and peg-like, giving place al- most at once to the rounded molars typical of the large areas of the oral cavity. In the adult fish the rakers are short and thick, projecting only slightly above the level of the ridge, and with all the spines short. The descending areas have reached half way down the side of the arch, and be- low them, and between the bases of the rakei’s, new, small, spiniferous patches have developed. On the hypobranchial the patches are so continuous that only with difficulty can all 4 rakers be distinguished. In fact the adult arch is almost solidly covered with a spiny coat, out of which the original rakers project as low mounds. Relative to the length of the fish, the rakers and gill-filaments are longer in the young than in the adult. Teeth. The teeth, as developed in an old transi- tional adolescent of 185 mm., are of two distinct types, first, sharp and curved ones, in rows or small clusters, but not villiform; and, second, low, blunt molar-like teeth in patches. The former are well-developed on the premaxillary and the mandible. In the young and half-grown fish they are in two distinct rows, but in the adults these be- come less apparent, and may merge into wider bands, 4 or 5 teeth deep. There is a slightly curved row of teeth on the vomer, which laterally merge into the narrow, elon- gate patches of the palatines. In the back of the mouth, the 1st epibran- chial shows only a very slight concentration into a patch, but on the 2nd, 3rd and 4th arches there are developed increasingly larger pharyngeal-epibranchial patches of teeth, all of the same type as those on the jaws. These, especially the ones on the 3rd and 4th arches, are directly opposed to the two large, triangular patches on the 5th ceratobranchials. The molar type of teeth is segregated in mid-mouth, in three large, palatal patches, the central one on the parasphenoid, and the two lateral on the entopterygoids (not, as Fowler has them in his type description of Dixonina nemoptera, on the sphenoid and pterygoids). The tongue is fleshy with only the most minute rugosities. The 1st basi- branchial is armed solidly with a great con- vex mass of the molars. The anterior half of the 3rd basibranchial is similar. The two types of teeth are everywhere distinct in the 185 mm. fish, except on the 3rd basibranchial patch where they are in- termediate, pointed, but much thicker and coarser than the jaw teeth. They are on their way to the change in the adult to the true molar type. The opposite of this is seen in the full- grown fish, where a scattering of true molar teeth, intermediates, and typical gill-raker teeth are found intermingled, in intimate association in small patches on the hypo- branchials of the 1st gill-arch. These are typical gill-raker patches, distinct from the large basibranchial molars. They seem to exhibit a distribution in reverse, secondarily outward and away from a dominant molar concentration. Fish Number 28,051 b, a young transi- tional adolescent of 80 mm. has the entire edge of the premaxillaries toothed for a dis- tance from the snout-tip back of 6.7, fol- lowed almost unbrokenly by a 2.8 mm. toothed edge of the maxillaries. On the lat- ter there are only about twelve. The teeth on the premaxillaries and on the mandible are in two distinct rows, and distinguished by the decided divergence of the angles of direction, the first row almost straight, and those behind pointing obliquely back and into the mouth. The outer row is slightly larger and more even, but all are strongly curved and quite slender. In front of the upper jaw, on the whole ventral surface of the snout, the skin is covered with minute but hard spicules. The three palatal patches are fairly well defined in this young fish, rounded in out- line and convex, but each tooth, although rounded, possesses a sharp point, while many of those along the outer rim of the palatines are curved and more slender, half- way between the two extremes of teeth. About 1 mm. behind the symphysis of the upper jaw, lies the triangular vomer, its apex projecting forward, and the dentulous area confined to the posterior base. These vomerine teeth form a slightly irregular line, about 16 to 18 in all. They almost, but not quite, join on each side with the teeth of the palatines. Adult female, No. 26,131, length 352 mm., has the premaxillaries toothed throughout, in about five rows in front, dwindling to one at posterior end. All are fairly slender, 48 Zoologica : New York Zoological Society [XXVII: 8 slightly curved and sharp. All trace of teeth has gone from the edge of the maxillaries. The vomer has a straight line of teeth, slightly curved in front and containing about four rows. This area merges into the lateral palatine teeth, which form elongate areas, curved along outer outline, straight inside, four to six rows, narrowing behind to two. Teeth as in premaxillary but straighter. Twenty-two mm. behind the vomerine teeth, begins the large, oval, median patch on the parasphenoid. This is 30 by 12 mm. On each side, also oval, but more elongate, with the inner margins almost touching the median patch, are the entopterygoid patches, 32 by 10 mm. The posterior end of the median area extends to between the second gill-arches, while the lateral patches cut into only the first arch. All three patches are de- cidedly convex. These teeth are molar-like, mosaics of low, smooth, rounded mounds, largest on the top of the convexities. On the base of the tongue, covering the basibranchial, 18 mm. back of the fleshy tip, is a fourth molar patch, 9 by 16 mm., very steeply convex, sending back a narrow ridge over the center of the first arch, and ending on the third. This patch fits neatly into, and fills the space between, the three palatal patches, forming a most efficient grinding apparatus. Digestive System. The oesophagus extends straight back from the pharynx for 40 mm. enlarging abruptly into the dead white stomach, a cylinder about 50 mm. long by 30 mm. in diameter. Posteriorly, this mid-section of the stomach narrows rather sharply into a long (50 mm.) tapering, blind finger. From the level of the oesophageal entrance, a large rounded diverticulum extends 35 mm. forward, with a diameter of 25 mm., lying directly over and ventral to the oesophagus. On the left side this forward extension of the stomach is exposed, its tissues distin- guished by being darker than the stomach proper. Its top and entire right side, and most of the same side of the stomach as well, are covered solidly with the large, con- centrically curved, white caeca. There are 15 main caeca, but 5 of these are bifurcated for one-fifth to one-half of their length. Each caecum is distinctly seDarated from its fellows by heavy bands of dark pigment. The attaching tissue is very slight, and the whole caecal mass readily peels away to its basal line of intestinal attachment. From the summit of the anterior di- verticulum, the intestine arises, extending straight back to the anus, a length of 145 mm., with the caeca occupying the first 60 mm. The lobes of the liver are very unequal, the left 65 by 25 mm. and the right 30 by 15 mm. The former extends around and down over the middle third of the stomach, to beyond the mid-ventral line, while the smaller lobe overlies the anterior caeca. The slight amount of fat, chiefly a linear mass along the ventral line of the caecum, is bright orange. In a 90 mm. Dixonina (28,051 b) the stomach is relatively much more slender than in the adult, and the posterior blind end is not finger-like, but an undifferenti- ated posterior extent of the stomach, very slightly less in diameter. The caeca are 13 in number and relatively larger, both indi- vidually and in general extent. The anterior part of the stomach, the hardly distinguish- able mid part and the entire posterior por- tion are all crammed with small mysids, shrimps and euphausids. Adipose Eye-lid. The adipose eye-lid was so loosely at- tached that a considerable number of sand grains had worked beneath it. With a little careful manipulation I got the entire mass off whole. It was attached most firmly an- teriorly at the tip of the snout, below the nostril, and above the eye. In fact the an- terior portion was attached while almost the whole posterior, much of the ventral and the postero-superior areas were loose. EXPLANATION OF THE PLATES. Plate I. Fig. 1. Dixonina pacifica, sp. nov. Holotype, adult female, No. 26,131, Port Culebra, Costa Rica, January 24, 1938. Stand- ard length 352 mm. Fig. 2. Outer right gill-arch of Dixonina pa- cifica, transitional adolescent. Stand- ard length 80 mm. Fig. 3. Outer right gill-arch of Dixonina pa- cifica, transitional adolescent. Stand- ard length 185 mm. Plate II. Fig. 4. Outer right gill arch of Dixonina pa- cifica, adult female, holotype. Stand- ard length 352 mm. Fig. 5. Bend of outer right gill-arch of Dix- onina pacifica, adult female, holotype. Standard length 352 mm. BEEBE. PLATE I. Fig. 1. Fig. 2. Fig. 3. ATLANTIC AND PACIFIC FISHES OF THE GENUS DIXONINA. BEEBE. PLATE II. Fig. 4. Fig. 5. ATLANTIC AND PACIFIC FISHES OF THE GENUS D1XONINA. . NEW YORK ZOOLOGICAL SOCIETY General Office: 630 Fifth Avenue, New York City OFFICERS Fairfield Osborn, President Alfred Ely, First Vice-president Laurance S. Rockefeller, Chairman, Executive Committee & Second Vice-president Cornelius R. Agnew, Treasurer SCIENTIFIC STAFF General William Bridges, Editor and Curator of Publications Zoological Park Lee S. Crandall, Curator of Birds Raymond L. Ditmars, Honorary Curator of Reptiles and Insects Leonard J. Goss, Veterinarian Claude W. Leister, Curator of Mammals Charles M. Breder, Jr., Director Christopher W. Coates, Aquarist Myron Gordon, Research Associate in Genetics Ross F. Nigrelli, Pathologist G. M. Smith, Research Associate in Pathology Homer W. Smith, Research Associate in Physiology Department of Tropical Research William Beebe, Director Jocelyn Crane, Research Zoologist Henry Fleming, Entomologist Jean Delacour, Technical Adviser John Tee-Van, Executive Secretary Aquarium William K. Gregory, Associate Gloria Hollister, Associate John Tee-Van, Associate Mary VanderPyl, Associate Editorial Committee Fairfield Osborn, Chairman William Beebe Charles M. Breder, Jr. William Bridges Jean Delacour Harry Sweeny, Jr. John Tee-Yan ZOOLOGIGA SCIENTIFIC CONTRIBUTIONS of the NEW YORK ZOOLOGICAL SOCIETY VOLUME XXVII Part 2 Numbers 9-16 Published by the Society The Zoological Park, New York CONTENTS PAGE 9. Notes on Tschudi’s Types of Peruvian Birds. By Herbert Friedmann & H. G. Deignan 49 10. A Revision of the Kingfishers, Ceyx erithacus and rufidorsus. By S. Dillon Ripley 55 11. On the Reproduction of Gobiosoma robustum Ginsburg. By C. M. Breder, Jr. (Plates I & II; Text-figure 1) 61 12. Trichodina spheroidesi and Trichodina halli spp. nov. Parasitic on the Gills and Skin of Marine Fishes, with Special Reference to the Life-history of T. spheroidesi. By Morton Padnos & Ross F. Nigrelli. (Plates I-III; Text- figures 1-4) 65 13. Mortality of Albino Embryos and Aberrant Mendelian Ratios in Certain Broods of Xiphophorus hellerii. By Myron Gordon. (Text-figure 1) 73 14. The Schooling Behavior of Mackerel: A Preliminary Experi- mental Analysis. By Arthur Shlaifer. (Plate I) 75 15. Food, Eggs and Young of the Carnivorous Snail, Euglandina rosea (Ferussac). By William Marcus Ingram & Walter Edward Heming. (Plate I) v. . . 81 16. The Anatomy and Morphology of the Hypophysis of Several Species of Ovo-viviparous Poeciliids. By Hugh E. Potts. (Plates I & II) 85 1942] Friedmann & Deignan: Notes on Tschudi’s Types of Peruvian Birds 49 9. Notes on Tschudi’s Types of Peruvian Birds. Herbert Friedmann & H. G. Deignan* A little over a century ago, J. J. von Tschudi, the earliest important ornithologi- cal explorer of Peru, made a sizable collec- tion of birds in that country. Among these specimens he found and described a large number of new forms, most of which are considered valid today. His collection went to the Museum of Neuchatel, Switzerland, from which institution, in 1866, the United States National Museum received 27 speci- mens of birds, mostly mounted — the con- signment being marked “Types of Tschudi’s Fauna Peruana.” We have recently redis- covered these birds scattered through the collections and have made use of the occa- sion to go over them carefully. We find that most of them must be re- garded as types, or at least as cotypes. Tschudi described his new forms in the Archiv fur Naturgeschichte, 1843 and 1844, and in his “Untersuchungen fiber die Fauna Peruana,” 1844-46. In no case did he desig- nate a definite specimen as the type or even intimate whether he had one or more speci- mens before him, or state an explicit, re- stricted, type locality. Inasmuch as his de- scriptions were all written after his collect- ing work was complete, it is reasonable to assume that all his birds were available to him at the time of his studies and that, therefore, in the absence of true holotypes, all of his paratypes (which would include all of his original series) must be consid- ered as cotypes, except where the descrip- tion, for one reason or another, does not fit a particular specimen. Berlepsch & Hellmayr ( Journ . fur Orn., 1905, pp. 6-20) reported on Tschudi’s types in the museum at Neuchatel, apparently un- aware of the fact that a considerable num- ber of specimens had been sent away many years before. In fact, they merely discussed the Tschudi types they found at Neuchatel and made no comment on the forms de- scribed by Tschudi but no longer to be found there. The birds now in the United States National Museum fill in a good num- ber of these omissions as well as revealing a number of cotypes of the forms still repre- * Published by permission of the Secretary of the Smith- sonian Institution. sented (in 1905) in Neuchatel. The speci- mens now in Washington, which we con- sider to be of type or cotype status, are listed below. Our specimens of forms which Berlepsch & Hellmayr found to be repre- sented (by “type” — only three!) at Neu- chatel in 1905 are considered cotypes ; those which they failed to find at Neuchatel are, in lieu of other known specimens, considered to be types. If other specimens should be found in other museums, some of the types herein listed would, of course, become cotypes. Penelope adspersa Tschudi = Ortalis guttata adspersa (Tschudi). Archiv fur Naturgeschichte, ix, (1), 1843, 386 (Peru, “frequenter in sylvis” ;= eastern Peru ) . Type-. U. S. Nat. Mus. 41932; Perou. Chapman (Bull. 117, U. S. Nat. Mus., 1921, 44) writes that two males from Rio Cosireni have the throat and breast darker with the margins of the feathers whiter and more clearly defined than in examples fi'om southeastern Colombia; and that, if the dif- ference should prove to be constant, they should be known as 0. g. adspersus. One of these Rio Cosireni birds is before us and agrees quite well with the unsexed type, ex- cept in having the upperparts more olive, less reddish-brown, and in being larger. The measurements of the two are as fol- lows: type, unsexed, wing 185, tail 206, culmen from base 21 mm; Rio Cosireni, $, wing 203, tail 211, culmen from base 24 mm. The type has the entire underparts slightly more brownish than the Rio Cosi- reni example, but its general brownishness may be partly due to foxing. Charadrius W inter feldi Tschudi = Aphriza virgata (Gmelin). Archiv fur Naturgeschichte, ix, 1843, 388 (“in Oceani Magni littoribus”) . Type : U. S. Nat. Mus. 41933; original label lost, but locality entered in catalogue in 1866 as Valparaiso. That Valparaiso, Chile, may be the actual type locality is not to be ruled out of con- sideration, as in his “Travels in Peru, Dur- 50 Zoologica : New York Zoological Society [XXVII: 9 ing the Years 1838-1842” (transl. by Thomasina Ross, 1849, p. 25), Tschudi de- scribes birds seen for sale and at large in the harbor of Valparaiso on his way north to Peru, and he may have obtained his speci- men of the surf bird there. The type is a bird in winter plumage and has the follow- ing dimensions: wing 162, tail 58, culmen Irom the base 26 mm. Columba frenata Tschudi = Oreopeleia bourcieri frenata (Tschudi). Archiv filr Naturgeschichte, ix, (1), 1843, 386 (“in sylvis Antium declivitatis orientalis,” Peru = eastern slopes of the Andes of Peru) . Type: U. S. Nat. Mus. 41931, unsexed; Perou; Wing 160.5, tail 94.5, culmen from base 21 mm. The type is somewhat foxed. It has been compared with a male from Rio San Miguel, Peru, from which it differs in being more rufescent above, the rump and tail lighter and brighter, and in having the breast paler, much less grayish, more of a pale cinnamon drab and the sides, flanks, and under tail coverts much more rufescent. In his description of O. b. subgrisea, Chapman ( Amer . Mus. Novit. no. 31, 1922, p. 2) re- fers to the underparts of his new form, “. . . as in O. frenata, the breast pale drab- gray instead of cinnamon-drab or drab; center of abdomen slightly paler than breast, pale smoke-gray rather than drab as in most specimens of bourcieri; lower tail- coverts grayer.” This is rather confusing, as the type of frenata is the brownest of all on the underparts. The Rio San Miguel bird, which was identified as frenata by Chapman [Bull. 117 U. S. Nat. Mus., 1921, p. 47) is intermediate between typical bour- cieri (from El Roble and Almaguer, Colom- bia, and Zaruma, Ecuador) and the type of frenata. We have seen no specimens of sub- grisea and can only call attention to the possibility of its range extending southward to northern Peru, making the Rio San Miguel specimen within its limits of varia- tion, or the possibility of northern Peru and southwestern Ecuador being inhabited by variable intergrades between typical bour- cieri and frenata, not constant enough to warrant nomenclatural standing. Columba meloda Tschudi = Zenaida asiatica meloda (Tschudi). Archiv fur Naturgeschichte, ix, (1), 1843, 385 (“frequenter in regionibus calidis praecipue declivitatis Antium occidentalis” = western slopes of the Andes in Peru) . Type: U. S. Nat. Mus. 41930, unsexed; Perou; wing 177, tail 125, culmen from base 27.5 mm. Conurus mitratus Tschudi = Aratinga mitrata mitrata (Tschudi). Archiv fur Naturgeschichte, x, (1), 1844, 304 (Peru = Chanchamayo Valley; Zimmer, Field Mus. Nat. Hist. Publ. Zool., ser. xvii, 1930, 263). Type: U. S. Nat. Mus. 41926, unsexed; Perou; wing 189, tail 151, culmen from cere 31.5 mm. Conurus rupicola Tschudi = Pyrrhura rupicola (Tschudi). Archiv fur Naturgeschichte, x, (1), 1844, 304 ( Peru ) . Cotype: U. S. Nat. Mus. 41925, dis- mounted specimen, unsexed; “Perou”; wing 133, tail 100+, culmen from cere 16 mm. Another cotype is in the British Museum (Cat. Birds Brit. Mus., xx, 1891, 225). In his account of his journey, Tschudi (Travels in Peru, transl. by Thomasina Ross, 1849, p. 176) states that he found this bird together with the next one, abound- ing in the coastal region south of Lima, which may be taken as a more restricted type locality. The pose of the present (dis- mounted) specimen agrees with that in the figure of Tschudi’s Untersuchungen Ueber die Fauna Peruana (PI. 26, fig. 1). Psittacus tumultuosus Tschudi = Pionus tumultuosus (Tschudi). Archiv filr Naturgeschichte, x, (1), 1844, 304 (Peru) . Type: U. S. Nat. Mus. 41927, dismounted bird, unsexed; Perou; wing 171, tail 79, cul- men from cere 26 mm. The type locality may be more definitely stated as the coastal region south of Lima, where Tschudi writes ( Travels in Peru, transl. by Thomasina Ross, 1849, p. 176) he found this parrot to abound “. . . in the val- leys along the coast,” and to “. . . commit great depredations in the maize fields.” Psittacus mercenarius Tschudi = Amazona mercenaria mercenaria ( Tschudi ) . Archiv fur Naturgeschichte, x, (1), 1844, 303 (Peru) . Type: U. S. Nat. Mus. 41928, unsexed; Perou; wing 198, tail 93, culmen from cere 30 mm. Caprimulgus ocellatus Tschudi = Nyctiphrynus ocellatus ocellatus (Tschudi) . Archiv filr Naturgeschichte, x, (1), 1844, 268 (Peru) . Type: U. S. Nat. Mus. 41912, unsexed; Perou; wing 131.5, tail 122, culmen from base 20.5 mm. Peters (Check-list Birds World, iv, 1940, 1942] Friedmann & Deignan: Notes on Tschudi’s Types of Peruvian Birds 51 196) considers brunnescens Griscom & Greenway to need confirmation before recognizing it. We have a single Brazilian bird which by locality may be either brun- nescens or ocellatus and find it to be duskier than the type of ocellatus as it should be according to Griscom & Greenway if it were brunnescens but otherwise it disagrees with their description, being brighter, more ru- fescent brown than ocellatus. These writers state that brunnescens has a general light chocolate brown coloration while ocellatus is bright rufous brown. The reverse is true of the two birds before us. The birds are of comparable antiquity, the Brazilian speci- men being from the U. S. Exploring Expedi- tion’s collection, so there should be no dif- ferential foxing. The validity of brun- nescens is therefore very doubtful, as indi- vidual variation seems to be greater than was supposed. Capita glaucogularis Tschudi = Capito versicolor glaucogularis (Tschudi) . Archiv fur Naturgeschichte, x, (1), 1844, 301 (Peru) . Cotype : U. S. Nat. Mus. 41920, unsexed (female by plumage); Perou; wing 71.5, tail 81, culmen from base 22 mm. While there seems to be no reason to con- sider this specimen as less likely a type than the others, it should be noted that in some details it agrees only approximately with the colored plate (plate 24, fig. 2 in Tschudi’s Untersuchungen Ueber die Fauna Peruana). Thus, in the plate the crimson pectoral band extends entirely across the posterior end of the blue throat whereas in the specimen before us it is not quite so extensive; the green of the upperparts is lighter, slightly more yellowish in the bird than in the plate and the centers of the upper wing coverts not as dark as in the figure. The specimen, which is a “taken down” mounted specimen may, of course, have faded somewhat. It is also possible that the plate was made from another specimen, which is our reason for calling our bird a cotype. A female of versicolor from Idma, Peru, is very similar to the type of glaucogularis and differs only in having the crown and occiput less yellowish and the yellow margin of the upper and posterior edges of the auricular area more distinct. Chamaeza olivacea Tschudi = Chamaeza brevicauda olivacea Tschudi. Archiv fur Naturgeschichte, x, (1), 1844, 279 (Peru; restricted by Hellmayr, Cat. Birds Amer., iii, 1924, 293, to Montana de Vitoc, Dept. Junin, Peru). Type-. U. S. Nat. Mus. 41916, unsexed, dismounted bird; “Perou”; wing 96.5, tail 65.5, culmen from base 22.5 mm. Anabales montanus Tschudi = Auabacerthia striaticollis montanus (Tschudi) . Archiv fur Naturgeschichte, x, (1), 1844, 295 (Peru wooded region of Peru, between 10° and 12° South Lat.). Cotype: U. S. Nat. Mus. 41923, unsexed, dismounted bird; “Perou”; wing 90, tail 75, culmen from base 17 mm. Other cotypes exist in Neuchatel, and in the British Museum (Sclater, Proc. Zool. Soc. Lond.,' 1871, p. 86). In identifying the specimen we follow the nomenclature of Bangs (Bull. Mus. Comp. Zool., lxx, 1930, p. 252) rather than of Hellmayr (Cat. Birds Amer., iv, 1925, 195). Anabates ochrolaemus Tschudi = Automolus ochrolaemus ochrolaemus (Tschudi). Archiv fur Naturgeschichte, x, (1), 1844, 295 ( Peru = forest region of Peru, between 10° and 12° South Lat.). Cotype: U. S. Nat. Mus. 41915, unsexed, dismounted bird; “Perou”; wing 83, tail 74, culmen from base 22 mm. Other cotypes are in the British Museum and the Museum at Neuchatel. Dendrocolaptes chunchotambo Tschudi — Xiphorhynchus chunchotambo chunchotambo (Tschudi ) . Archiv fiir Naturgeschichte, x, (1), 1844, 295 (Peru = Chanchamayo Valley, Peru) . Cotype: U. S. Nat. Mus. 41918, unsexed, dismounted bird; “Perou”; wing 103, tail 90, culmen from base 35 mm. Other cotypes exist in the Neuchatel and British Museums. The present specimen has been discussed by Zimmer (Amer. Mus. Novit. no. 756, 1934, p. 17). Elaenia viridiflava Tschudi = Tyranniscus viridiflavus (Tschudi). Archiv fiir Naturgeschichte, x, (1), 1844, 274 (Peru = “coast region of Peru,” errore = tropical zone of central Peru). Cotype: U. S. Nat. Mus. 41921, unsexed, dismounted bird; “Perou”; wing 58, tail 48.5, culmen from base 9.5 mm. There is another cotype in the museum at Neuchatel (referred to as the type by Hellmavr, Cat. Birds Amer., v, 1927, p. 474). This specimen is paler above, the feathers without the darker centers, when compared with Tschudi’s colored figure (Untersuch- ungen Ueber die Fauna Peruana, pi. ix, fig. 2). 52 Zoologica : New York Zoological Society [XXVII: 9 Leptopogon super ciliaris Tschudi = Leptopogon superciliaris superciliaris Tschudi. Archiv fur Naturgeschichte, x, (1), 1844, 275 (Peru = fringes of the forests of central Peru, — Montana of Vitoc, Dept. Junin, Hell- mayr, Cat. Birds Amer., v, 1927, p. 485). Type: U. S. Nat. Mus. 41917, unsexed, dismounted bird; “Perou”; wing 75, tail 63.5, culmen from base 14.5 mm. Pipra cliloromeros Tschudi = Pipra cliloromeros Tschudi. Archiv fur Naturgeschichte, x, (1), 1844, 271 (Peru = montanas of “northwestern” Peru; = valley of Vitoc, Dept. Junin; Hellmayr, Cat. Birds Amer., v, 1927, 25). Type : U. S. Nat. Mus. 41924, unsexed (adult male by plumage), dismounted bird; Perou; wing 59, tail 24, culmen from base 10.5 mm. Cyphorhinus thoracicus Tschudi = Leucolepis thoracica thoracica (Tschudi) . Archiv fur Naturgeschichte, x, (1), 1844, 282 ( Peru = montanas of Uchubamba, near Vitoc, Dept. Junin). Cotype: U. S. Nat. Mus. 41922, unsexed, dismounted bird; Perou; wing 68, tail 48, upper mandible broken. This specimen is listed as a cotype rather than as a type because Hellmayr (Cat. Birds Amer., vii, 1934, p. 287) writes “(type in Neuchatel Museum).” Yet, Berlepsch & Hellmayr ( Journ . fur Orn., 1905, 6-20), in their paper on the Tschudi types at Neu- chatel, do not list this species. Tschudi’s colored figure (Untersuchungen Ueber die Fauna Peruana, 1846, pi. 16, fig. 1) is very poor. The type has the face, chin, throat, and breast much more rufescent and much brighter than the published figure. Ptilogonys leucotis Tschudi = Entomodestes leucotis (Tschudi). Archiv fur Naturgeschichte, x, (1), 1844, 270 ( Peru ) . Cotype: U. S. Nat. Mus. 41908, unsexed, dismounted bird; Perou; wing 106.5, tail 106, culmen from base 21 mm. Baird (Rev. Amer. Bii’ds, pt. 1, 1866, p. 432-433) has described this specimen in detail. He writes that the bird is one “. . . of Mr. Tschudi’s types, presented to the [Smithsonian] Institution by the Museum of Neuchatel. It is moulting a considerable portion of its feathers, which somewhat ob- scures its characters, and it may even be a young bird not yet arrived at maturity.” It does not seem probable that the specimen is immature. It apparently is an adult in moult, especially on the chin and sides of the head. Berlepsch & Hellmayr failed to find any “type” of this bird in Neuchatel in 1905, but Hellmayr (Cat. Birds Amer., vii, 1934, 445) writes “type in Neuchatel Museum.” Tanagra analis Tschudi = Iridosornis analis analis (Tschudi). Archiv fur Naturgeschichte, x, (1), 1844, 287 (Peru = Valley of Vitoc, Dept. Junin; Hell- mayr, Cat. Birds Amer., ix, 1936, 179, foot- note) . Cotype: U. S. Nat. Mus. 41919, unsexed, dismounted bird; Perou; wing 82, tail 69, culmen from base 13 mm. Another cotype exists in the British Mu- seum (ex Sclater coll.) according to Sclater (Cat. Birds Brit. Mus., xi, 1886, p. 142). Hellmayr (Cat. Birds Amer., ix, 1936, p. 179) states “type in Neuchatel Museum,” but in his earlier paper with Berlepsch, made no mention of this species among the Tschudi types still in that museum. Zimmer ( Field Mus. Nat. Hist., Zool. ser., xvii, 1920, 443) writes that the “type may have come from near Lima since Tschudi says . . . the species was common in the fruit gardens of Lima, he did not find it further north or east . . .” However, Hellmayr (cit. supra) writes (from subsequent knowledge of the range of the bird) that Tschudi’s statement is a mistake, and suggests the restricted type locality given above. Tanagra frugilegus Tschudi = Thraupis bonariensis darwinii (Bonaparte) . Archiv fiLr Naturgeschichte, x, (1), 1844, 286 (Peru = prob. fruit gardens of Lima). Cotype: U. S. Nat. Mus. 41913, unsexed, (male by plumage) ; Peru (original lavel lost) ; wing 85, tail 73, culmen from base 14 mm. Another cotvpe went to the British Mu- seum (Cat. Birds Brit. Mus., xi, 1886, p. 165). Cissopis minor Tschudi = Cissopis leveriana leveriana (Gmelin). Untersuchungen Fauna Peruana, Aves, 1846, p. 211 (wooded region of Peru ; = Chacay- bamba ? ) . Cotype: U. S. Nat. Mus. 41914, unsexed, dismounted bird; Chacaybamba, Peru; 4.8.39 (August 4 or April 8, 1839) ; wing 112, tail 135, culmen from base 20 mm. The data as to place and date are on the original field label in what is probably Tschudi’s own caligraphy. The other speci- mens listed in this paper have only the original museum exhibition labels, which this bird has as well, and which merely read — “Perou. Voyage de M. Tschudi.” Hellmayr (Cat. Birds Amer., ix, 1936, 439) states that the type (= another co- 1942] Friedmann & Deignan: Notes on Tschudi’ s Types of Peruvian Birds 53 type) is in Neuchatel Museum, but in 1905 he and Berlepsch failed to find it there. Arremon frontalis Tschudi = Atlapetes brunnei-nucha brunnei-nueha (Lafresnaye) . Archiv fur Naturgeschichte, x, (1), 1844, 289 (Peru = eastern wooded region between 8° and 9° lat. south and Jaen de Bracamoras, Peru) . Cotype : U. S. Nat. Mus. 41911, unsexed bird; original label lost; wing 79, tail 85, culmen from base 21 mm. Although no type of this bird was found at Neuchatel in 1905 by Berlepsch & Hellmayr, the latter author (Cat. Birds Amer., xi, 1938, p. 413) states that a type is there. We therefore consider the present specimen a cotype. Included with these types in the consign- ment from the Neuchatel Museum were four other Tschudi specimens which are not of type status. For the benefit of students wishing to unravel the synonymies of the species involved, they are listed below, to- gether with references to the names under which Tschudi listed them. Ortygonax rytirhynchus rytirhynchus (Vieil- lot) . Rallus caesius (not of Spix) Tschudi, Unters. Fauna Peruana, 1844-46, p. 301. One spec., U. S. Nat. Mus. 41934, (Perou). Col u tuba plumbed delicata Berlepsch & Stolz- mann. Columba infuscata, Tschudi, Unters. Fauna Peruana, 1844-46, p. 275. One spec., U. S. Nat. Mus. 41929, (Perou). Thamnophilus melanochrous Sclater & Sal- vin. Thamnophilus luctuosus (not of Lichten- stein) Tschudi, Archiv. fur Naturgeschichte, x, (1), 1844, p. 278. One spec., U. S. Nat. Mus. 41910, (Perou). Molothrus bonariensis oecidentalis Berlepsch & Stolzmann. Icterus brevirostris (not of d’Orbigny and Lafresnaye) Tschudi, Archiv. fur Naturge- schichte, x, (1), 1844, 292. One spec., U. S. Nat. Mus. 41909, (Perou). V 1942] Ripley: A Revision of the Kingfishers 55 10. A Revision of tlie Kingfishers, Ceyx erithacus and rufidorsus. S. Dillon Ripley These two species of kingfisher are im- portant members of the Indo-Malayan avifauna. For many years their identifica- tion has proved a stumbling block and a very large amount of literature has collected around them. The most recent discussion of the problem, that of Chasen & Kloss (1929), has served to clear up many difficulties. I believe, however, that one or two interesting facts remain to be pointed out, and it is in an effort to do so that I have turned to these species in this paper. Kingfishers in general tend to show a con- stancy of plumage pattern which is strik- ing. Certain characteristic colors reappear again and again. Some colors, as brown and yellow, seem to be closely linked and substi- tute for each other with great regularity. Plumage patterns are nearly uniform throughout the family. In the case of these two species of Ceyx, the color and size re- semblances are so particularly close that the conclusion that these birds are very nearly related seems inescapable. The next nearest relation of these two species is Ceyx melanurus with three races from the Philippines. I am much indebted to Mr. J. L. Peters for reading over this manuscript as well as to the authorities of the United States Na- tional Museum, the Academy of Natural Sciences of Philadelphia and the American Museum of Natural History for the loan of specimens. In the following discussion all measurements are in millimetres, the wing pressed flat against the ruler, and the bill measured from the distal end of the external naris to the terminal point of the maxilla. For easy identification of these two species, it might be well to insert here a key by which identification can be made. A. Upper parts rufous with a lilac wash. a. Forehead with a dark blue-black spot . ... C. erithacus. b. Forehead spot absent . ... C. rufidorsus. a'. A patch of ultramarine in the supraocular region . . . C. erith- acus. b'. Supraocular patch lacking .... C. rufidorsus. a". Scapulars black with an ultra- marine wash . ... C. erithacus. b". Scapulars rufous with a lilac wash . ... C. rufidorsus. a' ". Wing coverts black tipped with ultramarine . . . .C. erithacus. b' ". Wing coverts rufous tipped with lilac . . . . C. rufidorsus. Ceyx erithacus erithacus (Linnaeus). Alcedo erithaca Linnaeus, Sys. Nat. 10, T. p. 115, 1758. Description'. For the adult bird see Sharpe (1892, p. 175). The rufous on the bend of the wing often extends to the lesser wing coverts. Juvenal birds are distinguished from adults by several characteristic features. The bill is rather short and pale, often dusky at the base. On the upper surface the plumage is the same as the adult, but the scapulars tend to be tipped rather than washed with ultramarine. Instead of being washed with lilac, the feathers of the lower back sometimes are tinted with cobalt. The tail is often, but not invariably, tipped with black. Below, juvenal and immature birds show a very characteristic plumage. The throat is pure white not tinted with yellow. The lower cheeks, flanks, thighs, under tail coverts, and a band across the breast and upper abdomen are rufous, sometimes with an orange tint. Measurements'. Wing, $ 53-58.5 (55.8), 2 55-60 (56.8). Tail, $ 20-23.5 (22.6), 2 2.15-24.5 (23.1) . Wing-tail index 38-42%. Bill, from naris, $ and 2,25-28.2 (27.2). Range : Ceylon through the lowland coastal parts of India to Nepal, Assam, Burma, fide Stuart Baker (1927), the Ma- lay Peninsula, Siam, Indo-China, southern China, Hainan, Andaman Ids, small islands in the Straits of Malacca and off the Malay Peninsula, coast of Sumatra. Specimens Examined: Sixteen. Discussion: This form is well established in continental Asia but only sparingly dis- tributed among the islands. Such a distri- bution indicates a later origin than that of rufidorsus which has extended into the 56 Zoologica : New York Zoological Society [XXVII: 10 Greater and Lesser Sunda ai'ea. The latter having become established in the Greater Sundas and adjacent islands, there has been a secondary infiltration perhaps in two waves by erithacus, which has succeeded in colonizing a few of the small islands not already reached by rufidorsus (Nicobars, Nias) and in competing with the latter species in the larger continental areas (Ma- lay Peninsula, Borneo). From the evidence afforded by specimens, Sumatra seems to be primarily the home of rufidorsus. I have examined the type of Ceyx enopopygius (Oberholser, 1912) and concur with Chasen & Kloss (1929) in their belief that it is an unusually bright imma- ture specimen. The measurements of this specimen (wing 57.5, tail 23.5, bill 27.7) are not different from normal erithacus. The only other examples of erithacus from Su- matra are two males listed in Laubmann (1925). These three records are from coastal localities and this coupled with Robinson & Kloss’ (1922) record of this species as being a bird commonly killed at one of the lighthouses in Malacca Strait leads me to suspect that the Sumatra rec- ords are accidental. These birds in contrast to Ceyx rufidorsus (Robinson, 1917) are subject to erratic local movements which re- sult in isolated records appearing all over the islands of Malacca Strait and the ad- jacent Sumatran coast. This is a subject which should be studied more carefully. It is possible that some physiological dispersal mechanism is at work here which may be akin to migration. Ceyx erithacus macrocarus Oberholser. Ceyx erithacus macrocarus Oberholser, Bull. 98, U. S. Nat. Mus., 23, 1917. Description'. Differs from erithacus by being larger. The forehead spot is smaller in two examples. A nestling, A.M.N.H. 637008, is colored as the adult except for the reduction of the ultramarine wash on the scapulars back and median wing coverts to a series of spots. The shaft and the terminal part of the tail are black. Below the bird is similar to young of erithacus. This specimen was collected Oct. 4, 1905. Measurements'. Wing, $ 58-60.3 (59.4), 2 61. Tail, $ 23.-24 (23.4), 2 24.5. Wing- tail index 39-40. Bill, $ and 2 30-31 (30.4). Range: Great Nicobar, Nicobar Ids. Specimens Examined: Five. Discussion: The most significant charac- ter of this race is its larger size, especially in bill measurements. It is interesting that the Andamans seem to be populated by typical erithacus. This is in contrast to the usual case in which Andaman and Nicobar populations tend to be identical. A single male from South Andaman (A.M.N.H. 637011) is characteristically of the smaller subspecies. This particular specimen is notable for a very heavy blue forehead spot which extends back broadly onto the crown. In order to check on the above measure- ments, I have estimated the probability of these two populations being identical by use of the formula of T for deviations from the mean in small samples as discussed by Simpson & Roe (1939). This formula allows a very critical estimation when, as in this case, the combined samples of adult speci- mens are less than fifteen. By using the formula on the wing, tail, and bill measure- ments, T proves to be very significant for the bill (6.0), significant for the wing (3.1) and insignificant for the tail (1.5). Thus even on the basis of the bill alone, macro- carus is a justifiable race. Ceyx erithacus motleyi Chasen & Kloss. Ceyx erithacus motleyi, Chasen & Kloss, Festsch. fur E. Hartert, Journ. f. Ornith., p. 106, 1929. Description: This and the following two races differ from erithacus and macrocarus by having the rufous tinted with lilac of the pileum and nape extending down on the back and continuous with the lower back and rump. The rest of the plumage, how- ever, is similar to the typical subspecies. The immature plumage differs from the adult as in erithacus. Measurements: Wing, $ 57.5-60.5 (59.5). Tail, $ 22-24 (22.8). Wing-tail index 36-39. Bill, $ 28.5-30 (29.1). Range: Borneo, Banguey Is, Labuan (?). Specimens Examined: Five. Discussion: The confusion resulting over the description of Ceyx dillwynni and sharpei from Borneo was finally resolved by Chasen & Kloss (1929). I believe, however, that they were mistaken in assuming Sharpe’s plate of dillwynni (1868-71) to be a representation of an immature specimen of rufidorsus. I have before me two male adult specimens (M.C.Z. 197135, 197136) which correspond very closely to Sharpe’s plate. I believe that they are hybrids between ru- fidorsus and motleyi. From a careful exami- nation of the specimens it is difficult to escape the conclusion that where erithacus overlaps into the range of rufidorsus, hybridization has occurred. Indeed Sharpe had apparently already begun to suspect this by the time that he was working on the Catalogue of the Birds in the British Mu- seum. In that volume (1892) he transfers his original description of dillwynni to a description of what is now motleyi and lists several dubious specimens which he re- marks may be “in changing plumage . . . or are hybrids.” A description of some of these specimens may be appropriate at this point. 1942] Ripley : A Revision of the King fisher s 57 (1) . M.C.Z. 197135, $, ad. Kalabakang R., Borneo, coll. July 7, 1937. H. G. Deignan. Above this specimen is similar to typical motleyi, but the forehead spot is very much reduced as is also the supraocular patch of ultramarine. TheScapulars are washed with ultramarine, but tipped with pale lilac. The lesser and median wing coverts are strongly rufous (more so than in Sharpe’s plate). The distal outer margins of the secondaries as well as the inner margins are edged with rufous. Below the bird is similar to the adult of either rufidorsus or erithacus. (2) . M.C.Z. 197136, <5, ad. Kalabakang R., Borneo, July 14, 1937. H. G. Deignan. Above this specimen lacks any hint of fore- head spot and again the supraocular patch of ultramarine is nearly gone. The scapulars are as in motleyi, but some of the upper median wing coverts are rufous and the greater wing coverts and outer margins of the secondaries are irregularly tipped with rufous. (3) . M.C.Z. 197134, $, ad. Sandaken, Borneo, June 18, 1937. H. G. Deignan. Above this specimen is uniformly rufous with a lilac wash. There is no trace of either fore- head or supraocular spot. The scapulars pre- sent a curious violet tone from the mingling of ultramarine and lilac washes. The lesser median wing coverts are rufous. The greater wing coverts are black with a bluish-lilac subterminal wash and rufous edgings. This bird stands almost exactly in an intermedi- ate position between motleyi and rufidorsus. (4) . M.C.Z. 69608, ad. Linibang, Sa- rawak, Borneo, no date. Above this bird has a forehead spot and supraocular spot. The scapulars are, however, predominately ru- fous-lilac. The lesser and median wing cov- erts are rufous, the greater strongly tipped with ultramarine. (5) . M.C.Z. 69610, ad. Borneo, 1915. H. W. Smith. Above this bird is similar to rufidorsus. However, it is an adult specimen and so should not have curious patchy black- ish and ultramarine tinted scapulars. The wing coverts are similar to rufidorsus ex- cept that there is a hint of ultramarine on the tips of some of the greater one. (6). M.C.Z. 69609, o im., Kuala Treban, Sarawak, Borneo, Feb. 28, 1918. An imma- ture hybrid is of interest. Above this speci- men shows a prominent forehead spot and the merest trace (two feather tips) of a supraocular patch. The scapulars are black with a few faint tips of ultramarine. The lesser and median wing coverts are largely rufous, the greater are black, washed with ultramarine and with faint rufous edgings. Below this bird is typically immature, hav- ing a white throat and strongly rufous cheeks, flanks, and breast band. The above specimens represent almost a complete transition from erithacus motleyi to rufidorsus rufidorsus. Each bird repre- sents a greater or less blending of the dis- tinctive characters of the two species. Here indeed is an interesting example of two closely related forms of kingfisher which do not quite satisfy any of the concepts of taxonomy or speciation. On the one hand the color of the back is so different that many taxonomists would consider these birds to belong to two species. Others might lump them, if it were not for the over-lapping of their ranges. From the speciation point of view they do not conform either to a Super- species or to a rassenkreis. And yet the birds would satisfy any geneticist as to their close ancestry by their apparent readi- ness to hybridize. Shown on a linear scale, the characters of these hybrids indicate quite clearly their intermediate position. Letting certain car- acters equal certain numbers we arrive at an arbitrary scale as follows : erithacus motleyi Forehead spot present — 0 Supraocular spot present — 0 Scapulars blue — 0 Lesser & median wingcoverts black — 0 Greater wing- coverts black — 0 rufidorsus rufidorsus absent — 10 absent — 10 rufous lilac — 10 rufous — 10 rufous — 10 motleyi rufidorsus Text-fig. 1. The numbers on the left are those of the individual specimens. The base line represents the scale from motleyi to rufidorsus. At the end of each specimen’s line is the figure representing the sum of its characters. 58 Zoologica: New York Zoological Society [XXVII: 10 By adding up each specimen’s score and dividing by five we arrive at the following: The original description of Ceyx dill- wynni Sharpe (1868) would have sufficed for the local race of eritliacus, if it had not been for the fact that not only did Sharpe not mention a supraocular patch as occur- ring on his type, but Salvadori (1869) spe- cifically says that it did not have one. I cannot find a single undoubted specimen of the species eritliacus which does not possess this spot. Any specimen, therefore, coming from Borneo or Sumatra or the Malay Peninsula (as will be seen later) which is an adult but does not possess either the forehead spot or supraocular patch, and yet which has ultramarine on scapulars or wing coverts must be considered to be potentially a hybrid. Ceyx erithaeus captus Ripley. Ceyx eritliacus captus Ripley, Proc. New Eng. Zool. Club, XIX, 15, Dec. 29, 1941. Description : This race differs from mot- leyi by its longer bill and slightly larger size, and by reduction of the forehead spot which is lacking in one specimen. This last character was not mentioned in the original description due to the fact that the question of the hybrid population of motleyi has not been elucidated. From eritliacus this race differs as motleyi. I have seen no immature specimens. Measurements: Wing, $ 59.5-62.5 (60.8). Tail, $ 23-24 (25.6). Wing-tail index 39-40. Bill 32-33.5 (32.6). Range: Nias I. West Sumatra. Specimens Examined: Thi-ee. Discussion: It is interesting to note that captus, as macrocarus, differs from its near- est relative by size. In this case it is the bill which is notably larger. Like macrocarus, also, the forehead spot and the supraocular patch are much reduced. Ceyx eritliacus vargasi Manuel. Ceyx eritliacus vargasi Manuel, Phil. Journ. Science, 69, No. 4, 383, Aug., 1939. Description: Differs from motleyi, which it otherwise closely resembles, by the reduc- tion of the lilac wash on the upper surface. In the specimen examined this wash occurs only in a supraorbital stripe ending in an ultramarine supraocular patch, and on the lower back, two areas where the color is most highly concentrated in motleyi. The ultramarine wash on the scapulars also is reduced to the tips of one or two feathers. The specimen examined, though otherwise in adult plumage, has a black shaft and a black stripe along the middle of the terminal half of the tail feathers. Measurements (one female) : Wing 56.5. Tail 24. Wing-tail index 42. Bill 28.5. Range: Mindoro I., Philippine Islands. Discussion: The existence of this popula- tion of Ceyx eritliacus was only discovered in 1939 by Manuel who noticed that two specimens of Ceyx had the blue supraocular patch not found in rufidorsus. It is worth noting that Ceyx melanurus behaves in a strictly representative way in the Philip- pines with rufidorsus, while eritliacus has incurred on the range of rufidorsus on Mindoro. Ceyx rufidorsus rufidorsus Strickland. Ceyx rufidorsa Strickland, Proc. Zool. Soc., p. 99, 1846. Ceyx innominatus Salvadori, Atti R. Accad. Sci. Toririo, IV, p. 465, 1869. Description: Above, rufous washed with lilac, scapulars and wing coverts as the back, primaries black, the first edged with rufous, secondaries black edged with rufous. Ordinarily this species lacks a dark fore- head spot (three times present in 25 ex- amples) and an ultramarine supraocular patch (twice present in 25 examples) . Below the throat is white, the rest of the under- parts being rich yellow. Immature birds differ from the adult by having less of the bright lilac wash on the rufous upper parts. A nestling (A.M.N.H. 637014) from Gunong Tahan, Pahang, Ma- lay Penin., collected in November, has black scapulars tinted with rufous and with one or two faint spots of ultramarine. The wing coverts tend to have rather more black than the adult. The tail is entirely rufous. Below the throat and belly are white, the cheeks, flanks, and a band across the abdomen rufous. An immature male from East Java (A.M.N.H. 637034) collected in August, has completely rufous scapulars. This bird is one of the specimens which has an ultra- marine spot above the ear. Below it agrees with the other specimens. An immature female from Borneo (A.M. N.H. 637055) collected in September, has blackish scapulars overlaid with rufous and a black-tipped tail. Measurements (adults) : Wing, $ 56.5- 60.5 (58.2), 2 59.5-60.5 (60). Tail, $ 22.7- 25 (23.7, 2 24-24.7 (24.2) . Wing-tail index 39-41. Bill, from naris, $ and 2, 27-32 (28). Range: Malay Peninsula, Rhio and Lingga Islands, Banka, Billiton, Sumatra, Siberut, Sipora, Java, Bali, Lombok, Sum- bawa, Flores, Kangean, Bawean, Borneo, North Natuna, Anamba Ids, and Philip- pines, Palawan, Balabac, Mindoro, Tawi- tawi, Bongao, Calamianes. Specimens Examined : Twenty-one. Discussion: Several specimens from the Malay Peninsula and Sumatra have been noted which appear to be hybrids. (1). An adult female (A.M.N.H. 637027) from the Deli district of Sumatra, Van Heyst coll., has a prominent forehead spot 1942] Ripley: A Revisio?i of the Kingfishers 59 and an infusion of ultramarine in the supra- ocular area. The scapulars, lesser and me- dian wing coverts are mixed with black and ultramarine. (2) . Another adult female (M.C.Z. 17707) from Benkoolen, Sumatra, has a prominent forehead spot and blackish scapulars and wing coverts irregularly spotted with traces of ultramarine. (3) . A juvenal male (U.S.N.M. 180199) from Kateman River, E. Sumatra, collected in August, has the forehead spot and black tipped with ultramarine scapulars of eritha- cus, but it lacks the supraocular patch. The terminal half of the tail along the shaft of the feathers is black. (4) . An immature specimen from Great Karimon Id., E. Sumatra, (U.S.N.M. 180198) collected in May, has a black fore- head patch and mixed scapulars, black and rufous, with lilac and ultramarine spots. (5) . A male molting into adult plumage (A.M.N.H. 637012) from western Pahang, Malay Peninsula, collected in January, has the blue forehead spot and supraocular patch of eritliacus. However, the wing cov- erts, as in some of the Bornean hybrids, are mainly rufous. These specimens indicate clearly that wherever the range of these two species overlap, there is a pronounced tendency to hybridization. From them it is clear that any specimen from Sumatra, the Malay Peninsula or Borneo, which seems to be adult but has either of the following com- binations, must be suspect. Hybrid Type A: Forehead spot and supra- ocular patch present, but scapulars and wing coverts largely rufous. Hybrid Type B : Forehead spot and supra- ocular patch absent, but scapulars and wing coverts largely black with an ultramarine wash. Specimens from Java tend to be slightly smaller, but tests for the significance of these data by the formula of T, show that the differences are not valid and the name innominatus cannot be upheld. Ceyx rufidorsus jungei ssp. nov. Type: M.C.Z. no. 178157, $ ad., collected by E. Jacobson and W. C. vanHeurn, July 28, 1913, Ajer Dingin, Simalur I. Diagnosis: From C. r. rufidorsus this race differs by its larger size. Measurements (of type) : Wing 62, tail 26, bill 31.5; (of series) : Wing, <5 62-64.5 (63.2), 9 62.5-63.5 (63). Tail, 2 25-26 (25.5), 2 25.5-26.5 (26). Wing-tail index 39-42. Bill, $ and 2,30.5-32 (31.3). Range: Simalur and Batoe Ids, Tanah Massa and Tanah Bala. Specimens Examined: Seven. Discussion: This race is decidedly larger than rufidorsus from the rest of the range. It is interesting to note that the bill meas- urements, though averaging larger, are not significantly so when the probability is com- puted by standard deviation. Birds from Siberut and Sipora, as listed by Chasen & Kloss (1926), agree in size with typical rufidorsus. Here again, as in the two races of Ceyx erithacus on small islands north and west of Sumatra, this population of kingfishers differs from its nearest rela- tives by larger size. This race is named in honor of Dr. G. C. A. Junge of the Leiden Museum who has always been interested in East Indian birds. Conclusion. Ceyx erithacus and rufidorsus are closely allied species and might be considered con- specific were it not for the fact that their ranges are overlapping. In the Malay Penin- sula, Sumatra and Borneo where this oc- curs, specimens were examined which show hybrid characters indicating that the earlier confusion in the nomenclature was probably due to this phenomenon. Ceyx rufidorsus is considered to be the older species due to its more extensive range. Literature Cited. Baker, E. C. Stuart 1927. The Fauna of British India, ed 2, Birds, IV, 260. Chasen, F. N. & Kloss, C. Boden 1926. Spolia Mentawiensis — Birds, Ibis, II, 12 Ser. 281. 1929. Some New Birds From North Borneo. Fest. fur Ernest Hartert, Jour, fur Ornith., Band 2, 106. Laubmann, A. 1925. Die Eisvogel der Insel Sumatra. Archiv fur Naturgesch. 90th Year (1924, Abteilung A, 7 Heft, 96-108. Oberholser, H. C. 1912. Descriptions of One Hundred and Four New Species and Subspecies of Birds from the Barussan Islands and Su- matra. Smithsonian Misc. Coll., 60, 7, 7. Robinson, H. C. 1917. On a Collection of Birds from Pulau Langkawi and Other Islands on the North-West Coast of the Malay Pen- insula. Journ. Fed. Malay States Mus., VII, pt. Ill, 146. Salvadori, T. 1869. Monografia del Genere Ceyx Lacepede. Atti della R. Accad. delle Scienze di Torino, IV, 467. Robinson, H. C. & Kloss, C. Boden 1922. Birds from the One Fathom Bank Lighthouse, Straits of Malacca. Journ. Fed. Malay States Mus., X, pt. 4, 254. Sharpe, R. B. 1868. On the Genus Ceyx. Proc. Zool. Soc. London, 587-599. 1868-71. A Monograph of the Alcedinidae or Family of Kingfishers. London, pi. 43. 1892. Catalogue of the Birds in the British Museum. XVII, 177. Simpson, G. G. & Roe, A. 1939. Quantitive Zoology. McGraw-Hill Co., p. 210. 1942] Breder: Reproduction of Gobiosoma robustum 61 11. On the Reproduction of Gobiosoma robustum Ginsburg. C. M. Breder, Jr. New York Zoological Society. (Plates I & II; Text-figure 1). Introduction. The small and active goby, Gobiosoma robustum Ginsburg, inhabits a wide range of territory on the west coast of Florida near the field station of the New York Aquarium located on Palmetto Key. There appears to be nothing whatever in the literature on the habits or behavior of this common fish. The observations and data presented herewith were gathered inci- dental to studies of the life histories of other species in this region (see Breder 1939a and b, 1941a and b ) . Mr. M. B. Bishop of Yale University, Mr. L. A. Ivrumholz of the University of Illinois and Mr. B. Dontzin of Cornell University all rendered valued as- sistance in connection with this study. This species, differentiated from related forms by Ginsburg (1933), represents the only member of the genus found occuring near the laboratory. Here it is abundant and may be taken in numbers among the man- grove roots and on open sandy beaches. Dredging brought it up from depths as great as twenty feet, which represents the deepest of this shallow bay in the vicinity of the laboratory. It was taken in such equipment over all of the various types of bottom, such as grassy places, sponge beds, scallop beds, sand bars and even over soft spots of flocculent mud. The individuals of this form are exceed- ingly variable. Fowler (1941) figures six patterns represented in his material. Some of the pattern differences are referable to sex but they are so overlaid with individual vagaries that exact description becomes dif- ficult. Nevertheless with a little experience it is possible to sex individuals with reason- able accuracy. Shropshire (1932) figures the young of a Gobiosoma under the name Gobiosoma moles- tum. It is impossible to tell just which form, under present terminology, he had. G. moles- tum Girard is now a synonym of G. bosci (Lacepede) but as Shropshire thought his material was not the latter it is possible that he was actually dealing with G. robus- tum as here understood. There is no confusion about the identity of the present material as the fish were taken guarding their eggs and furthermore all other specimens in our collections are referable to this one species, and it is the only form definitely known to be living in the vicinity. Dr. I. Ginsburg, the describer of G. robustum, was good enough to check over this material and compare it with his large series of both species. Nesting and Nesting Sites. Gobiosoma robustum may be found with its eggs from March to June, at least. At first this finding led to the assumption in the field that there were more than one species included. However, since a critical study of the incubating fishes shows them to be referable to a single species, it follows that this form has a long spawning period which reaches from the coldest to the warm- est period of the year. It may well be that actually spawning is in progress at all times, which, if true would account for the relatively small number of nests encoun- tered at any one time, compared with the abundance of individuals in the region. The nests so far located have all been in shells or sponges. Usually the eggs are found hanging from the underside of some surface, but are not necessarily in such a position. Plate I, Figure 1, shows the two valves of a Pecten shell with the eggs at- tached to what had been the lower shell and the guarding male with them. This fish con- tinued caring for the eggs with the shell in its open position. Courtship and Parental Care. Apparently only the males guard the eggs, as females have never been found in 62 Zoologica : New York Zoological Society [XXVII: 11 the vicinity of nests. The males fan and work over the eggs in a manner reminiscent of a fresh water darter. Beyond this noth- ing was noted in the matter of specialized behavior. Although the males would attack small animals, such as other gobies, the slightest disturbance of a larger sort would usually cause them to retire. They would return as soon as the disturbance subsided. Although we were not successful in hav- ing this species spawn in aquaria, several males established themselves in sheltered places which they would defend against the intrusion of their tank mates. From such locations they would sally forth to court nearby females. This was done with much spreading of the fins, especially the dorsal, accompanied by short darts near or at the female. This was usually carried on until she retreated. The coloration of the male at these times became intense and dark, nearly black, in most cases; most notably so on the widely extended first dorsal fin. The largest male seen was 34 mm. in standard length. This is the fish shown in Plate I. The smallest mature and ripe male seen was 16.5 mm. Ripe females examined ranged from 16.5 to 21.5 mm. Apparently the males not only reach a larger size but average somewhat larger than the females as well. Smaller individuals grading down to those of a few millimeters are present both summer and winter, again indicating a long if not continuous reproductive season. The Eggs. The elongate eggs are attached by one end to a matted base of adhesive threads. They are evenly elliptical but there is an apparent seasonal difference in their pro- portions. Those taken in March average rather differently than those taken in June according to the following schedule, meas- ured on living eggs and given in mm. dropping out. In this connection the differ- ences in the spawning times of Opsanus tau and Opsanus beta discussed by Breder (1941a) is suggestive. The winter eggs are shown in six stages of development in Text-fig. 1. When first found they appeared as in “A,” March 14, 3:45 p.m. The yolk and germinal parts are opaque and slightly yellowish. The elliptical envelope is clear and without markings. Attachment is at one end by means of a tangled mass of adhesive strands. The na- ture of this material is better shown in the photographs of PI. II. In many of the eggs the yolk was nearer to the upper tip than to the center. Numerous exceedingly small droplets, hardly visible at lower powers, were scattered over the surface of the yolk as indicated in the sketches. Twenty-two and a quarter hours later the eggs appeared as shown in Text-fig. 1, “B.” At this stage the swelling of the large head was espe- cially prominent. In an additional twenty-six and a quarter hours the embryo was well formed and appeared as in Text-fig. 1, “C.” Both yolk and embryo presented an opaque appearance. After twenty and one-quarter hours more the embryo had reached the pro- portions shown in Text-fig. 1, “D.” At no time was there a sufficient transparence to evidence much of the internal structure. The side view seen in Text-fig. 1, “E,” when the egg was twenty-one hours more ad- vanced, shows some of the vertebral frag- mentations. By the time an additional twenty-seven and one-half hours had passed the heart was evident and beating slowly and the tail gave spasmodic twitches. These figures cover a period of 117% hours during which time the temperature in the incubating dishes ranged from 15.5 to 18.5° C. It was noted that some of the eggs were not as advanced as others, indicating at least more than one spawning. These fig- No. of eggs Months Average Length Maximum Minimum Average Width Maximum Minimum 8 March 1.62— 1.70 1.55 0.62— 0.70 0.60 2 June 1.35 1.40 1.30 0.50 0.50 0.50 10 All 1.57 1.70 1.30 0.60 0.70 0.50 These differences would seem to be refer- able to seasonal effects, probably chiefly temperature, controlling the speed of de- velopment of the ovarian eggs. Another pos- sible interpretation is that this is an ex- pression of incipient speciation, starting first with a prolonged spawning season fin- ally reaching over the extremes of summer and winter temperatures. Those fish most responsive to temperature variations, at either end of the long season, might be be- ginning to show slight differences in repro- ductive items, in this case egg size, finally leading to the establishment of two fairly distinct forms with, perhaps, intermediates ures are based on the youngest eggs. By the time the last stage was reached they were practically equal, so far as gross examina- tion was concerned. Plate II, Figures 2 and 3, give some indication of the differences in the extent of advancement of the eggs, espe- cially if compared with the sketches of Text- fig. i. By March 24 all the eggs were dead, pre- sumably due to extreme temperature fluc- tuations in the laboratory. Discussion. The development of the eggs of Gobiosoma bosci (Lacepede) has been described in de- 1942] Breder : Reproduction of Gobiosoma robustum 63 A B C Text-fig. 1. Stages in the development of the eggs of Gobiosoma robustum. Camera lucida sketches of living material. A. An early egg as found in nest. B. 2214 hours older than “A,” with cephalization well advanced. C. 48 Vs hours older than “A,” with the tail reaching well beyond the yolk. D. 68% hours older than “A.” E. 89% hours older than “A.” F. 117 Vi hours older than “A,” with the heart beating and the embryo showing activity. tail by Kuntz ( 1916) and discussed at length by Hildebrand & Cable (1938). A compari- son of the present notes on G. robustum with these papers shows the details to be quite similar, as would be expected on two such close species. The eggs of the latter average slightly longer and wider than those of G. bosci, as is indicated by the fol- lowing tabulation of ranges in size, given in mm. 64 Zoologica : New York Zoological Society [XXVII: 11 Species G. bosci G. robustum All Summer Winter Long Diameter 1.15-1.37 1.30-1.70 1.35-1.40 1.62-1.70 Short Diameter 0.57-0.59 0.50-0.70 0.50-0.50 0.62-0.70 It will be noted that the summer measure- ments are closer to those of G. bosci than those of the winter. The G. bosci material was likewise taken in summer. Hildebrand & Cable (1938) found young from May to December on the North Carolina coast, in- dicating a long season for this species also but with a winter interruption which is probably more a matter of geography and temperature than a specific difference. This strengthens the view that spawning may take place throughout the year on the much warmer Florida west coast. They found the young sometimes in the surface tows but more frequently in the bottom tows, indi- cating that they do not spend much time as plankton, an item also noted in the G. robustum material. Dip nets have found them in very small sizes clinging to float- ing drift, such as bits of plant stems, and it seems that most specimens taken in sur- face tows come from such locations, as even in very small sizes they sink rapidly and are not given to much active swimming other than short darts between supporting objects. Apparently sustained swimming is a considerable effort for these fishes. The eggs of Gobiosoma and a number of related genera are remarkably similar as are their reproductive habits and it would be pointless at this place to discuss the per- haps minor differences of uncertain clarity until much more is known in greater detail about more of the numerous species. References. Breder, C. M., Jr. 1939a. On the life history and development of the sponge blenny, Paraclinus mar- moratus (Steindachner). Zoologica 24 (31) : 487-496. 1939b. On the trail of the tarpon. Bull. N. Y. Zool. Soc. 42(4) : 98-110. 1941a. On the reproduction of Opsanus beta Goode & Bean. Zoologica 26(21) : 229- 232. 1941b. On the reproductive behavior of the sponge blenny, Paraclinus marmoratus (Steindachner). Zoologica 26(22): 233-236. Fowler, H. W. 1941. A collection of fishes obtained on the west coasts of Florida by Mr. and Mrs. C. G. Chaplin. Proc. Acad. Nat. Sci. Philadelphia (1940), 92: 1-22. Ginsburg, I. 1933. A revision of the genus Gobiosoma (Family Gobiidae) Bull. Bingham Oceanographic Coll. 4(5) : 1-59. Shropshire, R. F. 1932. A contribution to the life history of Gobiosoma molestum. Copeia (1) : 28-29. Kuntz, A. 1916. Notes on the embryology and larval development of five species of teleo- stean fishes. Bull. U. S. Bur. Fish., 34(1914): 407-429. Hildebrand, S. F. & Cable, L. E. 1938. Further notes on the development and life history of some teleosts at Beau- fort, N. C. Bull. U. S. Bur. Fish., 48 (24) : 505-642. EXPLANATION OF THE PLATES. Plate I. Fig. 1. Male Gobiosoma robustum guarding nest in an opened scallop shell. Plate II. Fig. 2. Photomicrograph of the living eggs in about the stage shown in Text-figure 1, A. Fig. 3. Photomicrograph of the living eggs in about the stage of Text-figure 1, E. BREDER. PLATE I Fig. 1. ON THE REPRODUCTION OF GOBIOSOMA ROBUSTUM GINSBURG, BREDER. PLATE II. Fig. 2. ON THE REPRODUCTION OF GOBIOSOMA ROBUSTUM GINSBURG. 1942] Padnos & Nigrelli: New Species of Trichodina 65 12. Trichodina spheroidesi and Trichodina halli spp. nov. Parasitic on the Gills and Skin of Marine Fishes, with Special Reference to the Life-history of T. spheroidesi. Morton Padnos New York University & Ross F. Nigrelli New York Aquarium Plates I— III ; Text-figures 1-4. Introduction. Marine members of the family Urceolarii- dae have been described from several species of invertebrates and vertebrates. Among those parasitizing fish are the following: Trichodina scorpaena Robin (1879) from the gills of Scorpaena and Trigla; T. labro- rum Chatton (1910) from two species of Symphodus; and, T. fariai da Cunha & Pinto (1928) from the intestine of the smooth puffer, Spheroides testudineus. Fantham (1918, 1919, 1924, 1980) reported Trichodina from several fishes of which the following were definitely identified as new in his 1930 paper: Trichodina clini from Clinus taures, C. super ciliosus, C. capensis, C. cottoides, C. anguillaris, Pristopoma ben- nettii and Box salpa; T. blenni from Blen- nius cornutus, T. mugilis from Mugil capito and T. chelidonichthys from Chelidonichthys cupensis. Because of insufficient information on many of the above forms, it was extremely difficult to arrive at a definite conclusion as to the validity of these species. In the pres- ent studies, the writers encountered two forms of Trichodina parasitic on the gills and skin of puffers ( Spheroides maculatus ) and other marine fishes from the New York and New Jersey coast. These peritrich cili- ates are easily distinguished from one an- other on absolute size and other characters. The one, a smaller and more abundant form, is designated Trichodina spheroidesi. The other, larger in all respects and less fre- quently encountered, is called Trichodina halli. This species can easily be distin- guished from other Trichodina infesting marine fishes on the basis of body size and size and relationships of the organelles (see Table) . Material and Methods. Between the months of June and the early part of October, from 1938 to 1940, about three hundred of the puffers taken in pound nets in Sandy Hook Bay were examined for Trichodina. Two hundred and forty-eight, or 82%, of the fish were found parasitized, the intensity varying considerably. A number of gill samples were fixed in Schaudinn’s fluid without acetic acid, and others were fixed in 10% neutral formalin. The material was stained with Heidenhain’s iron-hematoxylin and Mallory’s triple stain. A few samples of those treated by the for- mer method were counterstained with “light green.” Only the Schaudinn-fixed specimens were found suitable for cytological studies. Studies on the adhesive disc and denticulate ring were made from both the Schaudinn and formalin fixed material. A combination of the Klein’s (1926) sil- ver impregnation and the De Fano’s reduc- tion techniques was used to determine the ciliary pattern. Especially good results for such studies were obtained with material fixed with formalin. The samples were washed in five changes of distilled water, impregnated with a 5% solution of silver nitrate for a period of eight to twelve hours, and placed in darkness. Following the re- moval of excess silver nitrate solution the material was reduced in De Fano’s solution. The films were tones in a 3% solution of sodium hydrosulphite and sodium anthra- quinone sulphonate (25:1). This method, 66 Zoologica : New York Zoological Society although not delicate enough to give a dis- tinct silver line system, nevertheless effec- tively demonstrated the ciliary pattern, in- cluding the basal granules. Measurements were taken of the height, the diameter of the organism, the diameter of the adhesive disc and of the denticulate ring. A count was made of the hook number on all mounted organisms. Host-Parasite Relationships. A variety of fishes indigent to the Sandy Hook area were found to harbor Trichodina (Nigrelli, 1940). The infestation on any one host species was never found as consistently or as intense as on the puffers. Insofar as could be determined, trichodiniasis among puffers was limited to fish on the New York and New Jersey coast, for examination of this species from the coast of Massachusetts during part of the period of this investiga- tion did not reveal infection. A few migratory Trichodina were found moving about on the body surface of puffers. However, the heaviest infestation was al- ways on the gills. This may be attributed to the small opercular opening which encour- ages the concentration of these ciliates in the gill chamber. The presence of red blood cells in food vacuoles of the parasites show that they are capable of considerable tissue destruction. This is indicated further by the fact that in exceptionally heavy infestation the gill epithelium was completely destroyed, leav- ing large denuded areas among the fila- ments. Such a condition results in the death of the host. Trichodina spheroidesi sp. nov. (Text-fig. 1). Description : Organisms turban shaped. Diameter ranges from 17-54 g; height from 12-42 u. Two parallel rows of adoral, long, flexible, closely set, cilia present; cilia begin a short distance from the base of the pedun- cle, make one and a quarter clockwise, spiral turns and terminate in the vestibulum just above the mid-sectional plane of the oral surface; cilia, shortened noticeably at a point near the entrance of the vestibulum, spirals twice along the wall of the vesti- bulum and at the lower end are twisted to- gether ; direction of this spiral, when viewed from the end of the vestibulum outward, also clockwise. Aboral region with two rings of cilia; one ring, fused to form a mem- branelle, attached to inner posterior surface of the velum; second ring of cilia, found between membranelle and outer surface of the adhesive disc, are more delicate and approximately half as long as those forming the membranelle. Adhesive disc is a deep saucer-shaped organelle, ventral in position, [XXVII: 12 bordered on the dorsal side by the denticu- late ring and on the posterior side by the inner row of cilia; diameter of adhesive disc ranges from 18-32 g; striae, present on in- ner and outer walls of the disc, are argento- philic. Denticulate ring non-argentophilic ; diameter of the ring varies from 14-22 g; denticles of the ring with hooks on outer border and slender rays on inner border, joined together by triangular projections (Text-fig. 4) ; number of hooks varies from 21-31. Macronucleus is typically horse-shoe Text-fig. 1. Trichodina spheroidesi. Side view. X 950. Semi-diagrammatic reconstruction from hematoxylin and silver nitrate preparations. Ad. C., adoral cilia; V., vestibulum; Ma., macronucleus; C. V., contractile vacuole; Mi., micronucleus; D. R., denticulate ring; Ad. D., adhesive disc; Ve., velum; I. C., inner ring of aboral cilia; Ah. C., outer ring of aboral cilia. shaped, with comparatively few basophilic granules ; it lies in the posterior half of the cell, parallel to the dorsal plane of the ad- hesive disc; the open ends of the macronu- cleus extend to the oral side and surround the descending portion of the gullet. Con- tractile vacuole lies close to the descending portion of the vestibulum and during trophic life, fission and post-con jugative reorgani- zation, it is found in the posterior half of the cell. Trophic micronucleus small, lies close to the left tip of the macronucleus; during conjugation its position is variable. Remarks : Many individuals of this spe- cies were found to be infected with the suc- torian parasite, Endosphaera engelmanni. In such cases, position of the nuclei and other cytoplasmic contents may be greatly displaced. The above description is based on several hundred normal ciliates. Trichodina spheroidesi is distinguished from other Trichodina described from marine fishes by the presence of the inner ring of aboral cilia. Trichodina halli sp. nov. (Text-fig. 2). Description: Hemisphere or dome-shaped J organisms. Diameter, taken at the base, j Text-fig. 2. A. Trichodina halli. Side view. X 950. Semi-diagrammatic reconstruction from hematoxylin and silver nitrate preparations. Legend same as in Text-fig. 1. Note the absence of inner aboral ring of cilia. B. Denticulate ring. X 475. ranges from 45-86 a. Adoral cilia present, similar in arrangement and extent to T. spheroidesi. Only one ring of aboral cilia evident; membranelle absent. Disc diameter varies from 41-81 p; striae present in two layers. Denticulate ring thicker than in T. spheroidesi; diameter ranging from 30-54 p; denticles fitted together by the insertion of a cone-shaped protuberance of one den- ticle into a corresponding depression in adjacent one (Text-fig. 2, B). Number of hooks varies from 26-34; hooks shaped and curved like a ship’s propellor blade and joined to the denticles at an angle so that when observed on edge the broad area is not evident and they appear to be crescent shaped. Macronucleus as in T. spheroidesi, except that arms are longer and there are more basophilic granules present. Position of the micronucleus and other structures typical. Remarks: Trichodina halli can be distin- guished from T. spheroidesi in the follow- ing ways: (1) size of organism and organ- elles considerably larger, (2) ratio of disc diameter to the diameter of the organism as a whole is less, (3) denticulate ring thicker, (4) shape and arrangement of denticles and hooks entirely different, (5) longer arms of the macronucleus, and (6) lack of inner ring of aboral cilia. The variation in size of both T . spheroid- esi and T. halli suggests that other species of Trichodina reported in recent years may have presented a like variation in range of measurement if sufficient numbers were studied. A review of the reported species shows that in very few instances was this range adequately determined. In the table below a comparison of measurements of various structures of T. spheroidesi and T. halli with the available data of some previously reported marine species is made. As may be seen from this table all the spe- cies, except T. halli, fall within the size range of T. spheroidesi. The distinguish- ing features of most of these forms consist mainly in host specificity (?) and minor variations as to shape, ciliary pattern, na- ture of adhesive disc and denticulate ring. Since such information is not given in sufficient detail these organisms cannot be keyed out. TABLE I. Measurements (in mm.) for Trichodina from Marine Fishes. Disc Ring Hook Species Height Width Dia. Dia. No. T. blenni 20-32 40-45 24-27 24-32 T. cl ini 20 37 20 24 T . chelidonichthys 19-27 30-45 19-32 30 T. mugilis 14-20 33-44 23-28 32 T. fariai 32 40 20-42 24-28 T. labrorum 18-22 30-34 21 T. spheroidesi 12-42 17-54 18-32 14-22 21-31 T. halli 45-86 41-81 30-54 26-34 Binary Fission in Trichodina spheroidesi. (Text-figs. 3 & 4). The stages of fission in T. spheroidesi correspond very closely to those described by Diller (1928) for the species of Tricho- 68 Zoologica : New York Zoological Society [XXVII: 12 Text-fig. 3. Binary fission in Trichodina spheroidesi. X 950. From iron-hematoxylm preparations. A, trophic stage; B, macronucleus in stage of contraction; C, D, and E, further contraction of the macronucleus; B and C, note swelling of micronucleus; U, micronucleus in metaphase; E, telophase; F. completed micronuclear division. 1942] Padnos & Nigrelli: New Species of Trichodina 69 G Text-fig. 4. Trichodina spheroidesi. X 950. Iron-hematoxylin preparations. G, II, I, J, binary fission stages showing the final division of the macronucleus and the reorganiza- tion of the denticulate ring. C. V., contractile vacuole; N. R., new ring; O. R., old ring. dina from tadpoles. The trophomacronueleus undergoes vacuolization and clefts appear in the ground substance (Text-fig. 3, B). The macronucleus begins to condense while the chromatin granules, large in the trophic stage, become smaller in size and randomly distributed in the matrix. As the macro- nucleus contracts further, the micronucleus swells, eventually becoming spheroidal, then divides mitotically (Text-fig. 3, C, D, E). A metaphase stage of a dividing micro- nucleus is shown in Text-fig. 3 D. The mito- tic division continues as the macronucleus pulls apart, and two daughter micronuclei are formed before the maci'onucleus is com- pletely divided (Text-fig. 3, F; Text-fig. 4, G). Plasmotomy takes place about the time of the late telophase. A stricture appears in the ventral portion of the organism, and the adhesive disc and denticulate ring separate into approximately equal halves (Text-fig. 4, G, H). Final cleavage of the macronucleus takes place and two daughter cells are formed (Text-fig. 4, H). Peshkowsky (1923) reported that during division of Trichodina steinii and T. mitra the adoral cilia, gullet and contractile vacu- ole are absorbed. Careful study of the silver impregnated material, however, leads the 70 Zoologica : New York Zoological Society [XXVII: 12 writers to believe that the adoral and the aboral zones of cilia, in T. spheroidesi at least, are retained throughout division. The fate of the vestibulum and the vestibular cilia could not be traced. In specimens stained with hematoxylin, it was observed that the contractile vacuole is retained also during cell division. It appears to cleave concomitantly with the cleavage of the macronucleus (Text-fig. 4, G). The anlage of the new denticulate ring (corona) is laid down as a delicate ring close to the posterior border of the adhesive disc before the onset of cell division (Text- fig. 4, G). This confirms the observations of Diller (1928). After the formation of the daughter cells the new corona gradually develops its denticles and hooks and takes up a more dorsal position on the adhesive disc. The rays are the last structure to be formed in the ring. Half of the old ring, carried over after cell division by each daughter cell, is pushed into the ventral plug of the cytoplasm between the aperture of the adhesive disc. Here it is slowly resorbed (Text-fig. 4, I, J). Conjugation in Trichodina spheroidesi. (PI. I— III) . Conjugation in T. spheroidesi is aniso- gamous and the process is very similar to that reported by Maupas (1888) for Vorti- cella nebulifera (see also Doflein, 1927). The aboral surface of the microconjugant is fitted over the adoral surface of the macro- conjugant (Pis. I, II). They may or may not be oriented in the same direction. After the conjugants have assumed their respective positions, the micronucleus of each begins to swell, eventually becomes vesicular and passes from the original posterior position to a more central location in the cell (PI. I, Figs. 1-3). The macronucleus begins to show signs of vacuolization (PI. I, Fig. 2). At the time of spindle formation, the macronucleus twists and pulls apart into large coarse fragments (PI. I, Fig. 3). These pieces continue to break up into small- er and smaller parts until minute spherical bodies with deeply staining granules are formed (PL I, Figs. 4, 5, 6, 7, 8; PL II, Fig. 9). Micronuclear activity immediately preced- ing the metaphase period is not clearly de- fined (PL I, Fig. 4). It is possible that the micronuclei have a decreased affinity for iron-hematoxylin stain during these stages. The metaphase spindle, however, is clear and granular chromosomes can be identified about the center (PL I, Fig. 5). During the final fragmentation of the macronucleus, two micronuclear divisions take place in each conjugant (PL II, Fig. 10). It is of interest to point out here that conjugation in T. spheroidesi differs from conjugation in Vorticella nebidifera in that there are only two micronuclear divisions in each conjugant instead of three. Protoplasmic continuity is established be- tween the conjugating individuals (PL II, Fig. 11), and the contents from the smaller individual pass into the larger one (PL II, Fig. 12). It is assumed that the gametic nuclei (PL II, Fig. 10) then combine to form the synkaryon, and the remaining nuclei are resorbed. The remains of the microconju- gant collapse and the ensuing processes of conjugation are confined to the single large exconjugant (PL II, Fig. 13). At the onset of its first division, the zygotic nucleus develops a larger spindle than any of the dividing nuclei in the early stages of conjugation. It is assumed that three mitotic divisions subsequently occur, resulting in eight micronuclei. Seven of these become the macronuclear anlage and one the functional micronucleus (PL II, Figs. 14, 15). The functional micronucleus divides (PL II, Fig. 15) and in the cell divi- sion which takes place, the macronuclear anlage are distributed between the daughter individuals ( PL III). The most frequent dis- tribution is three and four (PL III, Figs. 16, 17). However, the distribution may some- times be two and five (PL III, Figs. 18, 19) or one and six. Cell division continues until each of the daughter cells formed contain one macronuclear anlage (PL III, Fig. 20). The macronucleus then increases in size and develops its characteristic horse-shoe shape. Reorganization of the denticulate ring occurs in the macroconjugant shortly after the fusion of the protoplasmic contents of the conjugants has occurred. Figures 12 and 13 of Plate II show the newly formed ring together with the remains of the old den- ticulate ring. It should be pointed out that the original ring is present in preceding stages but is not shown in the figures for sake of clarity. It is of interest to note that the number of denticles in the new ring is invariably the same as the number present in the old ring. Such structures as the gullet, vacuole and cilia are present throughout conjugation. Discussion. Trichodina spheroidesi is distinguished from other Trichodina of marine fishes by the presence of the inner ring of aboral cilia. Wetzel (1927) and Precht (1935) re- ported a similar observation for T. pediculus and T. scoploplontis, respectively. The double layer of striae in the adhesive disc observed by the writers in the two forms from the puffers have been previ- ously reported by Mueller (1932, 1938) for Trichodina renicola and Vauchomia nephri- tica. Mueller (1938) refers to the inner group of striae as the posterior hard rays and the outer group as the anterior soft 1942] Padnos & Nigrelli: New Species of Trichodina 71 rays. He further commented that the soft rays comprise a system of myonemes which connect with the posterior girdle of cilia. This duplex nature of the striae Mueller featured to distinguish members of the Trichodinidae found in the urinary tract of fishes (e.g., T. renicola and V. nephritica from Esox niger and Esox masquinongy , respectively) from those found on the gills of certain fresh-water fishes reported by him in 1937. In the ciliates from the gills only the hard rays were present in the stri- ated ring. However, the occurrence of double striations in the forms described by the writers suggests that this feature may be more widespread than heretofore has been observed. The similarity of conjugation in Tricho- dina pediculus to that of certain members of the Vorticellidae was first pointed out by Busch (1855). Stages in conjugation among certain of the Urceolariidae have since been recorded and the same comparison made. Caullery & Mesnil (1915) reported stages in conjugation in Trichodina patellae (Cuenot) from a species of fresh-water mollusc. Peshkowsky (1923) described con- jugation in T. steinii and commented that this process was similar in all essential fea- tures to that observed in the Vorticellidae. Anisogamous conjugation was described by Zich (1928) for Urceolaria korschelti, and by Hunter (1936) for two types of Tricho- dina found in the intestine of sea-cucum- bers. Diller (1928) described in detail en- domixis in Trichodina from tadpoles. He suggested that some of the evidence of con- jugation presented by Caullery & Mesnil, Peshkowsky and others were most likely stages in endomixis. The endomictic stages described by Diller, however, conform close- ly to the nuclear reorganization in the post- con jugative stages described for Trichodina spheroidesi. Diller makes a distinction be- tween endomictic individuals and conjugat- ing forms on the basis of differences in shape of macronuclear fragments. Since the fragments in Diller ’s material are not unlike those observed in some of the stages in conjugation in T. spheroidesi this distinc- tion cannot be supported. Any other differ- ences that Diller may have noted are, in all probability, the result of examining too few samples of conjugation which he stated was present in his material. That he may have misinterpreted post-conjugation for endo- mixis is supported further by the similarity of his figure (PI. II, Fig. 14) of the first micronuclear division to the figure of the zygote nucleus in Vorticella nebulifera shown by Maupas (1888) (see Doflein, Fig. 310) and in Trichodina spheroidesi (PI. II, Fig. 13). Summary. 1. Trichodina spheroidesi and T. halli spp. nov. from the gills and skin of puffers ( Spheroides maculatus) and other marine fishes from the New Jersey and the New York coast are described. 2. The processes of fission and conjuga- tion are described for T. spheroidesi. References. Busch, W. 1855. A Manual of the Infusoria by Kent. 2: 647. Caullery, M. & Mesnil, F. 1915. Sur Trichodina patellae (Cuenot) symbiose avec des zooxanthelles ; struc- ture, division, conjugasion. C. R. Soc. Biol. 78: 674. Chatton, E. 1910. Protozoaires parasites des branches des Labres; Amoeba mucicola, Tricho- dina larorum n.sp., Appencide: Para- site des Trichodines. Arch Zool. Exper. 5: 239. Cunha, A. da & Pinto, C. 1928. Trichodina fariai n. sp. cilie peritriche endoparasite de Poisson marin. C. R. Soc. Biol. 98: 1570. Diller, W. F. 1928. Fission and Endomixis in Trichodina from tadpoles. Jour. Morph, and Physiol. 46: 521. Doflein, R. 1927. Lehrbuch der Protozoenkunde. 1 : 279. Fantham, H. B. 1918. Some parasitic protozoa found in South African fishes and amphibians. S. Afr. J. Sci. 15: 338. 1919. Some parasitic protozoan found in South Africa. II. S. Afr. J. Sci. 16: 185. 1924. Some parasitic protozoa found in South Africa. VII. S. Afr. .7. Sci. 21: 435. 1930. Some parasitic protozoa found in South Africa. S. Afr. J. Sci. 27: 376. Hunter, L. 1936. Some nuclear phenomena in the Trich- odina from Thyone briascus (Holo- thuroidea). Biol. Bull. 71: 408. Klein, B. M. 1926. Ergebnisse mit einer Silbermethode bei Ciliaten. Arch. f. Prot. 56: 243. Maupas, E. 1888. Recherches experimentales sur la mul- tiplication des Infusoires cilies. Arch. Zool. Exper. et Gen. Ser. 2. 6: 165. Mueller, J. T. 1932. Trichodina renicola, a ciliate parasite of the urinary tract of Esox niger. Roosevelt Wild Life Annals. 3: 139. 1937. Some species of Trichodina (Ciliata) from fresh water fishes. Trans. Amer. Micr. Soc. 56: 177. 1938. A new species of Trichodina from the urinary tract of the muskelonge, with a reparition of the genus. Jour. Para. 24: 251. 72 Zoologica : New York Zoological Society [XXVII: 12 Nigrelli, Ross F. 1940. Mortality statistics for specimens in the New York Aquarium, 1939. Zoolog- ica. 25: 525. Peshkowsky, L. 1923. Biological and structural studies of Trichodina. Russkii Archiv Protisto- logic. 2: 249. Precht, N. 1935. Epizoen der Keiler Bucht. Nova Acta Leopoldina Neue Folge. 3: 405. EXPLANATION OF THE PLATES. Anisogamous conjugation in Trichodina spheroidesi. All figures drawn from material stained with iron-hematoxylin. X 950. Plate I. Fig. 1. Start of conjugation. Fig. 2. Macronucleus in the process of frag- mentation. Note lumping of nuclear material. Micronucleus in pre-meta- phase stages of meiosis. Fig. 3. First macronuclear fragmentation. Micronucleus still in pre-metaphase stage. Fig. 4. Further macronuclear fragmentation. Fig. 5. Metaphase of meiotic nucleus clearly evident. Fragmentation of the macro- nucleus continues. Figs. 6, 7, 8. Continued meiotic division and completion of fragmentation of the macronucleus into many spherical and oval shaped bodies of various size. Plate II. Fig. 9. Continuance of the meiotic process seen in Fig. 8. Fig. 10. Second micronuclear division. Note persistent gametic micronucleus and the degeneration of other three micro- nuclei in each conjugant. Fig. 11. Gametic nuclei in an early stage of fusion to give rise to the synkaryon. The other micronuclei completely dis- appeared. Note the completion of cyto- Robin, C. 1879. Memoire sur la structure et la repro- duction de quelques infusoires. J. Anat. et Physiol. 15. Wetzel, A. 1927. Beitrag zur Kenntniss der Trichodina pediculus (Ehrb.). Zeitschr. f. Wiss. Biol. Abt. A. 9: 719. Zick, K. 1928. Urceolaria korschelti, n. sp., eine neue marine Urceolarine, nebst einem Uber- lich uber die Urceolarinen. Zeitschr. f. Wiss. Zool. 132: 335. plasmic continuity between the con- jugants. Fig. 12. The cytoplasmic contents of the micro- conjugant pass into the macroconju- gant. The new denticulate ring devel- ops while the old ring is resorbed. In this individual the macronuclear frag- ments are coarser than in the preced- ing stage. Fig. 13. Post-conjugant stage. Macroconjugant with large zygotic nucleus in meta- phase stage. New denticulate ring and remnant of the old ring still present. Fig. 14. Initial stage in development of the macronuclear anlage. There are sev- en of these larger bodies present, in- dicating that three divisions of the zygotic nucleus had taken place. Fig. 15. Further development of the macro- nuclear anlage. The start of the first binary fission. Micronucleus in meta- phase. Plate III. Figs. 16, 17. Daughter cells showing the three and four macronuclear anlage distri- bution. Figs. 18, 19. A two and five distribution of macronuclear material. Fig. 20. Each cell will continue to divide until only a single macronuclear anlage is present in each individual. The last step in this process is shown in this figure. PADNOS a NIGRELLI. PLATE I. TRICHODINA SPHEROIDES1 AND TRiCHODINA HALLI SPP. NOV. PARASITIC ON THE GILLS AND SKIN B OF MARINE FISHES, WITH SPECIAL REFERENCE TO THE LIFE-HISTORY OF T. SPHEROIDESI. PADNOS a NIGRELLI. PLATE TRICHODINA SPHEROIDESI AND TRICHODINA HALLI SPP. NOV. PARASITIC ON THE GILLS AND SKIN OF MARINE FISHES, WITH SPECIAL REFERENCE TO THE LIFE-HISTORY OF T. SPHEROIDESI. PADNOS a NIGRELLI. PLATE III. 20 TRICHODINA SPHEROIDESI AND TRICHODINA HALLI SPP. NOV. PARASITIC ON THE GILLS AND SKIN OF MARINE FISHES, WITH SPECIAL REFERENCE TO THE LIFE-HISTORY OF T. SPHEROIDESI. 1942] Gordon: Albino Embryos of Xiphophorus hellerii 73 13. Mortality of Albino Embryos and Aberrant Mendelian Ratios in Certain Broods of Xiphophorus hellerii.1 Myron Gordon New York Zoological Society ( Text-figure 1 ) . In a previous publication,2 the author has shown that when the albino swordtail is mated with the golden variety, their off- spring resemble neither parent but revert to the ancestral gray coloration of the wild species. The author showed further that when the gray offspring of the above mating reached maturity and were mated together, brother to sister, the second generation population consisted of three color classes: 202 were “wild” gray, 65 were golden and 67 were albino. These frequencies approached the theoretical 9:3:4 ratio and were interpreted on the basis of the recombination of two independent autosomal factors: St I for “wild” gray, st I for golden and St i and st i for albino. To get the above data the results of four matings were pooled, as Table I will show. Table I. Second Generation Offspring of Gray {Stst Ii ) Swordtails. Female No. Gray Golden Albino 70-4 37 12 8 70-5 38 13 6 70-6 50 15 20 70-7 77 25 33 Totals observed 202 65 67 expected 188 63 83 It was noted that there was a deficiency of albinos in the above totals, particularly in broods from females 70-4 and 70-5, yet the two independent factor hypothesis seemed most appropriate in explaining the results. This was confirmed when the Chi-Square test for significance was applied to the pooled data ; the value of Chi-Square was 1 I wish to express my thanks to the Department of Birds of the American Museum of Natural History for the use of their laboratory in the Whitney Wing of the Museum. 2 Back to their Ancestors, Journal of Heredity, 32: 385-390, 1941. found to be 4.18. However the value of Chi- Square for the first two broods alone was 9.84 and indicates the deviations found here cannot be attributed to chance. Broods were obtained from three addi- tional females after mating with their gray brothers. The three females, 70-8, 70-9 and 70-10, yielded ratios so obviously aberrant that counts of their young were not in- cluded in the original presentation because it was thought, at the time, that a contami- nating factor was involved in this portion of the experiment. Further study was under- taken to account for these unusual results. Reports from the laboratories of other workers and from our results as indicated in Table I pointed to a deficiency of albinos in genetic tests. This was attributed to the constitutional weakness of the albino indi- viduals. It was noted, for instance, that if the fishes of a brood are counted soon after birth and again after two months, the death rate of albinos is greater than that of the other color varieties. From this evidence it was suspected that the weakness of the albinos may be manifest early, perhaps in their embryonic stages of development. The three additional females, 70-8, 70-9 and 70-10, that produced hardly any albinos in their first broods (see Table II), were killed three weeks after giving birth to their last brood. The females were dissected, their embryos were removed and counts were made of them. In each case an appropriate proportion of albinos was found among the gray and golden embryos. This may be seen in Table II. Discussion and Conclusions. It is obvious that the numbers of albinos are decidedly deficient among the young fishes from females 70-8, 70-9 and 70-10; the extremely high value of Chi-Square, 42.46, indicates that the results obtained are not due to chance variation alone. On the other hand, the numbers of albinos 74 Zoologica : New York Zoological Society [XXVII: 13 9 “Wild” Gray 4 Albino Text-fig. 1. Embryos from gray females mated to gray males. Table II. Offspring of Gray ( Stst Ii) Swordtails. Female No. Gray Golden Albino 70-8 38 11 2 Y oung 35 10 13 Embryos 70-9 45 10 1 Young 40 9 12 Embryos 70-10 58 18 1 Young Totals: 40 12 17 Embryos Observed 141 39 4 Young Expected 103.5 34.5 46 Observed 115 31 42 Embryos Expected 106 35 47 found among the embryos of the same fe- males are in close harmony with expectancy. The slight deviations found may be at- tributed to chance, for the Chi-Square value is low, 1.75. In light of all the data, the gray females tested may be divided into three groups ac- cording to their ability to produce viable albino swordtails. Females 70-8, 70-9 and 70-10 fail in this respect almost completely; females 70-4 and 70-5 fail to produce an adequate number according to expectancy; females 70-6 and 70-7 have the ability to produce viable albino young in normal numbers. Thus the deficiency of albinos in certain broods must be attributed not only to the low viability of the albinos themselves but also to some failure in the ability of the albinos’ mothers to carry their complete broods through to birth. 1942] Shlaifer : Schooling Behavior of Mackerel 75 14. The Schooling Behavior of Mackerel: A Preliminary Experimental Analysis. Arthur Shlaifer Neiv York Aquarium & U. S. Fish & Wildlife Service, Woods Hole, Mass. (Plate I). Introduction. As indicated in the comprehensive review of the field by Allee (1931), animal aggre- gations and their significance have attracted increasing critical analysis in recent years. Curiously, the fish school — one of the most striking examples of well-integrated animal aggregations — has received remarkably meager critical attention. Spooner (1931) and Breder & Nigrelli (1935), considering various aspects of schooling in fishes, have noted this unfortunate lack of literature. Some notes on the schooling behavior of the herring, Clupea harengus L., by Newman (1876), constitute one of the earliest pub- lications in this field. Nevertheless, Parr (1927) after a fifty year interval had a virtually clear field for his theoretical analysis of the schooling behavior of mackerel. Fishes which do not school and have no visible aggregating tendency have been con- sidered “non-social” forms. However, it is unwise to so classify a form which mani- fests no obvious social tendencies. Statisti- cal evidence has demonstrated that the gold- fish, which had been summarily thus dis- missed, manifested a definite group effect; isolated individuals were found to have a higher rate of locomotor activity and oxy- gen consumption than did grouped gold- fishes (Schuett, 1934; Escobar, Minahan & Shaw, 1936; Breder & Nigrelli, 1938; Shlaifer, 1938). Hence, social tendencies may be relatively obscure and not readily determined by casual observation. At the extremes, it is a relatively simple matter to differentiate between a loosely ag- gregating fish and a closely schooling one. However, in many cases it is exceedingly difficult to determine whether a fish should be considered a closely aggregating form or a loosely schooling one. The mackerel im- poses no such difficulties. It remains in dense schools throughout life, except for possible dispersal at night. It is difficult to think of an instance, at least among verte- brates, in which individuality is as com- pletely lost as it is in a mackerel or herring school. Parr (1927) subjected the mackerel school to critical theoretical analysis. Sev- eral interesting conclusions were reached which will be considered later. His report, however, contains relatively little experi- mental data. It is the purpose of this re- port to treat the phenomenon experimentally and, wherever possible, to attempt correla- tion with Parr’s theoretical conclusions. The writer wishes to express his deep appreciation to the U. S. Fish and Wildlife Service at whose Woods Hole station the experiments were performed ; also to Dr. R. Buchsbaum whose photographic skill is responsible for the figures in the plate. Experimental Study. Number of Individuals Needed to Form a School: The experimental animals used in these and subsequent tests were 8-inch specimens of the chub mackerel, Pneuma- topliorus grex (Mitchill). Originally caught in the waters off Woods Hole, they were sub- sequently kept in appropriately large tanks in the laboratory. Several days were allowed for acclimatization before the animals were used. Two tanks were employed for experi- mental purposes. One was a rectangular as- sembled aquarium with transparent glass sides whose dimensions were 36" by 15" and 17" deep. The other was a rectangular wooden tank 44" by 23" and 9.5" deep. Sea water was kept running through these tanks at all times at a fairly rapid rate. The aver- age oxygen content was 5.60 cc. per liter and the temperature range was 18-20 de- grees C. In general, the mackerel survived well. Of those that died, many expired dur- ing the course of acclimatization in the laboratory. Most of those which survived this period lived for several weeks in no apparent distress. Repeated experiments demonstrate that 76 Zoologica : New York Zoological Society [XXVII: 14 two individuals suffice to form a mackerel school. If two fishes are removed from a group and placed in the wooden tank they immediately school, i.e., swim about the tank nearly always in line with each other. If one mackerel is kept isolated in a tank and another is grouped with it, they immedi- ately school. Apparently, the fish that first becomes aware of the presence of the other initiates the schooling reaction. Whether this will be the original fish or the intro- duced one is purely a matter of chance for it will be one or the other 50% of the time. Other fishes added to this group of two im- mediately join the school. Thus, only two individuals are required to begin the formation of a mackerel school which is, however, better integrated if com- posed of many individuals. Breder & Nigrelli (1935) found that two sunfishes, Lepomis auritus, grouped to- gether in a tank “aggregate” with each other after two days. Shlaifer (1938) demonstrated that the oxygen consumption and locomotor activity of an isolated gold- fish is significantly higher than is that of an individual in a group of two. It would seem, at least in the cases listed above, that there is a much greater psychic difference between an isolated fish and one in a group of two than between the latter and an individual in a much larger group. Effect of Various Types of Blinding: Parr (1927 ) found that when chub mackerel were blinded by the application of vaseline and lampblack to the eye, they did not school or mill. Blinded catfishes do not aggregate (Bowen, 1931) nor do blinded sunfishes (Breder & Nigrelli, 1935). Grouped gold- fishes which are normally less active than isolated individuals lose this group effect when they are blinded (Shlaifer, 1939). These results as well as other lines of evi- dence indicate the importance of vision in integrating social behavior in fishes. The experiments described below were de- signed to repeat and to extend the original work of Parr (1927) on this species. Mack- erel were blinded by piercing the cornea and were kept in the wooden tank described above. One day was allowed for recovery from operative shock. Blindness was ascer- tained by appropriate tests, e.g., failure to avoid a net, etc. All results noted for the various blinding experiments were con- firmed by repeated tests. As controls, the area in the vicinity of the eye of non- blinded mackerel was pierced by the same instrument used for blinding, thus approxi- mating similar conditions of shock. In no case did control individuals fail to school when grouped. If an individual is blinded on one eye and subsequently grouped with six schooling mackerel, it immediately joins the school. In fact, if only three minutes are allowed for recovery from the shock of the operation instead of the customary day, it also re- joins the school immediately. This half- blinded fish succeeds in maintaining its orientation with respect to the rest of the school fairly well. Sudden turns by the normal animals to the blinded side of the experimental fish may result in temporary loss of integration of that animal with the school ; however, it is quickly recovered. If a mackerel is subjected to bilateral blinding, it makes no attempt to join the school. Occasionally its random movements about the tank may disrupt the smooth in- tegration of the normal school but only for a moment. The grouping of a normal fish with a half-blinded one results in a schooling re- action. In general, the unilaterally blinded animal will orient itself so that the intact eye side is the one nearest the normal fish. Sudden turns by either animal will initiate a turn in the other, thus maintaining the school. If the half-blinded fish is then blinded on the other eye, the school disinte- grates. There is, of course, no reaction by the sightless form to the normal one. How- ever, it might be expected, inasmuch as there is no other mackerel in the tank but the blinded individual, that the normal fish would attempt to school with it. This is not the case. Sightless mackerel do not swim in a typi- cally normal manner; movement is slower and less uniform. Apparently, normal swim- ming movement is of great importance in the schooling reaction of mackerel. If two mackerel are blinded on the same eye, a school obtains though it is not as well integrated as is a school of two normal indi- viduals or one normal and one half-blinded form. In this case the maintenance of the school is dependent upon the behavior of the fish whose intact eye side is nearest the other animal whose swimming movement is apparently sufficiently normal to evoke a schooling reaction by the mackerel which is in visual contact with it. We are presented with the unusual case of a school of two fishes, one of which plays a passive role. Sudden turns and changes in direction may reverse the role of either fish. The mackerel which sees the other member of the group of two usually follows the turns of the pas- sive partner. On occasion, however, it may initiate a turn, in which case the school is broken for a second or so until one indi- vidual finds the other. If, in a group of two, one mackerel is blinded on its left eye and the other on its right one, schooling behavior is very erratic. If their blinded sides face each other no re- action obtains; if not, they school, though the school is likely to be broken by a sudden sharp turn by one animal which results in 1942] Shlaifer: Schooling Behavior of Mackerel 77 their blinded sides facing each other again. A totally blinded fish grouped with a half-blinded one evokes no good schooling reaction by virtue of the abnormal move- ment of the former. Two totally blinded in- dividuals grouped together will swim at random and may collide. Confirming Parr’s data (1927), three mackerel blinded in both eyes and placed in an exhibition tank containing a large school of mackerel swim aimlessly and make no at- tempt to join the group. The results obtained demonstrate the role of vision. Also indicated is the importance of normal swimming movement. Any experiment in which visual response is eliminated through blinding suffers from the criticism that the normal physiological state of the animal may be disturbed. This may be true even when a period deemed to be sufficient for recovery from shock ob- tains. A more natural condition is darkness and a description of the .behavior of mack- erel in this state follows. The Effect of Darkness on Schooling : Newman (1876) finds that the closely schooling herring, Clupea harengus, break up completely at night in a tank in captivity, each fish taking an independent path. The school reforms in the presence of sufficient light. Breder (1929) reports that compact schools of Jenkinsia, also a member of the family Clupeidae, are dispersed at night. Bowen (1931) finds that aggregations of catfishes are dispersed in darkness. Breder & Nigrelli (1935) report that aggregations of the sunfish, Lepomis auritus, break up with the coming of night. Shlaifer (1939) finds that the effect of grouping (decreased oxygen consumption) on goldfishes dis- appears in total darkness. The behavior of mackerel in darkness was investigated by two procedures, one obser- vational, the other photographic. One of the exhibition tanks of the Aquarium at the Woods Hole station of the U. S. Fish and Wildlife Service contains a school of forty to fifty 8-inch chub mackerel, as well as a turtle, flounder, blackfish, and skate. The tank, rectangular in shape, is 6.5 feet by 4 feet and 3 feet deep. Three of the verti- cal sides are composed of stone and cement and the fourth of transparent glass. The lights in the Aquarium were always off at night and after 9 p.m. the room was quite dark. The observer, looking at the tank from above or standing next to the trans- parent glass side, could see nothing in the tank; in fact, one’s hand held two inches from the eye was quite invisible. For several hours at night during several consecutive evenings observations were made on the schooling behavior of the mackerel in this state of darkness. At half-hour intervals a flashlight beam was directed at the bottom of the exhibition tank for only one or two seconds. If maintained longer, the fishes would react to the light, weak though it was, by forming a dense school. Hence, it was necessary to form an impression of the aggregating condition of the group in this very short time. The general impression gathered by these observations was that the school was fairly well dispersed. The fishes were never found to be closely schooling or milling but neither were they swimming about the tank at random as Newman (1876) reported for the herring in darkness. The mackerel swam in an elliptical orbit more or less in the same direction but with con- siderably greater distances between indi- viduals than is found in a normal school in the light. Considering the number of indi- viduals, the size of the tank, and the ten- dency of mackerel to swim in a uniform way for hours unless disturbed, the observed state of the school is probably what is to be expected in the absence of visual integra- tion. Further experiments are planned along these lines. The mackerel eye is apparently capable of detecting similarly moving forms at very low light intensities for when they could just barely be seen they were in fairly com- pact schools. This is in contrast to the find- ings of Breder & Nigrelli (1935) for sunfish aggregations which break up when they can still be seen distinctly. Another interest- ing fact is that the mackerel also is capable, apparently, of seeing in light at the deep red end of the spectrum. When a Wratten Series II Safelight, which transmits light in the deep red from about 650nn. to 700m.u., was suspended at night over the tank several inches from the water surface in otherwise total darkness, the mackerel formed fairly compact schools and mills. (See Plate I, Fig. 3.) The observational method described above is open to the criticism that the observer’s reaction must be instantaneous and is sub- jective. Accordingly, a series of flash-bulb photographs was made of the mackerel group in artificial light and in total dark- ness. These flash-bulb photographs are taken in a fraction of a second, much too fast for any disturbance caused by the blinding flash of light to be recorded in the photograph. The school was photographed from the side through the transparent glass. Darkness shots were taken only after a period of at least 15 consecutive minutes of darkness following the small amount of illumination from a flashlight incident to setting up the equipment. Darkness set in at about 9 p.m. and the first photograph was usually taken at about 10 p.m. Plate I contains two photo- graphs which represent typical results. Fig. 1 is a compact mackerel school under fairly strong artificial light. Fig. 2 was taken in total darkness at about 10:30 p.m. The con- trast in the denseness of the aggregations 78 Zoologica : New York Zoological Society [XXVII: 14 is obvious and portrays the results obtained by observation. The observational and photographic evi- dence reveals that mackerel schools appar- ently are dispersed in total darkness. How- ever, the ability of mackerel in aquarium tanks to school at very low light intensities leaves unsolved the condition of mackerel schools in nature where dim light may ob- tain at night. It is the intention of the writer to pursue this matter more exhaus- tively in future work. Visual Contact: Shlaifer (1939, 1940) found that the oxygen consumption and locomotor activity of isolated goldfishes in visual contact with others of the same spec- ies and variety was of the same order of magnitude as that of these individuals when actually members of a group. This confirmed previous results which demonstrated the visual integration of the group effect. Although darkness and blindness tests have indicated the importance of vision in the schooling behavior of mackerel, the follow- ing experiment was performed as a bit of additional evidence. The glass tank whose dimensions have been listed above was divided in half along its length by a plate of transparent glass. One mackerel was placed on each side. After a short period of acclimatization, the two fishes tended to swim close to the dividing glass plate in line with each other. In gen- eral, the animals would turn only when they reached one end of the tank and would then swim back to the other end. If, how- ever, one mackerel turned in the center of the tank, the fish on the other side of the glass in visual contact with it would usually also turn before reaching the end of the tank. This behavior was not invariable but occurred with sufficient frequency to be con- sidered significant. Thus, from three lines of evidence — blinding, darkness, and visual contact ex- periments— the important role of sight is demonstrated. Response to Form, and Movement : Spooner (1931), working with the bass, Morone labrax, which is a schooling foi'm, found that individuals would be attracted to dead, mounted, specimens of the species but not to rough models. Similar results were obtained for the goldfish (Shlaifer, 1939, 1940). Thus, there is indicated that there may be a visual response to objects of the proper form though they are devoid of movement. Response to form is also re- ported by Breder (1929) for the schooling herring, Jenkinsia, and by Breder & Nigrelli (1935) for the aggregating sunfish, Le- pomis auritus. On the other hand, many sexual behavior studies emphasize the im- portance of movement (Noble, 1934; Breder, 1936). Experiments demonstrate that a mackerel isolated in a tank with freshly killed and mounted specimens placed in the normal swimming position, does not react to them. On the other hand, if another normal mack- erel is introduced into the tank, the school- ing reaction is immediately evoked. Repeated experiments were performed with freshly killed mackerel which were manipulated by means of a long rod, the hooked end of which was inserted in the back of the fish. A normal mackerel was grouped with a mounted fish which was then manipulated so as to simulate a normal swimming animal. In only one case in 70 trials was there any response given to the dead specimen. Again, normal mackerel in- troduced into the experimental tank elicited immediate schooling. An olfactory basis for the lack of response to a dead, manipulated individual is not probable by virtue of the fact that the animals were freshly killed and were in fairly rapidly flowing water. it may be concluded that normal swim- ming movement is an important factor in the schooling reaction. True, with normal movement a mackerel may be attracted by the body form of its neighbor but, in the absence of normal movement, form alone will not suffice. The importance of normal movement is further emphasized by the failure of a mackerel to school with a blinded individual which does not swim in the usual manner. Evidently, mackerel are sensitive to differences in motion which we can also detect and quite possibly to minor differences which we cannot observe. Never- theless, further experimentation along these lines is in order. If simulation of swimming motion in a killed mackerel can be skillful enough to evoke a schooling reaction, neat checks might be obtained on response to form by altering in many ways the shape of the dead specimen. Response to Color: The reaction of fishes to colors is still the subject of considerable debate. Warner (1931) criticized the lack of control of the intensity factor and deemed most of the experimental work worthy of repetition. White (1919, 1927) demon- strated that mudminnows and sticklebacks can discriminate between wave lengths and not merely -intensities of light. Brown (1937) found that the large-mouthed black bass, Huro salmoides, responds to differ- ences in wave lengths. Noble & Curtis (1939) demonstrated that young cichlids may be born with a greater interest in moving red discs than in moving black, blue, green or yellow ones. Shlaifer (1939), though not differentiat- ing between wave length and intensity, found that the group reaction in the gold- fish was not in any way based on color dif- ferences or similarities. Accordingly, ex- 1942] Shlaifer : Schooling Behavior of Mackerel 79 periments were performed with mackerel to determine whether response to color was in any way involved in the schooling behavior. Mackerel were removed from the tank and paints of various colors were applied over all of the body but the eye. They were then allowed a period of recovery from shock and were grouped in various combinations. Specimens painted blue were grouped with ones painted white, black, etc. In all cases, schooling occurred immediately with no in- dication whatsoever of differential response to color. Effect of Isolation: Bowen (1932) found that the sight response of normal aggre- gating catfishes to one another was not completely eliminated in all individuals by 161 days of isolation. It was much less marked but was re-established in the course of a few minutes, usually after contact oc- curred. Catfishes isolated for only 52 days when grouped together showed no difference in behavior from those animals kept in a group. Five mackerel were isolated for 20 days, the maximum time available before the laboratory closed. At the end of this period, two of the five were grouped together as were the remaining three. Schooling oc- curred immediately. If mature mackerel could be kept in isolation for much longer periods, it would be interesting to observe subsequent schooling behavior. Even more interesting would be the subsequent school- ing reaction of mackerel reared in isolation from various early stages. Discussion. The experimental results reported above confirm the data of Parr (1927) on the visual integration of the mackerel school. Further, it is seen that only two normal in- dividuals are necessary to begin a school. The visual response is apparently not cor- related with color but is with normal swim- ming activity. Finally, several weeks of iso- lation do not induce any weakening of the schooling reaction. The fact that two individuals suffice to begin a school, at least under laboratory conditions, may have implications for larger aggregations in nature. If schools are com- pletely dispersed at night, their reforma- tion in daylight would be definitely facili- tated by the mutual attraction of only two fishes. In fact, again granting the break up of the school at night, without a schooling response by one solitary fish to another one, it is difficult to see how schools could reform. In reference to the apparent visual inte- gration of fish schools, Parr (1927) indi- cates that schooling pelagic fishes have eyes of large size and rather scantily equipped lateral line systems. He concedes that lateral line stimuli might come into play once the mackerel, by visual stimuli, approached each other. However, his data and those reported above on the behavior of blinded fishes would not tend to confirm this hypothesis. Parr attributes the schooling reaction of mackerel to a simple automatic eye reflex rather than to a social instinct involving the entire school. The apparently senseless milling reaction is caused, he believes, when the school as a whole tries to make a turn of more than 180 degrees and is thus turned back on itself. This behavior pattern tends to emphasize the rather mechanical nature of the school. Since the reaction of a mackerel to others of its kind is not to color it must be to form. However, form alone will not induce the schooling act if swimming movement is not normal. The fact that mackerel are evidently capable of detecting slight differences in movement would put the reaction of the fish on a slightly higher plane. Extensive hetero- typic grouping experiments are planned which would tend to shed much light on the factors of response to various types of body form and movement. Further interesting data might be obtained by observing the re- action by an isolated mackerel to its mirror image. Following the work of Spooner (1931), the reflecting surface of the mirror could be broken by lengths of tape at definite intervals ; in this way the reaction of the ani- mal to body form which is identical with its own but not complete might be ascertained. The condition of various types of fish schools and aggregations is summarized in a schematic diagram by Breder & Nigrelli (1935). Compared with other fishes possess- ing social tendencies, the schooling of mack- erel is striking by virtue of its fixity. Never- theless, the survival value of this mechani- cally highly integrated group is still not clear. An interesting feature of the mackerel school is the spacing of the individuals in the group. The distance between the ani- mals is more or less constant. A school may be greatly concentrated, however, by a sud- den disturbance which produces a “fright” reaction — after being momentarily dis- persed the fishes rush together in a compact mass which soon, however, returns to nor- mal proportions. Parr (1927) states that when the fishes in a closely schooling group approach each other too closely, their images may become too large and the accompanying strenuous accommodation of the eyes may produce a negative response, thus regulat- ing the spacing. Breder (1929) states the proposition that such fishes that depend on visual reactions for the formation of schools approach no closer to other objects than that distance at which they become clearly visible. This problem may be approached experimentally. 80 Zoologica : New York Zoological Society [XXVII: 14 Whether the schooling habit is ontoge- netically or phylogenetically acquired and whether the schooling or solitary state is the primitive one are important theoretical considerations. It is to be hoped that future investigation will shed some light on these problems. Summary. 1. Although larger groups are better in- tegrated, two individuals suffice to begin the formation of a mackerel school. 2. Blinding, darkness, and visual contact experiments indicate that the schooling re- action of the mackerel is visually integrated. 3. Mackerel display no schooling reaction to others of the same species, living or dead, which move or are moved in a manner not completely normal. 4. As far as tested, response to body color plays no role in the schooling reaction. 5. Isolation for three weeks does not eliminate or reduce the schooling proclivity. Literature Cited. Allee, W. C. 1931. Animal aggregations. Chicago: Uni- versity of Chicago Press. Bowen, E. S. 1931. The role of the sense organs in aggre- gations of Ameiurus melas. Ecologi- cal Monographs, 1: 3-35. 1932. Further studies of the aggregating behavior of Ameiurus melas. Biol. Bull., 63: 258-270. Breder, C. M., Jr. 1929. Certain effects in the habits of school- ing fishes, as based on the observation of Jenkinsia. Amer. Mus. Novitates, (382): 1-5. 1936. The reproductive habits of the North American sunfishes, Centrarchidae. Zoologica, 21 : 1-48. Breder, C. M., Jr. & Nigrelli, R. F. 1935. The influence of temperature and other factors on the winter aggrega- tions of the sunfish, Lepomis auritus, with critical remarks on the social be- havior of fishes. Ecology, 16: 33-47. 1938. The significance of differential loco- motor activity as an index to the mass physiology of fishes. Zoologica, 23: 1-29. Brown, F. A., Jr. 1937. Responses of the large-mouth black bass to colors. Bull. III. Nat. Hist. Survey, 21 : 33-55. Escobar, R. A., Minahan, R. P., & Shaw, R. J. 1936. Motility factors in mass physiology: locomotor activity of fishes under con- ditions of isolation, homotypic group- ing, and heterotypic grouping. Physiol. Zool., 9: 66-78. Newman, E. 1876. Mr. Saville Kent’s Lectures at the So- ciety of Arts on “The Aquarium: Con- struction and Management.” Zoolo- gist, 34: 4853-4856. Noble, G. K. 1934. Sex recognition in the sunfish, Eupo- motis gibbosus (Linne). Copeia, (4): 151-155. Noble, G. K. & Curtis, B. 1939. The social behavior of the jewel fish, Hemichromis bimaculatus Gill. Bull. Amer. Mus. Nat. Hist., 76: 1-46. Parr, A. E. 1929. A contribution to the theoretical anal- ysis of the schooling behavior of fishes. Occasional Papers, Bingham Oceano- graphic Coll., (1): 1-32. Schuett, J. F. 1934. Studies in mass physiology: the activ- ity of goldfish under different condi- tions of aggregation. Ecology, 15: 258-262. Shlaifer, A. 1938. Studies in mass physiology: effect of numbers upon the oxygen consumption and locomotor activity of Carassius auratus. Physiol. Zool., 11: 408-424. 1939. An analysis of the effect of numbers upon the oxygen consumption of Ca- rassius auratus. Physiol. Zool., 12: 381-392. 1940. The locomotor activity of the goldfish, Carrassius auratus L., under various conditions of homotypic and hetero- typic grouping. Ecology, 21(4): 488- 500. Spooner, G. M. 1931. Some observations on schooling in fish. Jour. Marine Biol. Assoc. United Kingdom, 17: 421-448. Warner, L. H. 1931. The problem of color vision in fishes. Quart. Rev. Biol., 6: 329-348. White, G. M. 1919. Association and color discrimination in mudminnows and sticklebacks. Jour. Exper. Zool., 24: 443-498. White Hineline, G. M. 1927. Color vision in the mudminnow. Jour. Exper. Zool., 47 : 85-94. EXPLANATION OF THE PLATE. Plate I. Fig. 1. A school of chub mackerel, Pneumato- phorus grex, in an exhibition tank under fairly strong artificial light. Fig. 2. A flash-bulb photograph of the same school in total darkness — dispersed. Fig. 3. Mackerel milling in light in the deep red — 650 pp to 700 pp. (The photographs include the upper four-fifths of the vertical depth and the central four-fifths of the length of the tank). SHLAIFER. PLATE I. FIG. 1. FIG. 2. FIG. 3. THE SCHOOLING BEHAVIOR OF MACKEREL: A PRELIMINARY EXPERIMENTAL ANALYSIS. 1942] Ingram & Heming : Carnivorous Snail, Euglandina rosea 81 15. Food, Eggs and Young of the Carnivorous Snail Euglandina rosea (Ferussac). William Marcus Ingram Mills College, California & Walter Edward Heming Whittier College. (Plate I) . The information contained herein is based on the observation of a single Euglandina rosea (Ferussac) which was collected April 1, 1940, at Fort Myers, Florida, by Dr. Wil- liam J. Hamilton, Jr., of Cornell University. On April 11 the snail was placed in a ter- rarium filled with humus and soil, at Ithaca, New York. Standing water was provided to keep the air in the container moist. The individual successfully survived until De- cember 9, 1940, when accidental overheating caused its death. During the interval of 252 days observations concerning its food habits, eggs and young were carried out. Food of Euglandina. During this period of approximately 8 months the following New York land mol- lusks were placed in the terrarium with the carnivorous Euglandina: one Mesomphix inornatus (Say), two Mesomphix cupreus (Rafinesque) , six Anguispira alternata (Say), and four newly hatched and two adults of Triodopsis albolabris (Say). Of these species only the M. inornatus, and the six A. alternata were devoured; the remain- ing snails were left untouched. The Euglandina was observed during its process of feeding upon the M. inornatus. The predatory snail in approaching the M. inornatus lifted its head and anterior foot region in the air and moved them from side to side. This weaving back and forth con- tinued for 60 seconds; during this time the greatly elongate lips were moved from side to side. Finally the Euglandina touched the body of the food snail, whereupon the latter contracted into its shell. The Euglandina then employed its anterior foot region to turn the Mesomphix shell over on its spire. This done, the predator entered the aperture of the food snail’s shell ; the posterior foot region serving to hold the rest of the body firmly to the substratum. As soon as the Euglandina thrust its head into the Mesom- phix shell aperture it began to feed with a piston-like motion, apparently forcing its head firmly against the soft parts of the prey so that the radula could obtain a firm purchase. The piston motion was accom- panied by a lateral movement, indicating that the Euglandina was working from the columellar region outward, and then back again. Because of the thinness of the M. inornatus shell, the feeding could be observed without great difficulty. As the body of Mesomphix was gradually consumed the Euglandina worked further into the prey’s shell, until its shell presented an effective block against further entrance. During the preliminary feeding the tentacle-like lips seemed to be thrust between the body of Mesomphix and its shell. Finally the colu- mellar muscle of the Mesomphix was torn loose and the entire upper visceral mass was observed to disappear into the buccal cavity of the Euglandina. The feeding process took 40 minutes from the time of entrance until all of the body of Mesomphix had disappeared. After the soft parts had been consumed the lips of the Euglandina were observed to move about the interior of the shell as though the animal was searching for food fragments which might have been over- looked. It seems likely that the sense of smell is well-developed in Euglandina and possibly serves this animal in locating food. Simpson’s data (1901) indicates that the sense of smell is highly developed and used to locate food by Triodopsis albolabris (Say). He records this species moving 18 82 Zoologica : New York Zoological Society [XXVII: 15 inches in order to feed on concealed lettuce. When the Mesomphix was placed in the terrarium with the carnivore, the latter was 10 inches removed and was fully contracted within its shell. In 3 minutes the Euglan- dina was active, and was moving- toward the Mesomphix. A similar phenomenon was ob- served when one A. alternata was placed in the terrarium; the Euglandina was again contracted and motionless, 6 inches away from the Anguispira. In five minutes it became active and moved toward the food snail. The six A. alternata and the one M. inor- natus utilized as food were all turned over on their apices before they were devoured. The following writers have discussed the snail food of Euglandina rosea. Baker ( 1903) reports Euglandina feeding upon the large Floridan pulmonate mollusks of the genera Liguus and Orthalicus. He observed that in some instances the voracious Euglan- dina will even bore a hole in the shell of the victim in order to reach the animal, instead of entering through the prey’s aperture. Rogers (1908) states that Euglandina de- vours individuals of its own species and preys chiefly upon Helices. A. Binney (1851) reports Euglandina feeding on half-putrid remains of a Helix, and on Limacies which were confined together in the same con- tainer. He, too, writes that it preys on its own kind. W. G. Binney (1885) concerning the feeding habits of E. rosea states, “By its [the radula’s] action the soft parts of its prey are rapidly rasped away or are forced in large morsels down the oesopha- gus. The animal has been seen to swallow entire the half-putrid remains of a Helix, and to attack Limaces confined in the same box with it, rasping off large portions of the integument, and in some instances de- stroying them. In one instance an individual attacked and devoured one of its own spe- cies, thrusting its long neck into the interior of the shell and removing all the viscera. I found many specimens of Polygra volvoxis, [ Polygra septemvolva Say], in the stomach of individuals collected by me at Saint Augustine, Fla.” Eggs of Euglandina. In the second week of October the Euglan- dina deposited 22 eggs on the upper surface of the humus in the terrarium. This location was exposed to light, although more secre- tive places were available. Five of the eggs did not hatch. Because of periodic absence from the laboratory complete data on the incubation period were not obtained. The approximate incubation period was 60 to 68 days; the hatching period for individuals continued for 8 days after the first indi- vidual had emerged, 60 days after the egg had been laid. The oval-oblong eggs were of nearly uni- form size, measuring 4.25 mm. in length and from 3 to 3.25 mm. in width (Fig. 1). The egg shell is brittle and hard; no inner egg membrane is present. The egg shell is extremely rough, and is quite porous. Young of Euglandina. The young develop with their long axes coinciding with the long axis of the egg. The young break through the egg shell by means of the radula. A circular opening (Fig. 1) is first filed in the egg shell at the end surrounding the aperture of the young within. The initial hole is enlarged by use of the radula until approximately one- third of the egg shell is cut away (Fig. 2). This done, the young emerges, leaving an intact two-thirds of the shell behind. The exact amount of time required for the young to cut away enough shell to permit their escape was not observed. It is estimated, however, that with the young in question it was between 6 and 10 hours. Little variation in size is shown in newly hatched young; of 15 that were measured the length varied .25 mm. ; with the greatest length 4 mm., and the least 3.75 mm. Varia- tion in width was also .25 mm. ; the widest shell was 3 mm.; and the narrowest 2.75 mm. The largest young was 3.90 mm. in length and 2.90 mm. in width; the smallest was 3.75 mm. long and 2.75 mm. wide. Each young has approximately two and one-fourth shell whorls upon hatching, as compared to the 6 to 8 whorls of fully ma- ture shells. The aperture of the young is relatively large in proportion to the length of the shell, running approximately five- sevenths of the total shell length (Fig. 3). In the parent of these young the aperture was approximately equal to one-half of the total shell length. W. G. Binney (1885) con- cerning the length of young shells that he studied states, “In young individuals the spire forms but a small proportion of the shell, but in the old it often forms one-third of the length.” A Binney ( 1851 ) comments on the shortness of the spire of young indi- viduals, and at its enormous increase in mature shells. In an attempt to supply food for the young snails 4 small immature T. albolabris were placed within reach. Adult individuals of A. alternata were also avail- able but individuals of neither species were touched. Apparently the lack of suitable ani- mal food resulted in the death of the young 15 days after hatching. Natural Habitat of Euglandina. It should be understood that the observa- tions included here were made in a habitat far from that in which Euglandina rosea is found. The data are, however, the most ex- tensive to be reported on the eggs and young of Euglandina, and the photographic method 1942] Ingram & Heming : Carnivorous Snail, Euglandina rosea 83 has not been used before to illustrate the eggs and young of this species. The natural habitat is reported by W. G. Binney (1885) as follows, “The habits of this animal are somewhat aquatic. It is found on the sea islands of Georgia and around the keys and everglades of Florida, and in these situations the shell often at- tains the length of 4 inches ; when found on the oyster hummocks and less humid locali- ties it seldom exceeds 1 inch in length. Mr. Say found it in the marshes immediately behind the sand-hills of the coast. It is most readily found in the center of the clumps of coarse grass on these marshes.” W. G. Binney (1885) lists the following distribu- tion: Atlantic and Gulf States from North Carolina to Texas; Macon Springs, Georgia; Bibb County, Alabama; and Jackson, Mis- sissippi. A. Binney (1851) indicates that it is common among the West Indian Islands. Pratt (1935) states that its distribution is from South Carolina to Texas. Acknowledgments. The writers are indebted to Dr. William J. Hamilton, Jr., for supplying the Euglan- dina. Credit is due Mr. Arthur Smith for taking the photographs of the eggs and the young. Bibliography. Baker, F. C. 1903. Shells of land and water, p. 51, Chi- cago, A. W. Mumford, Publishers. Binney, A. 1851. The terrestrial air-breathing mollusks of the United States, etc., pp. 301-303, Boston, Charles C. Little and James Brown. Binney, W. G. 1885. A manual of American land shells, Bull. U. S. Nat. Mus., no. 28, pp. 348- 350. Pratt, H. S. 1935. A manual of the common invertebrate animals, etc., p. 601, P. Blakiston’s Son and Co., Philadelphia. Rogers J E 1908. The shell book, p. 252, N. Y., Double- day, Page and Co. Simpson, G. B. 1901. Anatomy and physiology of Polygra albolabria and Limax maximus, etc., Bull. N. Y. State Mus., vol. 8, no. 40, pp. 273-276. 84 Zoologica: New York Zoological Society [XXVII: 15 EXPLANATION OF THE PLATE. Fig. 1. Egg of Euglandina rosea, showing ini- tial opening cut by radula of young within its eggshell. Note porous and rough character of the shell. X 21. Fig. 2. Egg of young three hours later, show- ing egg shell cut away, revealing the immature snail within. X 23. Fig. 3. Newly hatched young which emerged approximately 6 hours after the initial opening had been made in the egg shell. X 23. INGRAM & HEMING. PLATE I. FOOD, EGGS AND YOUNG OF THE CARNIVOROUS SNAIL, EUGLANDINA ROSEA (fERUSSAc). 1942] Potts: Hypophysis of Ovo-viviparous Poeciliids 85 16. The Anatomy and Morphology of the Hypophysis of Several Species of Ovo-viviparous Poeciliids.* Hugh E. Potts Department of Biology, W 'ashing ton Square College, New York University. ( Plates I & II) . Introduction. The constantly growing knowledge of the endocrine glands makes increasingly appar- ent the complex physiological interrelation- ships of the pituitary. A great deal has been reported both on the morphology and the physiology of the pituitary but most of the investigations have been concerned with mammals where such structures as the mammary glands, corpora lutea and pla- centa introduce many complicating factors. It would seem possible to obtain some clari- fication of the many problems arising from these studies by an investigation of the pituitary in forms such as lower vertebrates in which the presence of fewer accessory structures makes the problem less complex. A survey of the literature has shown a surprisingly incomplete knowledge of the pituitary of lower vertebrates, especially of teleost fishes (Stendell, 1914; deBeer, 1926; Charipper, 1937). This becomes especially apparent when one considers the various reports relating cell types in the anterior pituitary with the reproductive cycle in mammals (Rasmussen, 1921; Wolfe & Cleve- land, 1933, 1933; Charipper & Haterius, 1932; Kirkman, 1937) as compared with the meager literature of a similar nature con- cerning the bony fishes. With regard to the latter group, Matthews (1936) has shown that the pituitary gland of Fundulus is not static but undergoes variations in cell type distribution which may be related to the different seasons of the year. Cardoso (1934), also working on an oviparous form, Pimelodus clarias, has shown a relation between gonad size and injection of hypo- physeal suspensions. Houssay (1931), inves- tigating the ovo-viviparous teleost, Cnes- terodon decemmaculatus, described a rela- tionship between ovulation and secretions of the pituitary. In addition, Rojas et al. * Accepted in partial fulfillment of the requirements for the degree of Doctor of Philosophy, New York University. (1934) showed cyclic changes in the pi- tuitary of Jenynsia lineata, another ovo- viviparous teleost. Further, the Teleostei as a group seem to offer many opportuni- ties for descriptive investigations correlat- ing variable hypophyseal histology and cytology to reproductive phenomena. In this group are a vast number of diminutive tropical fishes which present many evolu- tionary phases of reproduction from ovi- parity to true viviparity. This group of viviparous fishes are of particular interest because of the relatively short and usually regular reproductive cycle. These animals may be kept in large numbers in small aqua- ria, assuring a plentiful supply throughout the year. In addition, most of them can be bred readily so that animals in definite stages of the reproductive cycle may be easily obtained at all times. With these facts in mind an investigation of the morphology and histology of the pituitaries of males and females of six spe- cies of ovo-viviparous teleosts was under- taken and forms the basis of the present report. The particular species examined were carefully chosen for availability and the ease with which they could be main- tained and bred under laboratory conditions. In addition all these forms showed regular reproductive cycles of approximately thirty days. It is proposed that these descriptions constitute the beginning of what should prove to be an exhaustive survey and furnish a basis for future experimental investigations. I wish to take this opportunity to ex- press my deep appreciation to Dr. Harry A. Charipper for his generous and helpful assistance throughout the course of this investigation. Materials and Methods. The pituitaries of six different species of ovo-viviparous poeciliids, Platypoecilus variatus, Xiphophorus heller i (red, and 86 Zoologica : New York Zoological Society [XXVII: 16 green varieties), Lebistes reticulatus, Limia tricolor, Mollienisia sphenops and Mollien- isia latipinna (black variety) were exam- ined histologically. Specimens of Mollien- isia sphenops and latipinna were obtained from a local dealer during the months of December, January and February. These were killed and fixed within two weeks of purchase. The other species were reared in aquaria at temperatures between 21 and 27 degrees centigrade and sacrificed at various times throughout the year. Each fish was anaesthetized in ice water and the head severed immediately posterior to the gills and also cut just in front of the eyes. A fixing fluid composed of 18.5 cc. basic Zenker’s to 1.5 cc. formalin gave the best results in about twenty hours. Follow- ing fixation, tissues were decalcified in phloroglucin 1% hours and washed in run- ning water 6 to 12 hours before dehydra- tion and embedding. Whole heads, or brain and pituitary, dis- sected free of all cartilage were then dehy- drated and embedded in hard paraffin or tissue mat, after which they were cut at 5 micra in median sagittal, cross and hori- zontal sections. The Masson technic was used almost ex- clusively. Timing and washing were stand- ardized as carefully as possible in order to get comparable results on all six species. Earlier preparations were stained in Masson A ten minutes, rinsed in distilled water and destained in 1% phosphomolybdic acid 1% hours. They were then transferred directly to Masson C for thirty minutes, dipped in 95% alcohol, absolute alcohol, half and half absolute alcohol and xylol and then pure xylol, and finally mounted in Canada bal- sam. In later preparations the staining times were changed to three minutes in Masson A, thirty minutes in phosphomolyb- dic acid and 1 hour in Masson C. The latter timing gave better results although the staining affinities did not seem to be criti- cally affected. Description. The pituitaries of the six species of poe- ciliids in this investigation show such strik- ing similarities in morphology and histology that a single description will be made, using Platypoecilus variatus as the standard. The description will deal first with morphology and then with histology, under the headings cerebral portion (pars nervosa) and epi- thelial portion (pars anterior, intermedia, and uebergangsteil). Similarities and dif- ferences of the other species will then be considered separately. I. Platypoecilus variatus A. Morphology. The pituitary is an ovoid gland, immedi- ately behind the optic chiasma, attached to the midventral floor of the diencephalon by a short hollow stalk. There is no sella turcica but the pituitary protrudes into a depres- sion, the hypophyseal fenestra, closed ven- trally by connective tissue (PI. II, Fig. 12). An extremely thin meninx primitiva ex- tends from the brain down over the stalk and over the pituitary, carrying with it numerous small blood vessels. There are four portions of the gland, dis- tinguishable by staining reactions and cell types, namely, pars nervosa, anterior, inter- media and uebergangsteil (PI. II, Fig. 7). No distinct septa separate the portions. The pars nervosa is a thickened and modified portion of the floor of the diencephalon, occupying the central dorsal region of the gland. The pars anterior, like a half sphere, has its posterior surface in contact with all the other parts of the gland. In its median dorsal portion, this contact is with the ner- vosa. Surrounding the nervosa completely ventrally and laterally, but only its ventral half anteriorly and posteriorly, is a convo- luted layer of cells, the uebergansteil, which contacts the posterior face of the pars an- terior ventral and lateral to the nervosa. The pars intermedia is a shell-like portion surrounding the uebergangsteil completely ventrally, laterally and posteriorly, forming in the latter region the blunt posterior end of the gland. Anteriorly it is contiguous with the pars anterior lateral and ventral to the contact of the latter with the ueber- gangsteil. B. Histology 1. Cerebral Portion ( Pars Nervosa) The cerebral portion is composed of tissue from the floor of the diencephalon. It con- sists of masses of neuroglia cells and many interlacing fibres continuous through the infundibular stalk with the brain ( PI. I, Figs. 2, 4). The outer walls of the stalk are covered by a thin meninx primitiva whereas the inner walls are composed of ependyma cells which line the infundibular cavity. The nuclei of ependyma and neuroglia cells are essentially similar, being rounded to ovoid and varying considerably in size. Each is surrounded by a well defined mem- brane and contains a centrally located, red- staining nucleolus and scattered pink to violet chromatin material. Nuclei are more abundant near the stalk and sparser in the central and more distal parts of the nervosa. The cytoplasm of the ependyma cells is pale gray. Many irregular processes from them extend into the infundibular cavity. The cytoplasm of the neuroglia cells is in- distinct and loosely fibrous, also staining a pale gray. Many tracts of fibres extend from the central portion of the nervosa into other parts of the gland as compact processes en- closing blood vessels which are thus distrib- uted to the epithelial portions of the gland. 1942] Potts: Hypophysis of Ovo-viviparous Poeciliids 87 Penetration by these processes is especially heavy in the uebergangsteil and intermedia (PI. I, Fig. 6; PI. II, Fig. 7). In the dorsal posterior region of the nervosa the fibres run caudad through the low posterior end of the basket-shaped uebergangsteil and ramify among the cells of the pars inter- media ( PI. II, Fig. 7) . Colloid bodies of varying size and shape are characteristically present in the nervosa in varying amounts (PI. I, Fig. 6; PI. II, Figs. 7, 11). In the central region they are usually fairly large bodies but in the pos- terior region they are generally fine and granular. Their staining reaction varies in different glands from violet to pink, and both colors may be present in the same gland. In some cases large basophilic masses are filled with brilliant red-staining globules. Free single cells, generally basophiles, are often found well within the nervosa. In some cases they appear normal but in others they show pycnotic nuclei and indistinctly outlined cytoplasm. The latter closely re- semble many of the larger colloid masses. 2. Epithelial Portion a. Pars Anterior The pars anterior is composed almost entirely of orange-red acidophiles, except for a tongue-like portion adjacent to the nervosa in the median dorsal portion of the gland. The acidophiles are closely packed, showing no definite arrangement into cords or tubules (PI. II, Fig. 7). They vary in shape from rounded to ovoid to spindle shaped. Their size is also variable. The cyto- plasm contains fine orange-red granules. The nuclei vary in size in relation to the size of the cells containing them. Their shapes range from spherical to elongate and their position may be central or apical. Within each nucleus are one or two red- staining spherical nucleoli surrounded by red-staining chromatin granules in a clear hyaloplasm. The tongue-like strip of tissue (PI. I, Fig. 2) is a sheet several layers thick, roughly “V” shaped, which fits into the uebergangsteil with which it is continuous laterally and ventrally. Its posterior surface is in contact with the pars nervosa from which it is clearly separated by a membrane, and its anterior surface is in contact with the acidophilic part of the pars anterior. Here the delineation is less definite as many of the acidophiles penetrate the tongue-like portion irregularly. Many blood vessels pass through this sheet, enclosed within proc- esses from the nervosa. The cells of this layer bordering the nervosa tend to be columnar while those within the sheet are poorly defined. The cytoplasm has a fleecy appearance and stains a pinkish-gray color. Nuclei are ovoid to spherical, containing a centrally located nucleolus and scattered granular chromatin, both of which stain similarly to the cytoplasm. b. Pars Intermedia The pars intermedia is composed entirely of basophiles. Its cytology is variable, the posterior part often showing a large number of basophiles, larger and more clearly defined than the cells of its middle portion, which closely resemble basophiles of the ueber- gangsteil. The cytoplasm of these larger ovoid to spherical cells ranges from violet to blue and appears homogeneous. Nuclei are roughly spherical, often having inden- tations which give them a vesicular appear- ance. They contain granular, faint pink- staining chromatin and a centrally located red-staining nucleolus. The rest of the in- termedia cells are smaller and usually indis- tinctly outlined, their stainability varying from pale gray-blue to a deep blue. Small spherical vacuoles are often present. No special arrangement is apparent in these cells (PI. I, Fig. 6; PI. II, Figs. 7, 8). The nuclei are small and ovoid to spherical with well defined membranes in which is a granu- lar chromophobic chromatin surrounding a centrally located red nucleolus. Nervosa processes enclosing blood vessels penetrate the middle region of the pars intermedia after passing through the uebergangsteil, while loose wavy masses of fibers penetrate the dorsal posterior portion directly ( PI. II, Fig. 7). Besides varying in cell type, the pars intermedia varies in size. In some it is large while in others it is much reduced, forming a thin layer over the uebergangsteil and a small posterior part. C. Uebergangsteil The uebergangsteil is the most changeable part of the gland, varying both as to size and as to proportions of its three cell types, basophiles, acidophiles and chromophobes (PI. I, Fig. 6; PI. II, Figs. 7, 8). Sometimes single cells or islets of cells lie free among the fibres of the pars nervosa. The most common type of cell is a large polygonal or round deeply staining carmine red cell with well defined borders. The cyto- plasm is filled with coarse carmine-colored granules. The nuclei are spherical to ovoid with a fine granular light red chromatin reticulum surrounding a centrally located brilliant red nucleolus. In some pituitaries these cells were present almost exclusively (PL I, Fig. 6). Other glands have an ueber- gangsteil which is heavily basophilic with a few scattered acidophiles and chromo- phobes. In such glands the stainable ma- terial of the acidophiles is often clumped and peripheral in location. The basophiles closely resemble the acidophiles in size and shape and their nuclei appear identical ( PI. II, Fig. 8). Their cytoplasm is optically homogeneous and varies in staining capacity 88 Zoologica : New York Zoological Society [XXVII: 16 from pale gray-blue to deep blue. In some glands basophiles are present only in one or two spherical, centrally located masses which may be exclusively basophilic or may contain some acidophiles also. Chromo- phobes are smaller, poorly outlined and faintly pink-staining cells, more prevalent, if present at all, in the ventral region of the uebergangsteil. Their nuclei are similar to those of the chromophiles. II. Xipliophorus helleri. The pituitary of Xipliophorus is typically deeper dorso-ventrally than that of Platy- poecilus variatus and is more concave at its dorsal surface (PI. I, Fig. 4). Cell types of the four portions of the gland correspond to those described above with the exception that several small patches of basophiles are usually present in the pars anterior (PI. I, Fig. 4). These have a small amount of fleecy, indistinctly outlined bluish-gray cytoplasm and nuclei with a coarsely granular, similarly colored chro- matin reticulum. They resemble quite closely the cells of the tongue-like portion of the pars anterior. Another difference from Platypoecilus is found in the arrangement of the cells of the uebergangsteil. In Xipliophorus cells of this region form a deep mass heavily penetrated by small curving fibrous tracts enclosing blood vessels, which, in median sagittal sec- tions, appear to divide the tissue into rounded islands (PI. I, Fig. 4). In Platy- poecilus the uebergangsteil is a thinner, more convoluted layer through which the nervosa processes pass in a more direct manner. III. Lebistes reticulatus The pituitary of Lebistes reticulatus is slightly more elongate and dorso-ventrally flattened than either of the preceding (PI. I, Fig. 5; PI. II, Fig. 10). Fewer nervosa processes penetrate the tongue-like portion of the pars anterior than is the case in Platypoecilus variatus. The uebergangsteil is more of a convoluted layer than in Platy- poecilus (PI. II, Fig. 10), and is heavily penetrated by nervosa processes with their blood vessels. One or two conspicuous masses of basophiles are characteristically present in the central region of the uebergangsteil. In the pars intermedia of Lebistes a broad fibrous tract often reaches from the pos- terior end of the nervosa, postero-ventrally to the median ventral border of the gland. This is not shown in the figures of Lebistes but may be seen in PI. I, Fig. 3, of Mollien- isia splienops, and PI. II, Fig. 9, of Limia tricolor. Cell types and arrangement are almost identical with those of Platypoecilus. How- ever, patches of pale basophiles, similar to those of Xipliophorus, are occasionally pres- ent in the pars anterior, especially in the Trinidad variety of Lebistes. IV. Limia tricolor The pituitary of Limia tricolor is similar in shape to that of Xipliophorus (PI. II, Fig. 9). The pars anterior is heavily pene- trated by nervosa processes as in the case of Platypoecilus. A broad tract of nervosa fibres cut through the uebergangsteil and pars intermedia to the median ventral sur- face of the gland as in the case of Lebistes. Cell types correspond to those in Xipho- phorus, sparse groups of basophiles being characteristically present in the pars anterior. V. Mollienisia splienops The pituitary of Mollienisia splienops is different in shape from all the others, being almost perfectly rounded, and flattened dorso-ventrally, (PI. I, Fig. 3; PI. II, Fig. 12). Like Lebistes and Limia it has a broad nervosa tract extending postero-ventrally through the uebergangsteil and pars inter- media to the median ventral surface of the gland. Occasional pale basophiles are found scattered in the pars anterior as in the case of Xipliophorus, and are also present along the nervosa processes entering the pars anterior. The uebergangsteil is more highly convo- luted than in Platypoecilus and Lebistes, and often a linear arrangement of carmine- colored acidophiles forms a border between the nervosa and basophiles and acidophiles of the deeper layer of the uebergangsteil. VI. Mollienisia latipinna The pituitary of Mollienisia latipinna is most similar in shape to that of Xiphopho- rus helleri (PI. I, Fig. 1). The extension of the infundibular cavity, the recessus hypo- physeus, penerates the gland more deeply than in any of the other species. Many ner- vosa processes with their blood vessels pene- trate the tongue-like portion of the pars anterior to be distributed throughout the basophilic part of the pars anterior. Along these processes, and scattered in small groups throughout the mass of acidophiles, are pale basophiles, a condition similar to that found in Mollienisia splienops. Discussion. I. Anatomy Pituitary morphology and histology are fundamentally similar in the six species of poeciliids examined. The glands most closely resemble those of the egg laying poeciliid, Fundulus, and the ovo-viviparous poeciliid, Jenynsia lineata, as described by Scruggs (1939) and Rojas et al. (1934) respectively. The infundibular stalk is highly variable among teleosts, being almost non-existant in Mormyrus (Stendell, 1914), the pituitary Potts: Hypophysis of Ovo-viviparous Poeciliids 89 1942] being held close to the brain, and short in Fundulus (Matthews, 1936) and Carassius auratus (Bell, 1937, 1938; Levenstein, 1939; Scruggs, 1939). In Lophius piscatorius (de Beer, 1926) the stalk reaches its highest development, extending anteriorly for a considerable distance. It may be solid as in Carassius or hollow as in Pungitus (Scruggs, 1939) and in the poeciliids de- scribed here. No structure comparable to a sella turcica is present in the teleosts. In Fundulus (Matthews, 1936) and Carassius (Bell, 1938; Levenstein, 1939) the pituitary is protected by the parasphenoid bone. The poeciliid crania of this investigation differ from the above, being of the platybasic type as described by Kingsley (1936). The cra- nium floor is composed of calcified connec- tive tissue which spreads out laterally over the trabeculae cranii. The pituitary projects ventrally into the hypophyseal fenestra, a depression in the floor of the cranium. A. Pars Nervosa The pars nervosa protrudes into the gland at different angles in various teleosts. In Carassius (Bell, 1938; Levenstein, 1939) and in Cyprinus (Scruggs, 1939) the gland is tilted forward and the nervosa runs an- teriorly. In Ameiurus nebulosus (Scruggs, 1939) the pituitary is tilted backward and the nervosa runs posteriorly. In the poe- ciliids investigated here the pituitary is directly beneath the infundibular stalk and the nervosa is directed vertically downward. A recessus hypophyseus, representing an extension of the infundibular cavity into the nervosa, is present in Jenynsia (Rojas et al., 1934), Fundulus (Matthews, 1936), and in the poeciliids of this investigation. It is lacking in some teleosts, as Carassius (Bell, 1938; Levenstein, 1939) and Amei- urus (Scruggs, 1939). There is also much variation in the num- ber of nervosa processes given off and in their penetration of the various portions of the gland. In Pseudopleuronectes america- nus and Cyprinus carpio (Scruggs, 1939) the nervosa processes are limited exclusively to the pars intermedia; in Fundulus, to pars intermedia and uebergangsteil ; in Caras- sius, Jenynsia and the poeciliids of this report, the nervosa processes go to all parts of the pituitary. B. Pars Anterior The proportions, contacts and positions occupied by the various portions of the pituitary in different teleosts are highly variable. In Carassius (Bell, 1938; Leven- stein, 1939; Scruggs, 1939) and in Cyprinus carpio (Stendell, 1914; Scruggs, 1939) the pars anterior is small, dorsal in position and makes contact with the uebergangsteil mostly and the nervosa only in a limited area. In Notemigonus (Scruggs, 1939) it is large and anterior in position, making con- tact with the uebergangsteil and the main portion of the nervosa. In Ameiurus nebu- losus the pars anterior is small. It is ante- rior and ventral in position, again making contact with uebergangsteil and nervosa (Scruggs, 1939). In the fishes of the inves- tigation the pars anterior is very large, anterior in position, and touches nervosa, uebergangsteil and pars intermedia. The variations, however, are more apparent than real, being due in large part to a tilting of the pituitary either backward or forward. A pars anterior has been reported for all teleosts examined except Esox niger (Scruggs, 1939) and Fundulus (Matthews, 1936). However, in his 1937 paper Matthews decided that the anterior portion of the pituitary, previously called the uebergang- steil, was actually the pars anterior. This was substantiated by Scruggs (1939) on the basis of staining reactions. In the case of Esox niger Scruggs found no portion taking a stain like that of the pars anterior of other teleosts; Stendell (1914), however, describes a small pars anterior for Esox lucius. C. Pars Intermedia The pai’s intermedia of teleosts has usually been identified by its close relation- ship with the pars nervosa. Stendell (1914) shows an intimate relationship of the two parts in the primitive Mormyrus, while in the higher teleost, Esox lucius, the pars intermedia is more posterior in position and retains its association with the nervosa by means of nervosa processes extending out into it. It extends forward ventrally, reach- ing or nearly reaching the posterior end of the pars anterior, thus surrounding the uebergangsteil ventrally. While Matthews (1937) divides the pituitary of Fundulus into only two epithelial portions, a pars anterior and a pars intermedia, Scruggs (1939) shows that the latter portion may be differentiated, by the Dawson & Fried- good (1938) method, into two portions com- parable to the pars intermedia and the uebergangsteil of other teleosts. The pitui- tary of the poeciliids of this investigation compare with that of Fundulus as found by Scruggs, having an uebergangsteil adjacent to the pars nervosa and a pars intermedia which surrounds it laterally, ventrally and posteriorly. Connection with the nervosa is direct posteriorly and by means of nervosa processes extending through the uebergang- steil centrally and anteriorly. D. Uebergangsteil The uebergangsteil is present in some cyclostomes and all teleosts. It is highly variable in the latter, the simplest condi- tion being found in Mormyrus where the 90 Zoologica : New York Zoological Society [XXVII: 16 pars anterior, uebergangsteil and pars in- termedia are arranged in linear order with no distinct boundaries between them. In Gasterosteus (Bock, 1928) connective tissue septa separate the parts. In Carassius au- ratus (Bell, 1938) the uebergangsteil is very large, bordered by connective tissue, and forms most of the anterior part of the gland. In Esox niger (Scruggs, 1939) no pars anterior is reported and the uebergang- steil occupies the antero-dorsal region of the gland, being in size, shape and location similar to the pars anterior of Carassius auratus. The uebergangsteil of the poeciliids investigated here has no connective tissue septa, is moderately large, and as in most teleosts, lies between the pars anterior and the pars intermedia, a portion of the latter extending over it, however, in the middle region of the gland. It is highly variable in individuals of the same species, being a large deep layer in some, a thinner layer in others. In Lebistes, Platypoecilus and Mol- lienisia sphenops it tends to be highly con- voluted while in the other species examined it is a thicker, less folded layer. II. Histology A. Pars Nervosa Stendell (1914) described the nervosa most completely. It is composed chiefly of neuroglia cells. The infundibular cavity and recessus hypophyseus are lined with primi- tive ependyma cells having protoplasmic extensions both basally and distally. Bock (1928) confirms these findings in Gasteros- teus. Through the neuroglia network Stendell describes lymph tracts, blood vessels and connective tissue. Stolon-like nervosa proc- esses extend to all parts of the gland. He interprets the structure of the nervosa as providing the means of absorption of pars intermedia secretions by way of lymph and blood vessels. Colloid masses among nervosa fibres, he believes represent secretion of degenerating pars intermedia cells. Collin (1924), in mammals, Florentin & Weiss (1931), Florentin (1934) and Rojas et al. (1934), in teleosts, hold the same theory of secretion and absorption. The structure of the pars nervosa of the poeciliids investi- gated here confirms that of the above au- thors. Nervosa processes ramify through- out all portions of the pituitary, especially the pars intermedia. Colloid masses of vary- ing size and amount are present in the ner- vosa and pars intermedia. Matthews (1936) , Levenstein (1939), Scruggs (1939) and others call attention to masses of colloid in the nervosa. Stendell (1914) also notes many free cells of the pars intermedia ly- ing in the nervosa and believes that they degenerate into colloid. In the present work a number of free cells were seen lying in the nervosa, some appearing degenerate and closely resembling colloid masses. B. Pars Anterior The pars anterior of teleosts differs greatly from that of other vertebrates, con- sisting almost entirely of acidophiles, with sometimes a scattering of basophiles. En- tirely basophilic pars anteriors have been reported in the eel, Cyprinus, Esox lucius and in Carassius by Tilney (1911), Stendell (1914) and Bell (1938). However, modern technics have shown these same portions to be predominantly acidophilic (Florentin & Weiss, 1931; Scruggs, 1939; Levenstein, 1939). These discrepancies thus appear to be due to differences in staining technics. In the present work the pars anterior is almost entirely acidophilic. Occasional sin- gle or small groups of pale basophiles may be present, especially in Xiphophorus, Limia and the Mollienisias. Their scarcity pre- cludes any great physiological significance. An additional type of cell, an orange acido- phile, is described in some species by Scruggs, using the Dawson & Friedgood technic. These are difficult to interpret since in some species they are the predominant or exclusive type of cell present in the pars anterior while in others a few are present among a background of carmine-staining cells. Also similar orange cells are described in the pars intermedia of some species. The arrangement of pars anterior cells differs greatly in the various groups of tele- osts, varying from a compact condition with no special arrangement as found in the poeciliids in this report to a tubular distri- bution in which the cells are arranged around lumina, as in the eel (Tilney, 1911) and in the Salmonidae (Scruggs, 1939). An intermediate condition, where the ceils are arranged in solid cords, is seen in the cod (Herring, 1908). In those forms in which cord or tubule arrangement is present the cords or tubules are separated from one another by connective tissue septa. Such separation is difficult to discern, or absent in the compact type of gland. A pale basophilic or chromophobic tongue- like portion of the pars anterior has re- ceived little attention; Bock (1928) reports it in Gasterosteus and Scruggs refers to it briefly in the Salmonidae, Fundulus and Pungitus. This portion is present in all the Poeciliidae investigated in the present re- port, occupying a position in the dorsal half of the gland, between the acidophilic part of the pars anterior and the pars nervosa. It is a layer several cells in thickness, the cells adjacent to the pai’s nervosa being ependyma-like, oriented with their long ax°s perpendicular to their contact with the nervosa. C. Pars Intermedia Most of the recent work on teleost pitui- taries shows the pars intermedia to be com- posed almost entirely of small pale-staining 1942] Potts: Hypophysis of O vo-viviparous Poeciliids 91 basophiles. Scruggs (1939), using the Dawson & Friedgood technic, also found a varying number of orange cells. In Esox, Scruggs describes basophiles, orange cells and carmine cells; in Fundulus, basophiles, chromophobes and orange cells. Matthews (1936) describes acidophiles of the pars intermedia lining the nervosa processes. These cells, however, seem to belong to the uebergangsteil as described by Scruggs in the same species, rather than to the pars intermedia. The poeciliids investigated in the present report closely resemble Fundu- lus in pituitary structure. In these the pars intermedia, with the Masson stain, is en- tirely basophilic. The uebergangsteil lies between part of the intermedia and the ner- vosa, and deep red cells of the uebergang- steil often line the nervosa processes. The arrangement of cells in the pars in- termedia is generally agreed to by all inves- tigators. In Anguilla and Esox ( Stendell, 1914) polygonal cells form a compact stra- tum traversed by many nervosa processes. Bordering the neiwosa processes the cells become cuboidal to columnar, forming an epithelium. Bell (1938) describes the cells of the pars intermedia of Carassius as ovoid or polygonal but forming an epithelial layer about the nervosa processes. Scruggs de- scribes the intermedia cells of Carassius as indistinctly outlined. In the present work the pars intermedia shows a compact ar- rangement with polygonal or indistinctly outlined cells. An epithelial arrangement may occasionally be found adjacent to ner- vosa processes. The pars intermedia varies both in struc- ture and in the proportion of the gland which it occupies. Rojas et al. describe the disappearance of typical cells in the pos- terior portion and the appearance of colloid droplets in this location. The same condition is noted in the pars intermedia of the poe- ciliids investigated here. The dorsal portion of the intermedia is sometimes filled with small colloid droplets and few typical cells are present. Herring (1908) was the first to note such colloid in his work on the cod, and since then Stendell (1914) and many others have called attention to its presence. Scruggs (1939) reports a decrease in the size of the pars intermedia of Fundulus in January as compared with June. The pars intermedia of the poeciliids investigated here shows considerable variation in the proportion of the gland which it forms. In some cases a portion of its territory is occu- pied by large basophiles which appear to belong to the uebergangsteil. D. Uebergangsteil The uebergangsteil is composed of acido- philes and basophiles in the eel and in Mormyrus, acidophiles predominating in the former and basophiles in the latter (Sten- dell, 1914). In the stickleback (Bock, 1928) chromophobes and a few basophiles are pres- ent. Scruggs (1939) finds no acidophiles in the uebergangsteil of the eel, only basophiles and chromophobes being present. In most of the teleosts examined by him, he reports deep-staining acidophiles and basophiles, and chromophobes. The poeciliids examined here have the three characteristic cell types. Uebergangsteil cells may form a compact mass with acidophiles and basophiles scat- tered in groups (Bell, 1937, 1938; Charip- per, 1937 ; Levenstein, 1939; Scruggs, 1939) or a folded epithelial configuration as seen in Scruggs’ figures of the C-entrarchidae and Poeciliidae. A similar condition is seen in the figures of Matthews (1936), Rojas et al. (1934) and in the poeciliids of this investigation. Cell type proportions are highly variable in the uebergangsteil of any single species. Bock (1928), Rojas et al. (1934), Matthews (1936) and Scruggs (1939) all call atten- tion to changes in cell types. Matthews finds seasonal changes in the proportionate num- ber of acidophiles and basophiles of Fundu- lus. A particular type of basophile is present in the posterior part of the pars intermedia only at certain times of the year. Scruggs also reports seasonal changes in the Fundu- lus pituitary and describes the migration of large deep-staining basophiles which invade the posterior part of the pars intermedia. In this investigation, large basophiles are present in many specimens in the posterior part of the pars intermedia, similar to those of the uebergangsteil. Also great variability is manifested in percentage of basophiles and acidophiles in different individuals, the uebergangsteil of some being almost entirely basophilic, others almost entirely acidophi- lic, still others having varying proportions of each. These variations are suggestive of a regulated cyclic variation which may be related to the reproductive cycle. However, cell counts failed to establish any definite relationship between stages of the repro- ductive and percentage of cell types. E. Colloid Herring (1908) first called attention to colloid in the cod pituitary. Stendell (1914 ) emphasized the functional relationship be- tween nervosa and pars intermedia and pos- tulates two methods of secretion by the intermedia, one by release of minute colloid droplets which are absorbed by the nervosa processes. In the second, intermedia cells wander into the nervosa and disintegrate, forming secretory material which is ab- sorbed by the nervosa. Rojas et al. (1934) and Florentin (1934) likewise describe holo- crine secretion in teleost pituitaries. Rojas states that cells of the posterior part of the pars intermedia degenerate and only colloid is left in their place. Matthews (1936) de- 92 Zoologica : New York Zoological Society [XXVII: 16 scribes acidophilic colloid in the nervosa which closely resembles the secretory inclu- sions of the acidophiles of the pars “inter- media” as described by him. In the poeciliids investigated here the posterior portion of the pars intermedia is highly colloidal in many specimens and few cells are present. In others the region is cellular and has little or no colloid. Also single cells or small groups of cells from the uebergangsteil are sometimes present in the nervosa, many of them apparently in the process of disinte- gration and closely resembling some of the colloid masses. Thus both types of secretion as interpreted by Stendell (1914) are con- firmed in this report. F. Blood Supply Bock (1928) and Bell (1938) describe poor vascularization in the pituitaries of Gasterosteus and Carassius respectively. On the other hand, a heavy vascular supply is described in the eel (Tilney, 1911) and Cyprinus and Esox (Stendell, 1914). The poeciliids investigated here most closely re- semble the eel in vascularity. Large vessels in the nervosa radiate outward in all direc- tions within the nervosa processes, and penetrate all parts of the gland. G. Homologies Establishment of homologies of the parts of the teleost pituitary with those of mam- mals is desirable as a prelude to experi- mental work. Stendell (1914), Charipper (1937), Bell (1938) and Levenstein (1939) homologize the teleost uebergangsteil with the mammalian pars anterior. Levenstein shows two types of chromophobes in the uebergangsteil having Golgi corresponding to those of the acidophiles and basophiles of this portion. This compares with the findings of Addison (1916), Atwell (1929), Severinghaus (1933) and Kirkman (1937) in mammals. The present work shows three types of cells in the teleost uebergangsteil similar to those in the mammalian pars an- terior. The pars intermedia of mammals and teleosts occupies a similar position and is almost exclusively basophilic. The pars anterior of teleosts is inter- preted by Bock as a portion unrelated to the pars anterior of higher vertebrates. Charipper (1937) and Bell (1938) compare it with the pars tuberalis of higher forms on the basis of location, cord-like arrange- ment of cells and basophilic staining reac- tion. Scruggs (1939), using the Dawson & Friedgood technic, finds the teleost pars anterior to be acidophilic, however, while the pars tuberalis of mammals is chromo- phobic with the same technic. In the pres- ent investigation a tongue-like portion of the pars anterior, occupying a position along the dorso-anterior surface of the nervosa, close to the infundibular stalk and brain, takes a pale basophilic or chromophobic stain. Bock (1928) and Scruggs (1939) describe a similar portion in several teleosts. This portion seems to compare more favor- ably with the pars tuberalis of higher forms. If this be the case, then the rest of the pars anterior can be considered only as a sepa- rate structure characteristic of teleosts and of some cyclostomes. Summary and Conclusions 1. The pituitaries of the six species of ovo- viviparous poeciliids conform, in the presence and general disposition of epi- thelial and nervous portions, with the structure reported for other teleost fishes. The four portions are the pars anterior, pars intermedia, uebergang- steil and pars nervosa. 2. Boundaries between the parts are estab- lished by abrupt changes in cell type rather than by connective tissue septa. 3. The pars nervosa consists of a solid mass of fibrous tissue and neuroglia cells. It occupies the dorsal and central portion of the pituitary gland and sends solid root-like processes to all the epithelial portions. Within the nervosa are vary- ing amounts of colloid, usually violet- colored, large, amorphous masses and some finer acidophilic granules. 4. The pars anterior is composed almost entirely of small orange-red acidophiles which show no special arrangement into cords or nests. A tongue-like strip of faintly basophilic or chromophobic cells separates the acidophilic portion from the pars nervosa throughout the dorsal half of the gland. 5. Only occasional single or small groups of basophiles are present in the pars anterior of Xiphophorus helleri, Limia tricolor, Mollienisia latipinna and Molli- enisia sphenops, and in a few Lebistes. 6. The pars intermedia consists of small faintly-staining basophiles showing no special arrangement. The posterior re- gion is less cellular and more heavily penetrated by fibres from the nervosa and takes a violet color with the Masson stain. The middle portion is heavily cel- lular and stains light blue. Many small vacuoles but little or no colloid are pres- ent in this region. 7. The uebergangsteil is a thick layer of cells surrounding the pars nervosa com- pletely laterally but only on its ventral half anteriorly and posteriorly. It may form a highly convoluted layer typical of Lebistes and Platypoecilus variatus, or a thicker, less folded layer as in Xi- phophorus helleri. In all cases many strands of the nervosa penetrate it to reach the middle portion of the pars intermedia. Large granular acidophiles, large and more homogeneous basophiles 1942] Potts: Hypophysis of Ovo-viviparous Poeciliids 93 and occasional chromophobes constitute the cellular population of this region. 8. There appears to be a regulated varia- tion of the pars intermedia and the uebergangsteil in which the proportions occupied by the two parts varies in- versely. Not only do the proportions of the two parts change but also the ratio of basophiles and acidophiles in the uebergangsteil changes. No correlation was established, however, between these changes and stages of the reproductive cycle. 9. The pars anterior, intermedia and ner- vosa have an abundant blood supply. The uebergangsteil has no direct blood sup- ply, but it is penetrated in many places by strands of the nervosa tissue, each of which ensheathes a blood vessel. The many foldings of the uebergangsteil layer bring most of the epithelial tissue in close contact with the blood vessels. In cases where the uebergangsteil is not convoluted the nervosa processes with their blood vessels show heavy anasto- moses with each other, thus dividing the epithelium into many patches or islands bordered by blood vessels. Bibliography. Addison, W. H. F. 1916. The Golgi apparatus in the cells of the distal glandular portion of the hypophysis. Anat. Rec., vol. 11, pp. 317-318 (Suppl.) . Atwell, W. J. 1929. Characteristics of the Golgi apparatus in the different types of cells in the anterior lobe of the cat’s hvponhysis. Anat. Rec., vol. 42, p. 44 (Suppl.). deBeer, C. R. 1926. The comparative anatomy, histology, and the development of the pituitary body. Oliver and Boyd, London. Bell, W. R. 1937. Studies of the endocrines of teleosts. I. The morphology of the hypophysis of the goldfish (Carassius auratus L.). Anat. Rec., vol. 70, p. 122 (Suppl.). 1938. Morphology of the hypophysis of the common goldfish (Carassius auratus L.). Zoologica, vol. 23, pp. 219-234. Bock, F. 1928. Die Hypophyse des Stichlings (Gaste- rosteus aculeatus L.) unter besonderer Beruchsichtigung der Jehrescyklischen Veranderungen. Ztschr. wiss. Zool., Bd. 131, S. 643-710. Cardoso, D. H. 1934. Relations entre 1 ’hypophyse et les organes sexuels chez les poissons. C. R. Soc. Biol., T. 115, pp. 1347-1349. Charipper, H. A. 1937. The morphology of the hypophysis in lower vertebrates, particularly fish and amphibia, with some notes on the cytology of the pituitary in Carassius auratus (the goldfish) and Necturus maculosus (the mudpuppy). Cold Spring Harbor Symp. Report, vol. 5, pp. 151-164. Charipper, H. A. & Haterius, H. O. 1932. The histology of the pituitary of the albino rat in relation to the oestrus cycle. Anat. Rec., vol. 54, pp. 15-27. Collin, R. 1924. Passage de la colloide hypophysaire dans la substance cerebrale chez le chien. C. R. Soc. Biol., T. 91, pp. 1334- 1335. Dawson, A. B. & Friedgood, H. B. 1938. The differentiation of two classes of acidophiles in the anterior pituitary of the female rabbit and cat. Stain Tech., vol. 15, pp. 17-81. Florentin, P. 1934. Histophysioloeie comparee de l’hypo- physe. L’excretion de la colloide hypo- physaire chez the Teleosteens. Aim. de Physiol, et Physico. Biol., T. 10, pp. 963-965. Florentin, P. & Weiss, M. 1931. Etude histologique de l’hypophyse de l’anguille (Anguilla anguilla L.). C. R. Soc. Biol., T. 107, pp. 718-720. Herring, P. T. 1908. A contribution to the comparative physiology of the pituitary body. Quart. Journ. Exper. Physiol., vol. 1, pp. 261-289. Houssay, B. A. 1931. Action sexuelle de 1’hypophyse sur les poissons et les reptiles. C. R. Soc., Biol., T. 106, pp. 377-378. Kingsley, J. S. 1926. Outlines of the comparative anatomy of vertebrates. P. Blakiston’s Son and Co., Phila. Kirkman, H. 1937. A cytological study of the anterior hypophysis of the guinea pig and a statistical analysis of its cell types. Amer. Jour. Anat.., vol. 61, pp. 233-287. Levenstein, I. 1939. The cytology of the pituitary gland of 2 varieties of goldfish (Carassius au- ratus L.), with some reference to variable factors in the gland which may possibly be related to the different morphological types. Zoologica, vol. 24, pp. 47-60. Matthews, S. A. 1936. The pituitary gland of Fundulus. Anat. Rec., vol. 65, pp. 357-369. 1937. The development of the pituitary gland in Fundulus. Biol. Bull., vol. 73, pp. 93-98. 94 Zoologica : New York Zoological Society [XXVII: 16 Rasmussen, A. T. 1921. The hypophysis cerebri of the wood- chuck (Marmota monax) with special reference to hibernation and inanition. Endo., vol. 5, pp. 33-66. Rojas, P. L., Castellengo, L. & Silva Alcorta, M. E. 1934. La hypofisis de Jenynsia lineata (Jen). Rev. Soc. Argent, de Biol., vol. 10, pp. 414-420. Scruggs, W. M. 1939. The epithelial components of the tele- ost pituitary gland as identified by a standardized method of selective stain- ing. J. Morph., vol. 65, pp. 187-214. Severinghaus, A. E. 1933. A cytological study of the anterior pituitary of the rat, with special ref- erence to the Golgi apparatus and to cell relationship. Anat. Rec., vol. 57, pp. 149-176. Stendell, W. 1914. Die Hypophysis Cerebri. Oppel’s Lehr- buch der vergl. Mikr. Anat., Teil 8, S. 1-162. Tilney, F. 1911. Contribution to the study of the hypo- physis cerebri with especial reference to its comparative histology. Mem. of the Wistar hist, of Anat. and Biol., No. 2, pp. 3-92. Wolfe, J. H. & Cleveland, R. 1933. Cyclic histological variations in the anterior hypophysis of the albino rat. Anat. Rec., vol. 55, pp. 233-249. 1935. Cyclic histological variations in the an- terior hypophysis of the dog. Z. Zell- forsch., Bd. 17, pp. 420-452. 1942] Potts: Hypophysis of O vo-viviparous Poeciliids 95 EXPLANATION OF THE PLATE. All material illustrated has been fixed with a modified Helly’s fluid, stained with the Masson stain, and cut at 5 micra (except Fig. 3 which was cut at 7 micra). Plate I. Fig. 1. Median sagittal section through the infundibular stalk of a female Mollien- isia latipinna, showing the recessus hypophyseus, nervosa processes enter- ing the pars anterior and other por- tions of the gland. X 140. Fig. 2. Median sagittal section through the pituitary of Platypoecilus variatus showing pars nervosa and part of pars anterior. Neuroglia cells and fibres are visible in the pars nervosa. The tongue- like portion of the pars anterior is clearly visible. X 610. Fig. 3. Median sagittal section through the infundibular stalk of Mollienisia sphe- nops showing nervosa tract descending from the posterior region of the pars nervosa to the ventral border of the gland. A patch of basophiles is present in the pars anterior, adjacent to a nervosa process. The convoluted ar- rangement of the uebergangsteil is apparent. Colloid is abundant both in the pars nervosa and in the posterior region of the pars intermedia. X 250. Fig. 4. Median sagittal section through the infundibular stalk of Xiphophorus helleri showing the four portions of the gland. Small gray patches of baso- philes and the gray tongue-like por- tion can be seen in the pars anterior. Also, small fibrous tracts carrying blood vessels can be seen penetrating the uebergangsteil. X 175. Fig. 5. Median sagittal section through the infundibular stalk of the pituitary of a female Lebistes reticulatus showing the different portions of the gland. The pituitary is more elongate and dorso-ventrally flattened than in the other species. X 300. Fig. 6. Cross section through the infundibular stalk of Platypoecilus variatus show- ing the sharp demarcation of the uebergangsteil (dark staining cells) from the middle region of the pars intermedia. Colloid bodies and nervosa processes can be seen in the nervosa. X 160. Plate II. Fig. 7. Median sagittal section through the infundibular stalk of a male Platy- poecilus variatus, the four portions of the gland. The pars anterior is com- posed almost completely of dark-stain- ing acidophiles. In the pars nervosa colloid bodies are seen among the neu- roglia cells and fibres. The uebergang- steil is composed of basophiles and acidophiles, some of which invade the pars intermedia ventrally. X 220. Fig. 8. Median sagittal section through the pituitary of Platypoecilus variatus, showing the large acidophiles and ba- sophiles of the uebergangsteil. Below them are the small basophiles of the pars intermedia. X 900. Fig. 9. Median sagittal section through the pituitary gland of Limia tricolor. A tract of pars nervosa fibres can be seen penetrating to the ventral border of the gland. Many colloid masses are present in this tract and in the dorsal posterior region. X 240. Fig. 1 0. Cross-section through the infundibular cavity of the pituitary of Lebistes reticulatus, showing the highly con- voluted uebergangsteil. X 300. Fig. 11. Cross-section through the mid-region of the pituitary of Platypoecilus vari- atus, showing the pars nervosa sur- rounded by the uebergangsteil. Colloid bodies are prominent in the pars ner- vosa. Numerous blood vessels carried in nervosa processes penetrate the uebergangsteil. X 460. Fig. 12. Cross-section through the infundibular stalk of the pituitary of Mollienisia sphenops, showing a part of the pars anterior at the right. Beneath the gland may be seen the hypophyseal fenestra. X 160. POTTS, PLATE I THE ANATOMY AND MORPHOLOGY OF THE HYPOPHYSIS OF SEVERAL SPECIES OF OVO-VI VI PAROUS POECILIIDS. POTTS. PLATE II. THE ANATOMY AND MORPHOLOGY OF THE HYPOPHYSIS OF SEVERAL SPECIES OF OVO-VI VI PAROUS POECILIIDS NEW YORK ZOOLOGICAL SOCIETY General Office: 630 Fifth Avenue, New York City OFFICERS Fairfield Osborn, President Alfred Ely, First Vice-president Laurance S. Rockefeller, Chairman, Executive Committee & Second Vice-president Cornelius R. Agnew, Treasurer SCIENTIFIC STAFF General William Bridges, Editor and Curator of Publications Jean Delacour, Technical Adviser John Tee-Van, Executive Secretary Zoological Park Lee S. Crandall, Curator of Birds Leonard J. Goss, Veterinarian Claude W. Leister, Curator of Mammals John Tee-Van, Acting Curator of Reptiles Charles M. Breder, Jr., Director Christopher W. Coates, Aquarist Myron Gordon, Research Associate in Genetics Ross F. Nigrelli, Pathologist G. M. Smith, Research Associate in Pathology Homer W. Smith, Research Associate in Physiology Department of Tropical Research William Beebe, Director Jocelyn Crane, Research Zoologist Henry Fleming, Entomologist Aquarium William K. Gregory, Associate Gloria Hollister, Associate John Tee-Van, Associate Mary VanderPyl, Associate Editorial Committee Fairfield Osborn, Chairman William Beebe Charles M. Breder, Jr. William Bridges Jean Delacour Harry Sweeny, Jr. John Tee-Van ZOOLOGICA SCIENTIFIC CONTRIBUTIONS of the NEW YORK ZOOLOGICAL SOCIETY VOLUME XXVII Parts 3 & 4 Combined Numbers 17-19 Published by the Society The Zoological Park, New York October 23, 1942 CONTENTS PAGE 17. The Synonymy of the Garter Snakes (Thamnophis) , with Notes on Mexican and Central American Species. By Hobart M. Smith 97 18. Membracidae (Homoptera) from British Guiana. By W. D. FUNKHOUSER 125 19. A Consideration of Evolutionary Hypotheses in Reference to the Origin of Life. By C. M. Breder, Jr 131 Index to Volume XXVII 145 1942] Smith: Synonymy of Thamnophis 97 17. The Synonymy of the Garter Snakes (Thamnophis} , with Notes on Mexican and Central American Species. Hobart M. Smith. The accumulation in recent years of rela- tively large series of garter snakes from Mexico and Central America has made pos- sible a more satisfactory definition of the various species occurring in that area than was possible when Ruthven completed his study of the genus in 1908 (Bull. U. S. Nat. Mus., no. 61). The redefinition of them at- tempted here has made necessary a re- examination of types and type descriptions, and a number of surprising facts have come to light. While the allocation of names was undertaken chiefly to clarify the synonymy of Mexican garter snakes, all names which have been proposed in or subsequently re- ferred to the genus Thamnophis have been allocated. Since these are not available in any one place, they are listed below with original place of description, type locality and present status. The discussions which follow this list concern only those species oc- curring in mainland Mexico and Central America. Specimen numbers, unless other- wise indicated, are from the U. S. National Museum. I am much indebted to Dr. E. H. Taylor for numerous courtesies, specimens and ad- vice ; and to Dr. K. P. Schmidt for the op- portunity to study material in Field Mu- seum of Natural History. angustirostris, Eutaenia. Kennicott, Proc. Acad. Nat. Sci. Phila., 1860, pp. 332-3. Parras, Coahuila. = T. rufipunctatus X T. melanogaster hybrid (or = T. angustiros- tris, a valid species). arabdotus, Thamnophis. Andrews, Zool. Ser. Field Mas. Nat. Hist., vol. 20, 1937, pp. 357-8. Catmis, Quintana Roo. = T. sumichrasti praeocularis. atrata, Eutaenia. Kennicott, U. S. Pac. R. R. Surv., vol. 12, 1860, p. 296. California. r= T. ordinoides atratus. aurata, Eutaenia. Cope, Proc. U. S. Nat. Mus., vol. 14, 1892, p. 659. Lake Valley, New Mexico. = T. eques cyrtopsis. Baronis Mulleri, Tropidonotus. Troschel, in Muller, Reisen Ver. Staat., Can., Mex., 1865, pp. 610-611. Mexico. = T. melanogaster canescens. (Not binomial, therefore not available.) baronis-mulleri, Tropidonotus. Boulenger, Cat. Snakes Brit. Mus., vol. 1, 1893, p. 226. Near Mexico City. T . m. melanogaster. bipunctatus, Tropidonotus. Schlegel, Essai Phys. Serp., vol. 2, 1837, p. 320. Nashville, Tennessee. = T. sirtalis sirtalis. biscutata, Eutaenia. Cope, Proc. Acad. Nat. Sci. Phila., 1883, p. 21. Klamath Lake, Ore- gon. = T . ordinoides ordinoides. bovalli, Thamnophis. Dunn, Herpetologica, vol. 1, 1940, pp. 191-2. Granada, Nicaragua. = T. sumichrasti sumichrasti. brachystoma, Eutaenia. Cope, Amer. Nat., vol. 26, 1892, p. 964. Franklin, Venango Co., Pennsylvania. = T. butleri. brunnea, Eutaenia elegans. Cope, Proc. U. S. Nat. Mus., vol. 14, 1892, p. 654. Ft. Bidwell, California. — T. ordinoides biscutatus. butleri, Eutaenia. Cope, Proc. U. S. Nat. Mus., vol. 11, 1889, p. 399. Richmond, Indiana. = T. butleri. canescens, Thamnophis melanogaster. See be- low. Chapala, Jalisco. cerebrosus, Thamnophis sumichrasti. See be- low. Escuintla, Guatemala. clialceus, Prymnomiodon. Cope, Proc. Acad. Nat. Sci. Phila., 1861, p. 558. Siam, in error. = T . sauritus clialceus. chrysocepliala, Eutaenia. Cope, Proc. Amer. Philos. Soc., vol. 22, 1885, pp. 173-174. Ori- zaba, Veracruz. = T. chrysocephalus. collaris, Trovidonotus. Jan, Elenco Sist. Ofkli, 1863, p. 69. Mexico. = T. eques eques. concinnus, Tropidonotus. Hallowed, Proc. Acad. Nat. Sci. Phila., vol. 6, 1852, p. 182. Ore- gon. =■ T. sirtalis concinnus. cooperi, Eutaenia. Kennicott, U. S. Pac. R. R. Surv., vol. 12, 1860, p. 296, pi. 15, fig. 1. Cathlapoot’l and Willopah Valleys, Wash- ington. = T . ordinoides ordinoides. couchii, Eutaenia. Kennicott, U. S. Pac. R. R. Surv., vol. 10, 1859, p. 10. Pitt River, Cali- fornia. = T. ordinoides couchii. cyclides, Thamnophis cyrtopsis. Cope, Proc. Acad. Nat. Sci. Phila., 1861, p. 299. Cape San Lucas, Baja California, in error , — T. eques eques. cyrtopsis, Eutaenia. Kennicott, Proc. Acad. Nat. Sci. Phila., 1860, p. 333. Rinconada, Coahuila. = T. eques cyrtopsis. digueti, Tropidonotus. Mocquard, Nouv. Arch. Mus. Hist. Nat. Paris, ser. 4, vol. 1, 1899, p. 327. Mulege and San Ignacio, Baja Cali- fornia. =: T. digueti. 98 Zoologica : New York Zoological Society [XXVII: 17 dorsalis, Eutaenia. Baird & Girard, Cat. N. Amer. Rept., 1853, pp. 31-32. Between Mon- clova and Rio Grande, Texas, in Coahuila. = T . sirtalis parietalis. eburatus, Tharnnophis. Taylor, Herpetologica, vol. 1, 1939, pp. 187-189, pi. 19, text fig. 2. Cerro San Felipe, 1700 meters, Oaxaca. = T. chrysocephalus. elegans, Eutaenia. Baird & Girard, Cat. N. Amer. Rept., 1853, p. 34. Eldorado Co., Cali- fornia. = T . ordinoides elegans. eques, Coluber. Reuss, Zool. Misc., 1834, pp. 152-155, pi. 8, fig. 2. Mexico. = T. eques eques. errans, Tharnnophis ordinoides. See below. Colonia Garcia, Chihuahua. Faireyi, Eutaenia. Baird & Girard, Cat. N. Amer. Rept., 1853, p. 25. Prairie Mer Rouge, Louisiana. = T. sauritus proximus. flavilabris, Eutaenia. Cope, Proc. Acad. Nat. Sci. Phila., 1866, p. 306. Tableland or south- ern mountains of Mexico. = T . macrostemma macrostemma. fidvus, Eutaenia cyrtopsis. Bocourt, Miss. Sci. Mex., Rept., 1893, pp. 777-8, pi. 62, fig. 2. Alta Verapaz, Guatemala. = T. sumichrasti fidvus. gigas, Tharnnophis ordinoides. Fitch, Univ. Calif. Publ. Zool., vol. 44, pp. 69-73, pi. 5, fig. 7, pi. 7, fig. 7. Gadwall, Merced Co., California. glaphyros, Tropidonotus. Jan, Elenco Sist. Ofidi, 1863, p. 70. North America. = T. radix. godmani, Tropidonotus. Gunther, Biol. Centr. Amer., Rept., 1894, p. 133. Omilteme, Guer- rero. = T. scalaris godmani. graminea, Eutaenia sirtalis. Cope, Proc. U. S. Nat. Mas., vol. 11, 1889, p. 399. Brookville, Indiana. = T. sirtalis sirtalis. halophilus, Tharnnophis. Taylor, Herpetologica, vol. 1, 1939, pp. 183-187, pi. 19, text-fig. 1. Seven kilometers north of Zacualtipan, Hidalgo. = T. phenax halophilus. hammondii, Eutaenia. Kennicott, Proc. Acad. Nat. Sci. Phila., 1860, p. 322. San Diego and Ft. Tejon, California. = T. hammondii. Haydenii, Eutaenia. Kennicott, Expl. Surv. W. 100th Mer., vol. 12, pt. 2, 1860, p. 298. Ft. Pierre, Nebraska. = T. radix. Henshawi, Eutaenia. Yarrow, Proc. U. S. Nat. Mus., vol. 6, 1883, p. 152. Ft. Walla Walla, Washington. = T. ordinoides vagrans. hueyi, Tharnnophis ordinoides. Van Denburgh and Slevin, Proc. Calif. Acad. Sci., ser. 4, vol. 13, 1923, p. 2. Arroyo Encantado, San Pedro Martir Mts., Baja California. hydrophila, Tharnnophis ordinoides. Fitch, Amer. Midi. Nat., vol. 17, 1936, p. 648. Trail Creek- Jackson Co., Oregon. ibibe, Coluber. Daudin, Hist. Nat. Rept., vol. 7, 1803, pp. 181-3. Carolina. = T. sirtalis sirtalis. infernalis, Coluber. Blainville, Nouv. Ann. Mus. Hist. Nat. Paris, vol. 4, 1835, p. 291, pi. 26, figs. 3-3a. California. = T. sirtalis in- fernalis. insigniarum, Eutaenia. Cope, Proc. Amer. Philos. Soc., vol. 22, 1885, p. 172. Chapulte- pec, Distrito Federal, Mexico. = T. macro- stemma macrostemma. jauresi, Tropidonotus. Dumeril & Bibron, Erp. Gen., vol. 7, p. 606. No locality. = T. sirtalis sirtalis. Kennicotti, Tropidonotus. Jan, Elenco Sist. Ofidi, 1863, p. 70. North America. = T. sir- talis parietalis ( ?) . leptocephala, Eutaenia. Baird & Girard, Cat. N. Amer. Rept., 1853, pp. 29-30. Puget Sound. — T. ordinoides ordinoides. lineolata, Eutaenia elegans. Cope, Proc. U. S. Nat. Mus., vol. 14, 1892, p. 655. Southern California. = T. ordinoides vagrans. macrostemma, Eutaenia. Kennicott, Proc. Acad. Nat. Sci. Phila., 1860, p. 331. Mexico City, D. F. = T . macrostemma macrostemma. marciana, Eutaenia. Baird & Girard, Cat. N. Amer. Rept., 1853, pp. 36-37. Red River, Arkansas [Oklahoma], = T. marcianus. megalops, Eutaenia. Kennicott, Proc. Acad. Nat. Sci. Phila., 1860, p. 330. Tucson, Ari- zona; Santa Magdalena, Sonora. = T. macro- stemma megalops. melanogaster, Tropidonotus. Peters, Monatsb. Berl. Akad. Wiss., 1864, pp. 389-390. Mexico. = T . melanogaster melanogaster. melanota, Eutaenia sirtalis. Higley, Trans. Wise. Acad. Sci. Arts Lett., vol. 7, 1889, p. 163. Walworth Co., Wisconsin. = T. sirtalis sirtalis. melanotaenia, Eutaenia radix. Cope, Proc. U. S. Nat. Mus., vol. 11, 1889, pp. 400-401. Brookville, Indiana. = T. radix, mesomelanus, Tropidonotus. Jan, Elenco Sist. Ofidi, 1863, p. 73. Mexico. = T. melanogaster melanogaster. multimaculatus, Atomarchus. Cope, Amer. Nat., vol. 17, 1883, p. 1300. San Francisco River, New Mexico, near Arizona Boundary. — T. rufipunctatus . nigrolateris, Eutaenia. Brown, Proc. Acad. Nat. Sci. Phila., 1889, pp. 421-2. Tucson, Arizona. = T. marciana. nigrilatus, Eutaenia. Cope, Proc. U. S. Nat. Mus., vol. 14, 1892, p. 665. Tucson, Arizona. = T. marciana (emendation of nigrolateris Brown) . obalskii, Tropidonotus. Mocquard, Bull. Mus. Hist. Nat. Paris, 1903, no. 5, pp. 211-2. Black Lake, Canada. = Tharnnophis sirtalis sirtalis. obscura, Eutaenia sirtalis. Cope, Expl. Surv. W. 100th Mer., vol. 5, p. 546. Westport, New York; Lacque Parle, Minnesota; Ft. Benton, Montana ; California. = T . sirtalis sirtalis. The name is here restricted to the five co- types (U.S.N.M. No. 974) from Westport, New York. ocellata, Eutaenia cyrtopsis. Cope, Bull. U. S. Nat. Mus., no. 17, 1880, pp. 22-23. Helotes, Bexar Co., Texas. = T. eques cyrtopsis. olympia, Tharnnophis leptocephalus. Meek, Zool. Ser. Field. Mus. Nat. Hist., vol. 1, 1899, p. 235. Olympic Mts., Washington. = T. ordinoides ordinoides. ordinatus, Coluber. Linnaeus, Syst. Nat., ed. 12, vol. 1, 1766, p. 379. Carolina. == T. sir- talis sirtalis. ordinoides, Tropidonotus. Baird & Girard, Proc. Acad. Nat. Sci. Phila., 1852, p. 176. Puget Sound. = T. ordinoides ordinoides. 1942] Smith: Synonymy of Thamnophis 99 ornata, Eutaenia. Baird, Rept. U. S. Mex. Boundary, 1859, p. 16, pi. 9. Indianola, Texas; Lower Rio Grande, Texas; San An- tonio, Texas. = T. sirtalis parietalis. pallidula, Thamnophis sirtalis. Allen, Proc. Post. Soc. Nat. Hist., vol. 29, 1899, pp. 64-67. Intervale, New Hampshire. = T. sirtalis sir- talis. parietalis, Coluber. Say in Long’s Exp. Rocky Mts., vol. 1, 1823, p. 186. Stone Quarry on west side of Missouri River, 3 miles above the mouth of Boyer’s River. = T. sirtalis parietalis. phenax, Eutaenia. Cope, Proc. Acad. Nat. Sci. Phila., 1868, p. 134. Cordoba, Veracruz (in error ? ) . = T . phenax phenax. Pickeringii, Eutaenia. Baird & Girard, Cat. N. Amer. Rept., 1853, pp. 27-28. Puget Sound. = T. sirtalis pickeringii. plutonia, Eutaenia vagrans. Yarrow, Proc. U. S. Nat. Mus., vol. 6, 1883, p. 152. Arizona. = T. ordinoides vagrans. postremus, Thamnophis eques. See below. Hda. El Sabino, Michoacan. praeocidaris, Eutaenia. Bocourt, Le Natural- iste, 1892, p. 278. Belize, British Honduras. = T. sumichrasti praeocidaris. proximus, Coluber. Say, in Long’s Exp. Rocky Mts., 1823, p. 187. Stone Quarry on west side of Missouri River, 3 miles above the mouth of Boyer’s River. = T. sauritus proximus. pidchrilatus, Eutaenia. Cope, Proc. Amer. Philos. Soc., vol. 23, 1885, p. 174. Probably Guanajuato. = T. eques eques. ij radix, Eutaenia. Baird & Girard, Cat. N. Amer. Rept., 1853, p. 34. Racine, Wisconsin. = T. radix. : rozellae, Thamnophis. Smith, Proc. Biol. Soc. Wash., vol. 53, 1940, pp. 56-57. Palenque, Chiapas. = T. sumichrasti sumichrasti. rubristriata, Thamnophis. Meek, Zool. Ser. Field Mus. Nat. Hist., vol. 1, 1899, p. 235. Olympic Mts., Washington. = T. ordinoides ordinoides. rufipunctatum, Chilopoma. Cope, Rept. U. S. Expl. Surv. W. 100th Mer., vol. 5, 1875, p. 544. Southern Arizona. = T. rufipunctatus. rutiloris, Eutaenia. Cope, Proc. Amer. Philos. Soc., vol. 22, 1885, pp. 388-9. Cozumel Island. = T. sauritus chalceus. sackenii, Eutaenia. Kennicott, Proc. Acad. Nat. Sci. Phila., 1859, p. 98. Florida. = T. sauri- tus sackenii. saurita, Coluber. Linnaeus, Syst. Nat., ed. 12, vol. 1, 1766, p. 385. Carolina. = T. sauritus sauritus. scalaris, Thamnophis. Cope, Proc. Acad. Nat. Sci. Phila., 1860, p. 369. Jalapa, Veracruz. = T . scalaris scalaris. scaliger, T ropidonotus. Jan, Elenco Sist. Ofidi, 1863, p. 70. No type locality. — T. scalaris scaliger. semifasciata, Eutaenia sirtalis. Cope, Proc. U. S. Nat. Mus., vol. 14, 1892, pp. 662-3. Aux Plaines, Illinois. = T. sirtalis sirtalis. sirtalis, Coluber. Linnaeus, Syst. Nat., ed. 10, vol. 1, 1758, p. 222. Canada. — T. sirtalis sirtalis. stejnegeri, Thamnophis. McLain, Contr. Neotr. Herp., 1899, pp. 4-5, pi. Salamanca, Guana- juato. = T. macrostemma megalops. sumichrasti, Eutaenia. Cope, Proc. Acad. Nat. Sci. Phila., 1866, p. 306. Orizaba, Veracruz (in error?). — T. sumichrasti sumichrasti. taenia, Coluber. Schoepf, Reise Ver. Staat., 1788, vol. 1, p. 496. New York (?). = 7\ sirtalis sirtalis. tetrataenia, Eutaenia sirtalis. Cope, U. S. Expl. Surv. W. 100th Mer., vol. 5, 1875, p. 546. Puget Sound, Washington; Pitt River, California. = T. sirtalis tetrataenia. trilineata, Eutaenia sirtalis. Cope, Proc. U. S. Nat. Mus., vol. 14, 1892, p. 665. Port Town- send, Oregon. = T. sirtalis concinnus. trivittatus, T ropidonotus. Hallowell, Proc. Acad. Nat. Sci. Phila., 1853, p. 237. Cosum- nes River, California.=7\ ordinoides elegans. twiningi, Eutaenia radix. Coues and Yarrow, Bull. U. S. Geol. Surv., vol. 4, pp. 279-280. Two Forks of Milk River, Montana. = T . radix. vagrans, Eutaenia. Baird & Girard, Cat. N. Amer. Rept., 1853, p. 35. California, in er- ror. = T. ordinoides vagrans. vicinus, Thamnophis. See below. Temaxcal, Michoacan. vidua, Eutaenia inf emails. Cope, Proc. U. S. Nat. Mus., vol. 14, 1892, p. 658. San Fran- cisco, California. = T. ordinoides atratus. Jan’s Tropidonotus intermedins (Elenco Sist. Ofidi, 1863, p. 70, and Arch. Zool. Anat. Fis., vol. 3, 1865, p. 209), from unknown lo- cality, was placed, with question, in the synonymy of marcianus by Boulenger (Cat. Snakes, vol. 1, 1893, p. 210). This disposi- tion cannot be correct. The snake is de- scribed as follows (translation). “This ser- pent, of unknown provenance, differs prin- cipally from T. natrix, which it resembles not a little, by having 8 supralabials and 21 series of scales. Behind the head are seen two spots that simulate a kind of collar, and on the body six series of small spots dis- posed alternately. Both upper and lower labials have a black border on lip; the pre- dominant tint is olive-color above and yel- lowish below. The specimen examined is 74 centimeters long, of which 14 centimeters and 5 millimeters form the tail.” It is in- ferred from this description that the anal is divided as in T. natrix, with which inter- medins is identical except in the characters mentioned; it is likewise apparent that no light stripes exist in the described specimen (lacking in natrix). I believe the name is based upon some species which does not oc- cur in the western hemisphere. Thamnophis phenax phenax Cope. The five snecimens known of this form are from “Cordoba” (Nos. 30498-9) and “Alpine Region, Orizaba” (No. 707913]). I believe the former locality is incorrect, as it is situated at an elevation of some 2,000 ft., in a humid forest zone. The subspecies is different from all other Thamnophis in having the large dorsal 100 Zoologica : New York Zoological Society [XXVII: 17 blotches completely crossing the back. One specimen is of great interest, as in it the pattern is partially broken, and shows a stage intermediate between the striped pat- tern of most species and the singular one of phenax. Only the nuchal blotch is single; on the anterior half of the body the blotches are divided medially and alternate with each other, and a very irregular, zig-zag median stripe is made evident; at the middle of the body a median series of blotches, similar to the lateral blotches, becomes evident; and on the extreme posterior part of the body this median series of spots is divided into two, forming a total of four series of alter- nating spots on the body. An additional series of poorly defined spots occurs on each side, involving the outer two or three rows of scales, but these are visible also in the other specimens of phenax. The pattern of this aberrant specimen is highly suggestive of a primitive status for the typical phenax pattern, from which the striped patterns, with six series of spots, observable in all other Thamnophis, may have been derived. Most distinctive of halophilus and phenax is the peculiar head pattern, by which they easily are distinguished from close relatives. Thamnophis scalaris Cope. The most diminutive garter snake in Mex- ico is Thamnophis scalaris, a species dis- tinguished not only by its small size but also by its reduced number of supralabials (normally seven, rarely eight), very small, low loreal, a relatively small eye and head, and the tendency toward formation of a single row of large spots on each side to replace the usual double row. A median stripe is always present, but the lateral stripe is poorly defined or invisible, on the second and third scale rows (or second only) when visible. These characters define a form restricted to high elevations (conifer zone) on the central Mexican plateau.1 Since the areas to which these snakes are adapted are separated from each other by broad barriers in the form of semi-arid deserts and plains, a gradual divergence in different directions from the presumably Scale Counts of phenax phenax. Number Sex Scale Rows Ventrals Caudals Supralabials Infralabials Proc. Ptoc. 30499 2 19-19-17 161 65 8-8 9-9 1-1 3-3 7079 2 19-19-17 151 60 8-8 11-11 1-1 3-3 7079 8 17-19-17-15 160 78 7-7 10-10 1-1 3-3 7079 8 19-19-17 158 76 8-8 9-10 1-1 3-4 30498 8 19-19-17 158 73 7-7 9-10 1-1 3-3 Thamnophis phenax halophilus Taylor. In addition to the type, one other speci- men is known, from Tequeyutepec, 7 miles west of Jalapa, Veracruz, at 5,600 ft. (No. 110801). It is a male, with 19-19-17 scale rows, 156 ventrals, 71 caudals, 8-8 supra- labials, 11-12 infralabials, one preocular, three postoculars. The head has the pattern of phenax, with small, parietal spots and most of the remainder of the head light. A pair of large nuchal spots is present, and following these on each side is a double series of very small spots which decrease in size and distinctness posteriorly. In structural characters and form of body the present form is not distinguishable from phenax. It also has the same, peculiar head pattern. The differences between the two are in body pattern ; halophilus shows but little evidence of spots, while in phenax they usually are present, as a single series. A specimen of phenax shows a condition inter- mediate between the two pattern types, how- ever, and accordingly I have little doubt that halophilus is properly associated as a sub- species of phenax. The trend exhibited by halophilus — toward reduction of pattern by subdivision and obsolescence of the spots — is exhibited also by a subspecies of scalaris ( godmani ) . more uniform character of the original stock has taken place in the various isolated populations. As a result three very well de- fined populations may be distinguished. Thamnophis scalaris scalaris Cope. Diagnosis. Supralabials usually seven; eye relatively small ; scale rows usually 17-19-15; ventrals 134 to 145 in males, 136 to 145 in females; caudals 69 to 77 in males, 53 to 65 in females; usually a single row of spots on each side at least anteriorly; two vertical light lines, one on nape and one be- hind eye, enclosing a lateral extension of dorsal head color reaching to upper edge of labials; dorsal stripe splitting the two nape spots, or else a complete transverse light band preceding the fused nape spots. Specimens Examined. Thirty, as follows: Orizaba (Nos. 7076, 12115-6, 30497), Cruz Blanca (No. 110806, EHT-HMS No. 4989), Mt. Orizaba (FMNH Nos. 1517, 1523), Las Vigas (EHT-HMS No. 27916), Cofre de Perote (EHT-HMS No. 5567), Veracruz; Mt. Orizaba, western side, 10,000 ft. (EHT- 1 The specimens mentioned by Ruthven (op. cit., p. 129) from Guatemala are not scalaris but belong to a subspecies of sumichrasti (fulvus). They are discussed under the latter name in the following. 1942] Smith: Synonymy of Thamnophis 101 HMS No. 5566), Puebla-, El Chico National Park (EHT-HMS No. 23512), Guerrero (EHT-HMS No. 5568), Hidalgo; Mt. Popo- catepetl (No. 110815), Lake Zempoala (No. 110816, EHT-HMS Nos. 5299, 5565, 21535), Mexico-, Tres Cumbres (EHT-HMS No. 4964), Km. 54, 26 kilometers east of Cuer- navaca (EHT-HMS No. 4666 ), Morelos. No locality, EHT-HMS No. 23513. Data avail- able also on ANSP No. 11694, Jalapa, Vera- cruz (Ruthven, op. cit., p. 129), and the type (Cope, loc. cit.) Variation. Usually the spots are in a sin- gle series on each side at least anteriorly, but some variation occurs. In certain speci- mens two rows are visible the entire length of the body (save the nuchal spots), and in some of these the posterior spots become very indistinct. The essential features of the head pattern are the two vertical light bars (one postocular and one on nape) which en- close an extension of the dorsal head color reaching the upper edges of the supra- labials; the nape spots are completely sepa- rated by the continuation of the middorsal light stripe to the occiput, or else the two nuchal light bars are prolonged and are fused medially in front of the fused nuchal spots. In some respects the type, as described by Cope, is not exactly typical of the sub- species; among other discrepancies are the 8-8 supralabials. There is, however, no other known species of Thamnophis in that area to which the name may apply. One other specimen with 8-8 supralabials has been seen. Thamnophis scalaris godmani (Gunther) . Diagnosis. Supralabials usually seven ; eye relatively small; scale rows usually not over 17, usually 15 or less in front of anus; ventrals 134 to 158; caudals 61 to 88 in males, 60 to 73 in females; head dark above, the color extending posteriorly to a trans- verse black line across nape, and postero- laterally to level of angle of mouth. Specimens Examined. Twenty-six from the states of Puebla ( San Diego, No. 110810 ; Laguna San Bernardino (EHT-HMS Nos. 27932-4) ; Pajaro Verde (No. 110811, EHT- HMS No. 23801); Oaxaca. (Oaxaca, Nos. 46534, 46604, EHT-HMS Nos. 23774-5; Cerro San Luis, EHT-HMS No. 4997, Cerro San Felipe Summit, EHT-HMS Nos. 15978, 15989) ; Veracruz (above Acultzingo, Nos. 110807-9, EHT-HMS Nos. 5277-9, 5280-1, 27932-4) and Guerrero (Omilteme, EHT- HMS Nos. 23779, 23781). Ten other speci- mens (including the types) are recorded by Boulenger (Cat. Snakes Brit. Mus., vol. 3, 1896, p. 600) from Omilteme, Guerrero, and certain data given. Description (from No. 46534). Head Scale Counts in scalaris scalaris. Number Sex Scale Rows Ventrals Caudals Supral. Infral. Proc. Ptoc. 1517 $ 17-17-15 138 74 7-7 9-9 i-i 2-2 1517 $ 17-17-15 137 75 7-7 8-8 1-1 3-3 1517 $ 17-17-15 138 75 7-7 8-9 i-i 2-2 7076 $ 17-17-15 142 73 7-7 9-9 i-i 2-3 7076 $ 17-19-15 137 72 7-7 9-9 i-i 3-3 7076 $ 17-19-15 134 76 7-7 8-9 i-i 3-3 11694 $ ? 17-19-15 138 73 7-7 9-9 i-i 3-3 30497 s 17-19-15 145 75 7-7 10-10 i-i 3-3 110806 $ 17-19-15 136 77 7-8 9-10 i-i 3-3 4989 $ 17-19-17 138 — 7-8 9-10 i-i 3-3 5568 $ 17-19-17 137 74 7-7 8-8 i-i 3-3 23513 $ 17-19-17 140 73 7-7 9-9 i-i 3-3 5565 $ 19-19-17 133 69 7-7 9-10 i-i 2-3 21535 $ 17-19-15 142 74 7-7 9-9 i-i 3-3 5299 $ 17-19-16 137 70 7-8 10-10 i-i 2-3 23512 2 17-19-16 145 63 7-7 9-10 i-i 3-3 5567 2 17-19-15 139 60 7-7 9-9 i-i 3-3 5566 2 19-19-16 140 58 7-7 9-9 i-i 3-3 110815 2 17-19-17 144 61 7-7 10-10 i-i 3-3 110816 2 17-19-17 136 58 7-7 10-10 i-i 3-3 4666 2 17-19-16 141 — 7-7 8-9 i-i 2-3 4964 2 17-19-17 140 54 7-7 10-10 i-i 3-3 1517 2 17-19-16 141 61 7-7 9-10 i-i 3-3 1517 2 17-19-15 136 65 7-7 9-10 i-i 3-3 1517 2 17-19-15 140 66 7-7 9-9 i-i 3-3 1517 2 17-19-15 138 60 7-7 8-8 i-i 2-3 1523 2 17-19-17 144 61 7-7 9-9 i-i 2-3 7076 2 17-19-15 145 58 7-8 10-10 i-i 3-3 12115 2 17-19-15 135 59 7-7 9-9 i-i 3-3 12116 2 17-19-17 136 53 7-7 9-9 i-i 2-3 Type 2 ? ?— 19— ? 143 59 8-8 10-10 i-i 3-3 102 Zoologica : New York Zoological Society [XXVII: 17 small, not markedly distinct from neck, short; suture between internasals equal to that between prefrontals; internasals about three-fourths as long as prefrontals, a little over half their area; latter extending far onto sides of head; preocular not in contact with frontal; latter hexagonal, shorter than its distance from tip of snout, more than twice as long as broad ; nasal completely di- vided; loreal small (fused with prefrontal on one side) ; a single large preocular; three postoculars ; temporals 1-2 ; supralabials 7-7, third and fourth entering eye; eye small, its vertical diameter a third greater than its distance from labial border, less than height of fifth labial; longitudinal diameter of eye (3.1 mm. ) over half its distance from tip of snout (5.7 mm.) ; infralabials 9-10, 4-5 in contact with anterior chinshields; latter shorter and a little broader than posterior chinshields, which are separated medially. Dorsals in 17-17-16 rows, all of which are keeled ; ventrals 147 ; caudals 64 ; total length 459 mm., tail 106 mm. Top of head uniform brown, the color ex- tending onto nape a distance of four scale lengths, where it is bordered by a black, transverse line; lips nearly white, except for black lines along the sutures between the labials; a very distinct, broad black line bor- dering penultimate labial above and pos- teriorly; dorsal color extending postero- laterally even with angle of mouth, uninter- rupted save by a vague lighter area imme- diately posterior to penultimate labial; be- ginning after transverse nape line, a light (brown) vertebral line on middorsal scale row, visible to tail; sides of body brown, of about same shade as top of head, becoming lighter on first and second scale rows, the color disappearing on ends of ventrals; two rows of very small, scarcely distinguishable, alternating spots on each side; the spots of lower row involving the third and fourth scale rows, the upper ones the seventh and eighth scale rows; the spots are more dis- tinct anteriorly than posteriorly. Belly and subcaudal surface dark slate; chin and throat cream; no black ventral marks save on the extreme anterior edges of ventrals (visible only by lifting overlapping portion of preceding ventrals). Variation. The 25 other specimens ex- amined show scarcely any divergence from the pattern of the described specimen. One, softer than the others, shows the lateral dark spots more plainly. In three the nuchal spots characteristic of scalaris and many other Thamnophis are visible, being darker than the dorsal head color; in No. 46534 they are fused indistinguishably with the head color. The vertical dark bars are not regularly present on the anterior labials, but a curved line on the edge of the sixth labial is regularly present and the most distinct of all. The specimen with 19 scale rows has a short series of abnormally small scales in- tercalated on either side of the vertebral Scale Counts in scalaris godmani. Number Sex Scale Rows Ventrals Caudals Supral. Infral. Proc. Ptoc. 5279 s 17-17-15 153 76 7-7 10-10 1-1 2-3 46534 $ 17-17-16 147 64 7-7 9-10 1-1 3-3 46604 $ 17-19-17-15 147 — 7-7 10-10 1-1 3-3 110807 $ 17-17-15 145 65 7-7 10-10 1-1 3-3 110808 $ 17-17-15 149 75 7-7 10-10 1-1 3-3 Brit. M. $ 148 80 7-7 Brit. M. $ 146 74 7-7 Brit. M. s 146 81 7-7 Brit. M. s 146 78 7-7 Brit. M. s 149 73 7-7 Brit. M. $ 145 81 7-7 5277 $ 17-17-15 152 76 7-7 10-10 1-1 3-3 5280 $ 17-17-14 158 77 7-7 10-10 1-1 3-3 5281 s 17-17-15 150 73 7-7 10-11 1-1 3-3 23775 $ 17-17-15 143 65 7-7 10-10 1-1 3-3 23779 s 17-17-15 142 88 7-7 10-10 1-1 3-4 23774 $ 17-17-15 149 61 7-7 9-9 1-1 3-3 15989 $ 17-17-15 146 65 7-7 10-10 1-1 3-3 5278 2 17-17-15 151 68 7-8 10-10 1-1 3-3 23781 2 17-17-15 136 71 7-8 10-10 1-1 3-3 15978 2 17-18-16 143 63 6-7 10-10 1-1 3-3 23801 2 17-17-15 143 64 7-7 9-10 1-1 3-3 110809 2 17-17-11 141 64 8-8 9-10 1-1 2-3 11022 2 17-17-14 140 57 + 7-7 10-10 1-1 3-3 11551 2 17-17-15 144 60 7-7 9-10 1-1 3-3 Brit. M. 2 142 66 7-7 Brit. M. 2 145 73 7-7 Brit. M. 2 134 61 7-7 Brit. M. 2 141 — 7-7 4997 2 17-17-15 141 — 7-7 9-10 1-1 3-3 1942] Smith: Synonymy of Thamnophis 103 row. One specimen has the penultimate and antepenultimate labials mostly fused. The one with 8-8 supralabials owes its higher count to the presence of a small (but com- plete) labial between the second and third labials. Comparisons. In scale characters this form is most like s. scalaris, but it differs from that by usually having no more than 17 scale rows; typical scalaris usually has 19 near the middle of the body. In details of head pattern the two are remarkably and constantly different. Typical scalaris has the median light stripe usually extending to the occiput, separating the nuchal blotches; if it does not reach the occiput the light areas bordering the nape spots anteriorly meet dorsally. Also a light, vertical, postocular band extends dorsally, to the supraocular; the dorsal head color extends laterally be- tween this postocular and the nuchal light lines, and terminates abruptly at the upper edge of the supralabials. Thamnophis scalaris scaliger (Jan). Diagnosis. Supralabials usually seven; eye relatively small; scale rows usually 19-19-17, never 17-17-15, occasionally 17- 19-17, rarely 17-19-15; ventrals 136 to 151 in males, 135 to 150 in females; caudals 51 to 74 in males, 47 to 65 in females; a single row of spots on each side, seldom two rows ; spots as a rule larger, more rectangular, than in s. scalaris; head pattern as in latter. Specimens Examined. Seventeen, from Mexico City (No. 12730) and Road between Tacubaya and Desierto de los Leones (EHT- HMS No. 5298), Distrito Federal; 15 kilo- meters west of Toluca (Nos. 110812-4), Toluca (No. 32281), Rio Frio (EHT-HMS Nos. 4990, 21524), Llano Grande near Rio Frio (EHT-HMS Nos. 23510-11), and 8-10 miles west of Villa Victoria (EHT-HMS Scale Counts in Number Sex Scale Rows Ventrals 12675 8 19-17-15 144 12730 8 19-17-17 151 32281 8 19-19-17 143 110814 8 17-19-17 143 5570 8 19-19-17 140 5571 8 19-19-17 136 46553 9 17-19-17 139 110812 2 19-19-17 135 110813 2 19-19-17 138 110817 9 19-19-17 143 110818 9 19-19-17 149 5569 9 19-19-17 146 23510 2 19-19-17 142 4990 9 19-19-17 143 23511 2 19-19-17 150 21524 9 19-19-17 144 5298 9 18-19-17 137 No. 5571), Mexico; Mt. Malinche (EHT- HMS Nos. 5569-5570), Tlaxcala; Guana- juato No. 12675) ; 2 miles east of Rio Frio, Mexico, in Puebla (Nos. 110817-8); and Nahuatzen (No. 46553), Michoacan. Variation. In cephalic pattern this sub- species is like s. scalaris. In general its body pattern also is the same, except that the lateral spots are more frequently large and single. Comparisons. This form differs from s. scalaris primarily in the number of scale rows anteriorly and in front of anus; males may usually have moi’e numerous caudals and ventrals. The race is conceived to have a central area of distribution, away from the periphery of the plateau ; s. scalaris oc- cupies the eastern escarpment, which for the most part is separated by arid plains from the central area where s. scaliger oc- curs; s. scalaris also occurs on the southern edge of the plateau in Morelos and Mexico. Further specimens from central Michoacan will be necessary to determine whether the Nahuatzen, Michoacan, specimen is prop- erly allocated with s. scaliger. The name scaliger is applied to the cen- tral subspecies of scalaris with some ques- tion. The scale rows are said to be 19 in the type, which therefore is not the same as godmani. However, it is impossible to cer- tainly ascertain which of the other two sub- species the type may represent. The only in- dication whatever is the statement that the spots are large and “subquadrate.” While this statement is applicable to some speci- mens of typical scalaris, on the other hand in general it more clearly describes the cen- tral subspecies. Boulenger’s description of scaliger (Cat. Snakes, vol. 1, 1893, pp. 203- 204) is of no assistance, but is apparently a composite based perhaps upon scaliger and godmani as well. scalaris scaliger. Caudals Supralabials Infralabials Proc. Ptoc. 56 7-7 9-9 1-1 2-2 52 7-7 9-9 1-1 2-3 52 7-7 9-9 1-1 3-3 51 7-7 9-10 1-1 2-3 74 7-7 10-10 1-1 3-3 71 7-7 10-10 1-1 2-3 57 + 7-7 10-10 1-1 3-3 47 7-7 9-9 1-1 2-3 — 6-7 8-8 1-1 2-2 57 7-7 10-10 1-1 3-3 62 7-7 8-9 1-1 3-3 55 8-8 9-12 1-1 3-3 — 7-7 9-9 1-1 3-3 58 7-7 10-10 1-1 3-3 64 7-7 10-10 1-1 3-3 65 8-8 10-10 1-1 3-3 49 + ? 7-7 9-10 1-1 2-2 [XXVII: 17 104 The three subspecies of scalaris contrasted as follows: Zoologica : New York Zoological Society may be scaliger Posterior scale rows, less than 17 6% (1 in 17) Maximum scale rows 19 100% (17) Anterior scale rows 172 12% (2 in 17) Caudals in males over 60 33% (2 in 6) Ventrals in males 143 or more 67% (4 in 6) Infralabials 10 or more on each side 42% (7 in 17) scala 70% (21 87% (27 93% (2S 100% (U 7% (1 23% (7 2 Some care must be used in determining this character, for the distance on neck in which the scale rows are re- duced to 17, in .y. scalaris and s. scaliger , is very short. This tabulation does not include pattern characters, by which godmani can easily be distinguished from the other two. Curiously, no differences of great significance are dis- cernible in the ventral and caudal counts of the females of the three subspecies. Thamnophis chrysocephalus (Cope). This is one of the few species of the genus entirely lacking a vertebral light stripe, and the only one that combines this character with 17-15 scale rows. Also characteristic of the species is the very light color of the head, sharply differentiated from the black nuchal area. The lateral stripe has irregular edges and involves the second and third scale rows. Ordinarily the doi’sum, between the lateral stripes, is light brown or brownish- gray, with two series of very poorly defined spots on each side; the spots in the outer series are generally better defined than those of the inner series; posteriorly the spots become entirely indistinguishable ; an- terior edges of ventrals irregularly marked with black. However, another color phase is represented by the type of eburatus, in which the whole dorsum (between the stripes) is dark (black), and the entire venter black. This phase appears to be only a variation, since (1) it occurs in a series of specimens from other localities, one in seven from “Orizaba,” and one in three from the vicinity of Acultzingo, Veracruz; and (2) it does not regularly occur in the region of the type locality. Moreover the specimens having the very dark color are otherwise precisely like other specimens with the lighter markings; the differences are only in degree of pigmentation, and do not involve different patterns. Twenty-three specimens examined are from the following localities. Veracruz : Orizaba (Nos. 7077, 30494) ; above Acult- zingo (EHT-HMS Nos. 21536-8; U.S.N.M. No. 110774); Xuchil (F.M.N.H. No. 1519). Oaxaca: Cerro San Felipe (EHT-HMS No. 5556); Totontepec (No. 46445); Mt. Zem- poaltepec (No. 46446). Puebla: Pajaro Verde (Nos. 110775-6). Guerrero: Omil- teme (Nos. 46342, 47747; EHT-HMS Nos. 23778, 23780, 23782). ris godmani in 30) 100% (18) r in 31) 6% (1 in 18) S in 30) 100% (18) l) 100% (16) in 15) 94% (16 in 17) in 31) 72% (13 in 18) Thamnophis vicinus sp. nov. Holotype. EHT-HMS No. 21539, female, from a locality near Temaxcal, Michoacan, about 20 kilometers east of Morelia, col- lected by E. H. Taylor, August, 1939. Para- types. Nine, including No. 15897, a topo- type, and Nos. 15893-6, 15992-5, from Morelia, Michoacan, all collected by H. Dev- lin Thomas. Diagnosis. Similar to e. eques, having 19-19-17 scale rows, 149 to 160 ventrals and 77 to 89 caudals; differing from e. eques in the complete absence of a middorsal light stripe which is replaced by a series of dark spots, and perhaps in having a smaller size. Description of Holotype. Head a half wider than neck, somewhat flattened; eye large, its longitudinal diameter (3.2 mm.) three-fourths its distance from tip of snout (4 mm.) ; full width of rostral visible from above; length of portion of rostral visible from above about half length of internasals; latter about half as wide anteriorly as pos- teriorly, a little shorter than maximum length of prefrontals; frontal somewhat shield-shaped, a little longer (4.2 mm.) than its distance from tip of snout (3.8 mm.), subequal to length of median suture be- tween parietals, four-fifths maximum length of parietals (5.3 mm.), its width about two- thirds (2.7 mm.) its length; supralabials 7-8, three on one side to a point below mid- dle of eye, four on other; antepenultimate labial slightly the largest; nasal completely divided, naris in anterior section, which is a little higher and a little larger than pos- terior section ; a quadrangular loreal, a lit- tle higher than wide; a large preocular, nar- rowly separated from frontal; three post- oculars, more or less subequal in size, the upper two in contact with parietal; tem- porals 1-2-3, the primary much larger than others. Infralabials 10-10, five in contact with anterior chinshields, two (5th and 6th) with posterior; latter separated from each other medially throughout their length, and divergent posteriorly; posterior chinshields perhaps somewhat longer than anterior, lat- ter a little the wider. Dorsal scales strongly keeled, except those in outer row, which are weakly keeled; dorsals in 19-19-17 rows, those scales to- ward middorsum with a single apical notch, none with pits; ventrals 151; anal entire; caudals 77; total length 299 mm., tail 76 mm. 1942] Smith: Synonymy of Thamnophis 105 Top of head light slate, this color merging with cream on sides of head and white on lower parts of labials; posterior edges of supralabials with a vertical black line, that on antepenultimate labial most conspicuous; a large, single, black nuchal spot extending laterally two scales below angle of mouth, anteriorly to within one scale length of rictus oris and to posterior margin of parie- tals, and posteriorly five scale lengths (me- dially) ; middorsum of body brown, this color reaching to the third scale row, where it is replaced by a very light brown or cream color; the first scale row and the lower half of the second is gray; on the pos- terior part of the body the lateral light stripe descends to the first and second scale rows; nowhere is the lateral light stripe sharply defined, and on the anterior third of body it is broken by encroaching black spots and with difficulty can be discerned as a stripe; three series of more or less rounded black spots, none with sharply de- fined edges, on the body; spots in the me- dian series alternating with those of the lateral series; on the first scale row and lateral edges of ventrals another series of smaller black spots alternating with the lat- eral spots; latter spots extending to first scale row on anterior fourth of body, but posterior to this part become restricted to the area above the lateral light line; all spots becoming very poorly defined near middle of body, and posteriorly scarcely dis- tinguishable ; on the posterior part of the body the dorsum appears almost uniform brown ; dorsal surface of tail brown, un- spotted. Ventral surface of head cream; belly and subcaudal surfaces slate, clouded, without distinct black markings. Variation. The paratypes are very much like the holotype in coloration and scutella- tion. All completely lack evidence of a mid- dorsal stripe, and the lateral stripes are poorly defined because of encroachment upon them by the lateral dark spots. The middorsal area is occupied by a series of large dark spots. The largest specimen (a male) measures 592 mm. in total length, the tail 155 mm. Variation in Scale Number Sex Scale Rows Ventrals Caudals 21539 2 19-19-17 151 77 15992 2 19-19-17 149 79 15895 2 19-19-17 156 80 15894 2 19-19-17 152 78 15993 $ 19-19-17 162 15893 $ 19-19-17 160 84 15897 $ 19-19-17 161 88 15994 $ 19-19-17 155 84 15896 $ 19-19-17 160 89 15995 $ 19-19-17 156 86 Remarks. This species bears a very close resemblance, in coloration, to chrysocephalus, with which the specimens at first were as- sociated, without question, until counts were made. These reveal a relationship closer to eques than to chrysocephalus, since the lat- ter regularly has 17-17-15 scale rows. An examination of the maxilla also shows a closer agreement with eques, since the latter (4 specimens) has 24 to 27 teeth, chryso- cephalus (2 specimens) 29 to 30, and the type of vicinus 26. Its chief difference from e. eques is the total absence of the middorsal light stripe ( a middorsal series of dark spots instead ) ; correlated with this is the reduction of the lateral light stripes. The total absence of variation in the nu- merous specimens examined of eques with respect to the character of the middorsal stripe leads me to believe that the specimens referred to vicinus, even though known from a very small area, are not merely vari- ations from the eques norm but represent a distinct population. The uniformity of char- acter of the several specimens of vicinus also points to the probable distinctness of that species. Since vicinus is so like e. eques from the same area in scutellation, the two might be considered subspecies ; this arrangement is not supported, however, by the existence of perfectly typical e. eques at exactly the same localities as those in which vicinus has been collected. For the present they must be con- sidered distinct species. It seems that vicinus furnishes evidence of a close relationship between eques and chrysocephalus in the west, and of the deri- vation of one of these from the other (or from their very close ancestors). There is a southward gradient in the reduction of the median stripe, from eques cyrtopsis in the north, with a broad stripe, to typical e. eques with a narrow stripe, vicinus without a stripe, and chrysocephalus on the extreme south, also without a stripe. A similar gradi- ent in reduction of the number of ventrals also is evident. That vicinus forms a con- necting link in this chain is evident, but evi- dence that chry soceplialus and eques actually intergrade is completely lacking, and it is Counts of vicinus. Supral. Infral. Proc. Ptoc. Temp. 7-8 10-10 1-1 3-3 1-2-3 8-8 10-10 1-1 3-3 1-2-3 8-8 10-10 1-1 3-3 1-2-3/4 8-8 10-10 1-1 3-3 1-2-3 8-8 10-10 1-1 3-3 1-2-3 8-8 10-10 1-1 4-4 1-2-3 8-8 9-9 1-1 3-4 1-2-3/2 8-8 10-10 1-1 4-4 1-2-3/4 8-8 10-10 1-1 3-3 1-2-3 8-8 9-9 1-1 3-3 1-2-3 106 Zoologica : New York Zoological Society [XXVII: 17 assumed that such intergradation does not occur; the completely overlapping ranges bear out this assumption. However, given that orthogenetic trends do exist in this group of the genus, it is not even yet obv- ious in which direction evolution has oc- curred; conclusions by Fitch and Ruthven point toward a north-south direction in this case, while various facts mentioned else- where in this paper support the opposite view. Thamnophis eques eques (Reuss). This form is characterized by having 19 scale rows ; median light stripe on body very distinct and occupying no more than a sin- gle scale row; lateral light stripes on scale rows two and three ; a light head followed by two large, black nape spots (fused or not) ; belly nearly immaculate; ventrals 146 to 171 in males, 145 to 163 in females; caudals 74 to 104 in males, 68 to 95 in fe- males. It ranges from central Durango and southern Sinaloa southward to the edge of the plateau in Michoacan, west to Hidalgo and central Veracruz; highlands of central Oaxaca and the Sierra Madre del Sur of central Guerrero. Thamnophis e. eques can be differentiated from the northern race (e. cyrtopsis) by having 166 or fewer ventrals in males (97%, 31 out of 32) while in e. cyrtopsis only 14% (3 in 21) have 166 or fewer; in females 100% of e. cyrtopsis have 163 or more ven- trals, while in e. eques 3% (1 in 38) have more than 162 ventrals. Thamnophis eques, as previously known, shows a north-south trend in (1) reduction of ventral scale counts; (2) reduction of the median stripe and complete restriction of it to the vertebral scale row; and (3) re- duction in an average number of supra- labials. Most marked of all these trends is the reduction in ventral scale count, which shows a rather abrupt change in central Mexico. The approximate line at which this change occurs does not correspond exactly in position with the zone in which a change in the width of the dorsal stripe occurs; nevertheless, because of the ease of defini- tion of the scale character, this is the one upon which the definition of the two races is Scale Counts in e. eques (females). Museum Number Ventrals Caudals Supral. Infral. Proc. Ptoc. State FMNH 17110 159 — 8-8 10-10 1-1 3-3 D. F. USNM 46482 156 91 8-8 10-10 1-1 3-3 Dgo. EHT-HMS 5282 153 89 7-8 10-10 1-1 3-3 Gto. USNM 9892 156 91 8-8 10-10 1-1 3-4 tt “ 9899 159 68 7-7 9-10 1-1 3-3 tt “ 14433 151 68 7-7 10-10 1-1 3-3 “ “ 14434 156 85 8-8 10-10 1-1 3-3 “ “ 25363 164 73 7-7 10-11 1-1 3-3 “ “ 26147 154 91 8-8 10-10 1-1 3-3 tt tt 26148 155 — 8-8 10-10 1-1 3-3 “ EHT-HMS 21542 149 79 8-8 10-10 1-1 4-4 Gro. tt 23783 148 78 8-8 10-10 1-1 3-3 “ tt 23792 145 79 8-8 11-12 1-1 3-4 tt tt 15974 155 70 7-7 10-10 1-1 2-3 Hgo. tt 15985 158 90 8-8 10-10 1-1 3-4 tt u 23776 163 95 8-8 10-10 1-1 3-3 “ tt 23785 145 85 7-8 10-10 1-1 3-3 Jal. tt 4998 156 98 8-8 11-11 1-1 2-3 it tt 5080 156 70 7-7 10-10 1-1 3-3 Mex. tt 15975 154 72 7-7 10-10 1-1 3-3 ft tt 15890 152 86 8-8 10-11 1-1 3-3 Mich. a 15891 149 87 8-8 8-10 1-1 3-3 USNM 110779 155 — 7-7 10-10 1-1 3-3 tt tt 110780 157 71 — — — — tt EHT-HMS 21525 157 72 7-7 10-11 1-1 3-3 u tt 5276 151 83 8-8 10-10 1-1 3-3 Mor. “ 15977 150 80 8-8 10-10 1-1 3-4 a USNM 46605 149 80 8-8 10-10 1-1 2-3 Oax. EHT-HMS 15979 151 80 8-8 10-11 1-1 3-3 it tt 23784 158 81 7-7 10-10 1-1 3-3 tt tt 23789 146 82 8-8 9-10 1-1 3-3 it “ 23790 152 77 7-8 9-10 1-1 3-3 it 23791 159 — 8-8 10-10 1-1 3-3 tt USNM 46457 150 — 8-8 10-10 1-1 3-3 Sin. “ 46432 160 — 7-7 10-10 1-1 3-3 Ver. “ 25038 149 75 8-8 10-10 1-1 3-3 a tt 30496 153 — 8-8 10-10 1-1 2-4 EHT-HMS 15986 152 — 8-8 10-10 1-1 3-4 9 1942] Smith: Synonymy of Thamnophis 107 largely based. The reduction in numbers of supralabials from 8 to 7 in the southern specimens is not sufficiently universal to warrant use as a key character to define a southern race. Nevertheless, specimens with 7 supralabials almost invariably belong to the southern race. While for convenience I have here spoken of “north-south” trends, they should more properly be termed “south-north” trends for in general this is the direction of evolution in Thamnophis and especially in this section. Scale counts have been taken on 104 Mexi- can specimens of eques eques and e. cyrtopsis. The scale rows in these are uniformly 19- 19-17, except in one from “Orizaba” (with 17-19-15 rows), one from (?) Guanajuato (17-19-17 rows), one from Samachique, Chihuahua (17-19-15), two from Moja- rachic, Chihuahua (19-19-15, 19-19-16) and one from Chilpancingo, Guerrero (19- 19-16). There is no indication that these variations have special significance other than indicating the extremes of variation in the form; the extra scale rows dropped are usually missing only for a very short dis- tance (one scale on one side, six on the other, in one case). The identity of Reuss’ name is not abso- lutely certain. It has generally been asso- ciated with the Mexican form here treated under that name, and at the present, until the type can be re-examined, there is no al- ternative. Jan’s collaris definitely belongs to e. eques, however; in it the middorsal light line is only one scale row wide (see Jan and Sordelli, Icon. Gen., livr. 25, pi. 5, fig. 2). Specimens examined are from the follow- ing localities. Distrito Federal: Rio San Juan de Dios (FMNH No. 17110) . Durango: Durango (No. 8066) ; Huasamota (No. 46482). Guanajuato: Acambaro (EHT-HMS Nos. 5282-3) ; Guanajuato? (Nos. 9892, 9899 [type pulchrilatus ], 14433-4, 25363, 26147-8). Guerrero: 7 miles east of Chil- pancingo (EHT-HMS Nos. 21542, 23783, 23786, 23792) ; Omilteme (EHT-HMS No. 15974); south of Zacualtipan (EHT-HMS Nos. 15985, 23776-7). Jalisco: 20 kilometers south of Guadalajara (EHT-HMS No. 23785); Belen (EHT-HMS Nos. 4965, 4998). Mexico: Nochitongo Ditch, 30 miles north of Mexico City (No. 19003); San Martin (EHT-HMS Nos. 4968, 5080-1); Villa Victoria (EHT-HMS No. 15975) ; Zempoala National Park (No. 15973). Mi- choacan: 15 miles east of Morelia (EHT- HMS Nos. 15889-91, 21533-4 ) ; Los Reyes (No. 46463); Tacicuaro (Nos. 110777-83, EHT-HMS Nos. 21525-32). Morelos: Cuer- navaca (EHT-HMS Nos. 5276, 15977). Oaxaca: Huajuapam (No. 46605); Summit of Cerro San Felipe (EHT-HMS No. 15979) ; Oaxaca (EHT-HMS Nos. 23784, 23789-91). Scale Counts in e. eques (males). Museum Number Ventrals Caudals Supral. Infral. Proc. Ptoc. State USNM 8066 164 — 8-8 10-10 1-1 3-4 Dgo. EHT-HMS 5283 163 101 8-8 10-10 1-1 3-3 Gto. 23786 157 85 7-7 10-10 1-1 4-4 Gro. 23788 152 81 + 8-8 9-9 1-1 2-3 ii “ 23777 161 95 8-8 9-10 1-1 3-4 Hgo. it 4965 159 99 8-8 10-10 1-1 3-3 Jal. USNM 19003 164 95 8-8 10-10 1-1 3-3 Mex. EHT-HMS 4968 166 90 7-7 10-11 1-1 3-3 “ “ 5081 159 82 7-7 9-10 1-1 3-4 ii “ 15973 162 78 7-7 10-10 1-1 2-2 “ a 15889 157 97 8-8 10-10 1-1 3-3 Mich. “ 21533 171 82 7-7 9-10 1-1 3-3 “ “ 21534 158 81 7-7 10-10 1-1 3-4 a USNM 46463 155 93 8-8 10-10 1-1 3-3 a a 110777 162 81 7-7 10-10 1-1 3-3 “ a 110778 159 80 7-7 10-10 1-1 3-3 “ “ 110781 166 84 7-7 10-10 1-1 3-4 “ 110782 146 — 7-7 9-9 1-1 2-2 “ a 110783 153 100 8-8 10-10 1-1 3-3 “ EHT-HMS 21526 159 89 8-8 10-10 1-1 3-3 “ ii 21527 165 80 7-7 11-11 1-1 3-3 a ii 21528 160 79 7-7 10-10 1-1 2-2 “ ii 21529 161 88 8-8 10-10 1-1 3-3 “ ii 21530 159 77 7-7 10-10 1-1 3-3 “ ii 21531 161 76 7-7 — — — “ ii 21532 163 74 7-7 10-10 1-1 3-3 “ USNM 46423 160 99 8-8 10-10 1-1 3-3 Zac. “ 32279 166 76 7-7 8-9 1-1 3-3 7 “ 32280 161 78 7-7 10-10 1-1 3-3 7 EHT-HMS 5273 164 104 8-8 9-10 1-1 3-3 7 a 15976 166 — 7-7 10-10 1-1 3-3 7 u 23620 163 — 7-7 8-9 1-1 3-3 108 Zoologica : New York Zoological Society [XXVII: 17 Sinaloa: Rosario (No. 46457). Veracruz: Las Vigas (No. 46432); Mirador (No. 25038); Totalco (EHT-HMS No. 27908); Orizaba (No. 30496). Zacatecas: San Juan Capistrano (No. 46423). No locality: (Nos. 32279-80; EHT-HMS Nos. 5273, 15976, 15986, 23620). The species has been re- ported also from Moro Leon, Guanajuato, and Tezuitlan, Puebla. Thamnophis eques cyrtopsis (Kennicott) . This race differs from e. eques of the southern part of the Mexican plateau chiefly in the greater average number of ventral scales. In e. cyrtopsis 86% of the males have 167 or more ventrals, and 100% of the fe- males have 163 or more ventrals; in e. eques, on the other hand, only 3% of the males have 167 or more ventrals, and only 3% of the females have more than 162 ventrals. In addition, over most of its range e. cyrtopsis is characterized by having the middorsal light stripe involving parts of ad- jacent scale rows on various parts of the body; in e. eques the middorsal stripe is al- ways confined strictly to the vertebral scale row. Toward the south specimens of e. cyrtopsis have a stripe like e. eques, so the ventral count must be relied upon to dis- tinguish the two races. Very rarely do 7 supralabials occur in e. cyrtopsis, while in e. eques 7 occur as fre- quently as 8. The range of e. cyrtopsis in Mexico is the eastern half of Sonora southward to north- ern Durango and along the Sierra Madre Occidental to northern Nayarit; eastward through Chihuahua to eastern Coahuila, and southward on the central plateau to south- ern San Luis Potosi and probably northern Zacatecas. Specimens of this race have been ex- amined from the following localities. Chi- huahua: Arroyo del Alamos, Casas Grandes (No. 42876) ; Basuriachi (FMNH No. 11823) ; Cajon Bonito Creek (No. 21056) ; Chihuahua (No. 14256) ; Guadelupe y Calvo (No. 46356-8); Mojarachic (EHT-HMS Nos. 18962-3, 23063-6, 23787) ; San Luis Mts. (Nos. 21057-8); Samachique (FMNH Nos. 11822, 11824, 15727). Coahuila: 21 miles north of Saltillo (No. 105303) ; Rin- conada (No. 8067, type cyrtopsis) . Durango : Guanacevi (No. 46367). Nayarit: Santa Teresa (Nos. 46420-1). San Luis Potosi: Hacienda La Parada (No. 46410). Sonora: La Posa, 10 miles north of Guaymas (EHT- HMS Nos. 4866-73) ; Guadelupe Canon (No. 21059). Scale Counts of eques cyrtopsis. Museum Number Sex Ventrals Caudals Supral. Infral. Proc. Ptoc. State FMNH 11823 2 170 — 8-8 10-10 1-1 2-3 Chih. USNM 46356 2 175 81 7-8 10-10 1-1 3-3 U EHT-HMS 23063 2 169 84 8-8 10-10 1-1 3-3 u “ 23065 2 176 95 8-8 10-10 1-1 3-3 u “ 23066 2 168 86 8-8 11-12 1-1 3-4 a “ 23787 2 174 90 8-8 11-11 1-1 3-3 “ USNM 21058 2 167 85 8-8 9-10 1-1 2-2 u “ 46367 2 172 88 8-8 10-10 1-1 2-3 Dgo. “ 46420 2 172 84 7-7 10-10 1-1 3-3 Nay. “ 46421 2 166 84 7-7 10-10 1-1 3-3 U EHT-HMS 4866 2 163 78 8-8 10-10 1-1 3-3 Son. “ 4872 2 171 86 8-8 10-10 1-1 2-2 “ a 4873 2 166 77 + 8-8 10-10 1-1 3-3 U USNM 42876 $ 177 91 8-8 10-11 1-1 3-3 Chih. “ 21056 $ 169 94 8-8 10-10 1-1 3-3 it u 14256 S 172 102 8-8 10-10 1-1 3-3 U “ 46357 S 174 92 8-8 10-10 1-1 3-3 (( “ 46358 $ 174 — 8-8 10-10 1-1 3-4 u EHT-HMS 18962 $ 174 89 8-8 10-10 1-1 3-3 a U 18963 $ 169 99 8-8 10-10 1-1 3-3 * it 23064 S 173 98 8-8 10-10 1-1 3-3 u USNM 21057 $ 173 91 8-8 10-10 1-1 3-3 (( FMNH 11822 $ 174 92 8-8 10-10 1-1 3-3 “ “ 11824 $ 177 95 8-8 10-10 1-1 3-3 “ “ 15727 $ 163 — — — — — USNM 105303 $ 169 89 8-8 10-10 1-1 3-3 Coah. U 8067 $ 173 88 8-8 11-11 1-1 2-3 “ 46410 $ 174 91 7-8 10-10 1-1 3-3 S.L.P. EHT-HMS 4867 $ 172 91 8-8 11-11 1-1 3-3 Son. u 4868 $ 169 88 8-8 10-11 1-1 3-3 “ u 4869 $ 163 84 8-8 10-11 1-1 3-3 “ u 4870 $ 171 — 9-10 9-10 2-2 3-3 * “ 4871“ $ 172 — 8-8 11-11 1-1 3-3 “ a 21059 $ 164 93 8-8 10-10 1-1 3-4 1942] Smith: Synonymy of Thamnophis 109 Thamnophis eques postremus subsp. nov. Type. E. H. Taylor-H. M. Smith Coll. No. 5275, El Sabino, Michoacan, collected by H. M. Smith. Paratypes. Three, Nos. 5274, 5285-6, Uruapan, Michoacan. Diagnosis. Similar to e. eques, except ventrals fewer, 138 to 141 in the only speci- mens known (4 females) ; caudals 71 to 72; supralabials 7 to 8; lateral light stripes not apparent; middorsal stripe but very little lighter than ground color, scarcely or not evident; dark spots very much reduced, al- ternating with each other in two series on either side; middorsum not traversed by spots; outer row of dark spots on first and second scale rows greatly reduced, generally scarcely visible; no distinct dark spots on venter. Description of Holotype. Length of por- tion of rostral visible from above a little more than half length of median suture be- tween internasals ; latter a little longer than prefrontals, but much narrowed anteriorly, their combined width anteriorly somewhat less than a third the length of the posterior border of the rostral; frontal pentagonal, anterior edge straight, posterior edges forming an obtuse angle of about 110°; length of frontal subequal to its distance from tip of snout and from posterior tip of parietals; nasal divided, nostril largely in anterior section and bordering posterior section; loreal large, nearly square; a large preocular, rather widely separated from frontal; three postoculars, upper a little the largest, lowest smallest and wedge-shaped; temporals 1-3-3, the primary much the largest. Supralabials 7-7, 5th largest, 3d and 4th entering orbit; infralabials 10-10, 5 bordering anterior chinshields, two (5th and 6th) bordering posterior chinshields; length of suture between 1st infralabials subequal to length of mental ; anterior chin- shields a little shorter and a little broader than posterior chinshields; latter separated from each other their full length. Dorsal scales in 19-19-17 rows, strongly keeled, those in outer row with relatively weak keels ; apical notches present on dorsal scales, but no typical apical pits; ventrals 139; caudals 72; total length 575 mm.; tail 144 mm. ; female. Maxillary teeth 27. Dorsal surface of head slate gray, un- marked; sides of head the same color, be- coming cream at labial border, but posterior borders of all labials with a distinct vertical black streak, most prominent on 5th labial; nape with a large black blotch fading an- teriorly into the slate color of head, pos- teriorly sharply terminating about six scale lengths from edge of parietals; nape spot extending slightly below level of angle of mouth, and somewhat indented on middorsal line. Ground color of body slate gray, a little lighter than head color; a poorly defined, short area cream in color immediately pos- terior to dark nuchal spot; a very feebly defined light line on vertebral scale row, very slightly lighter than ground color and lacking darker edges; lateral light stripes not visible but their position on the second and third scale rows indicated by the re- striction of dark marks to areas lateral and medial to these rows. Sides of body with two series of small, poorly outlined but distinct, alternating black spots widely separated from each other and not encroaching what- ever on the vertebral line or upon the second or third scale row; on the outer, anterior edges of certain scales of the outer row of dorsals is a small black spot; these tend to occur on alternate scales. Venter cream, immaculate save for a small dark streak near either end of each ventral, placed on the anterior edge and concealed by the overlapping edge of the preceding scale ; these small dark marks not visible except when scales are spread apart ; chin and gular region cream, unmarked save for a dark streak on the posterior border of a few of the posterior infralabials; sub- caudal surface immaculate. V aviation. One of the paratypes is marked exactly like the holotype. Another (No. 5286 ) has somewhat larger dark spots on the dorsal surface ; the corners of the spots nearly touch on either side, but in no case do the spots cross the middorsal line ; those in the paravertebral series alternate with each other. The last specimen, a juvenile, agrees with No. 5286 in pattern. Variation in Scale Counts of eques postremus. Number 5275 5274 5286 5285 Sex 9 9 9 9 Dorsals 19-19-17 19-19-17 19-19-17 19-19-17 Ventrals 139 138 140 141 Caudals 72 71 72 72 Supralabials 7-7 7-7 8-8 7-7 Infralabials 10-10 10-10 10-10 10-10 Preoculars 1-1 1-1 2-2 1-1 Postoculars 3-3 3-3 3-3 3-3 Temporals 1-3-3 1-3-3 1-3-3 1-3-3 Total length (mm.) 575 571 441 192 Tail length (mm.) 144 141 105 47 110 Zoological New York Zoological Society [XXVII: 17 In the specimen with 2-2 preoculars the normal preocular is split across the middle ; it is a type of anomaly not infrequent in eques and its relatives. Remarks. In the low ventral count, e. postremus is like s. sumichrasti, s. praeocu- laris, s. fulvus and perhaps s. cerebrosus. The last and s. praeocularis have distinct spots on the belly and distinct, broad ver- tebral light stripes. In s. sumichrasti there is a median series of dark spots alternating with a series of lateral spots, as in vicinus. The closest resemblance to s. postremus of all these is held by s. fulvus, which also lacks belly spots and distinct stripes. In it, however, the middorsal stripe is visible, dis- tinct and relatively broad (one and two half rows to three full rows) on the anterior part of the body, and becomes narrower pos- teriorly, disappearing completely on the posterior part of the body. In e. postremus the dorsal stripe is restricted completely to one scale row (as in e. eques ) and is of equal indistinctness throughout its length. Thamnophis sumichrasti sumichrasti (Cope) . This form is characterized by the absence of a dorsal stripe, presence of a medial series of dark spots, a very poor definition of the lateral light stripes, and by having 19 scale rows and 139 to 157 ventrals, 58 to 72 cau- dals. In form, scutellation and type of pat- tern it is undoubtedly a close relative of eques, which differs by having a distinct dorsal stripe and higher average ventral and caudal counts. The two are considered as distinct species because there is no evi- dence whatever of an intergradation be- tween the curious pattern of sumichrasti — a median and two lateral series of spots — with that of eques, which has a median light stripe and four lateral series of spots. T. sumichrasti does intergrade with races which do have patterns similar to that of eques {viz., s. fulvus, s. praeocularis), but from that fact it cannot be inferred that sumichrasti must also intergrade with eques. For some time I labored under such an inference, but the discovery of vicinus has made it evident that intergradation be- tween forms with patterns as different as those of sumichrasti and eques is not to be taken for granted. T. vicinus, with a pattern like sumichrasti but with a ventral count like eques, occurs in the same localities as the latter species, yet remains quite dis- tinct. Since it is so close a parallel of su- michrasti ( differing chiefly in ventral count) , I believe it unwise to assume that the latter is a subspecies of eques, any more than is vicinus. Thus it appears that there are two centers of dispersal of the complex eques group : one in Guatemala, the other in Mexico. The most primitive of the groups may be vicinus and s. sumichrasti, respectively, each of which has given rise to striped forms some of which closely parallel each other (e.g., e. postremus and s. fulvus). The chief differ- ence between the history of the forms de- veloping from these two centers of dispersal is that vicinus has become completely dis- tinct from its striped derivative, eques, while sumichrasti still remains linked with its striped derivatives. Chiapas specimens of sumichrasti show definite evidence of in- tergradation of that form and s. fulvus. In spite of evidence of separate centers of dispersal for sumichrasti and eques, the two species obviously have had a common origin, and chrysocephalus seems very near the an- cestral type from which they were derived. T. chrysocephalus shows a strong similarity to vicinus, differing chiefly in number of scale rows and maxillary teeth. The two cotypes of sumichrasti are very small and soft. They completely lack stripes, however, and have a series of spots crossing the vertebral scale row. Moreover they have 19 scale rows (19-19-17) and the low ven- tral and caudal counts typical of the sub- species. They may actually have been se- cured near Orizaba, as stated by Cope, but it appears more probable that they were col- lected on the Isthmus of Tehuantepec, on the Atlantic slopes. At best the locality data cannot be relied upon until verified by fur- ther specimens from the region of Orizaba, for in numerous instances the data for Sumichrast’s specimens are known to have been misrepresented. The name sumichrasti has previously been applied to a 17-scale-rowed form ( chrysoce- phalus + scalaris godmani) and for this reason Dunn and I nearly simultaneously applied different names to the stripeless, 19-scale-rowed form to which actually be- longs the name sumichrasti. These three names {sumichrasti, bovalli, rozellae) clear- ly refer to the same form. Scale Counts of s. sumichrasti. Museum Number Sex Scale Rows Ventrals Caudals Supral. Infral. Proc. Ptoc. USNM 25234 2 19-19-17 149 66 8-8 10-10 1-1 3-3 USNM 26501 8 19-19-17 157 63 8-8 9-9 1-1 3-3 USNM 26502 8 19-19-17 148 68 8-8 8-8 1-1 3-3 USNM 46549 2 19-19-17 139 72 8-8 9-10 1-1 3-3 USNM 108597 2 19 (20) -19-17 140 58 7-7 9-9 1-1 3-4 ITSNM 108598 8 19-19- (15)17 148 — 8-8 9-9 1-1 3-3 USNM 108599 8 19-19-15 151 — 7-7 9-9 1-1 3-4 1942] Smith: Synonymy of Thamnophis 111 The localities now represented by speci- mens are: Tabasco : Montecristo (USNM No. 46549). Veracruz : ?Orizaba (USNM Nos. 26501-2). Chiapas: Palenque (USNM No. 108597) ; Aguacate, 24 kilometers north of Palenque (USNM No. 108598-9). Guate- mala: (USNM No. 25234). Nicaragua: Granada (Dunn, loc. cit.) Thamnophis sumichrasti praeocularis (Bocourt) . The types of praeocularis are described with 140 ventrals, 68 to 74 caudals, and 19 scale rows. The type of arabdotus (FMNH No. 26994) has 140 ventrals, 72 caudals, and 19-19-17 scale rows (female). Two fe- males in the U. S. National Museum (Nos. 46528-9) from Puerto Morelos, Yucatan, have 137 and 136 ventrals, respectively; caudals 70, 71; scale rows 19-17-15, 19- 19-16. The patterns in all these specimens are remarkable alike, including a very broad median stripe; large nuchal blotches fol- lowed by smaller blotches on the body, of which a few anterior are fused to form sin- gle lateral cross-bands ; and a series of black spots on each side of venter, a spot to the side of each ventral. This uniformity in pat- tern, as well as in most features of scutella- tion, leads me to believe that the presence of three preoculars in the types (both?) of praeocularis is an anomaly; there is no other feature which could distinguish them from arabdotus as another species or sub- species. Thamnophis sumichrasti cerebrosus* subsp. nov. Holotype. U. S. Nat. Mus. No. 12734, fe- male, Escuintla, Guatemala, collected by H. J. Stuart. Paratypes. U. S. Nat. Mus. No. 12735, topotype; F.M.N.H. No. 410, “Guatemala.” Diagnosis. Dorsal stripe covering one and two half scale rows, with continuous lateral edges; lateral stripe poorly defined or absent ; two regular rows of dark spots on venter, at least toward posterior part of belly; head light above, with discreet verti- cal dark marks on edges of supralabials; ventrals 144 to 167. Scale rows regularly 19-19-17. Description of Holotype. Internasals a little shorter than wide, their combined con- tact with rostral considerably greater than a nasorostral suture; prefrontals a little longer than internasals; frontal very slightly shorter than its distance from tip of snout or length of parietal suture ; greatest width of frontal four-fifths its length, least width about half its length; one loreal, about as high as long; a large preocular; three post- * Light-headed , in reference to the relatively light ground color and poorly defined dark marks on head. oculars, lower smallest; temporals 1-3, an- terior very large; supralabials 7-7, 3d and 4th entering eye, 5th highest and largest; ten infralabials, five in contact with an- terior chinshields, two with posterior; pos- terior chinshields separated from each other, a third longer than anterior chin- shields. Dorsal scales in 19-19-17 rows, all strongly keeled except those of outer row, which are weakly keeled ; ventral 144 ; anal divided; subcaudals 66 (female). Head very light brown; two small, white pineal spots; posterior edges of 2d to the 5th supralabials inclusive black-edged, the anterior scales less prominently; otherwise head without markings. Two dark nuchal blotches partially fused medially; a very distinct, sharply outlined middorsal light stripe covering one and two half scale rows, extending from nape onto tail; below this two series of alternating spots on each side, the lower reaching to and including the upper part of the third scale row; irregular dark areas on many scales of the three outer rows. Belly light, with small, black spots near the anterior end of each ventral, ar- ranged in a series on each side, near ends of ventrals ; other scattered dark spots and irregular dark areas near bases of many ventrals, especially those near middle of body; chin with a very few, tiny dark flecks below. Variation. The topotypic paratype has an anomalous pattern, and the head scutella- tion is not normal. Apparently it should have 7-8 supralabials, as the outlines of the scales indicate this number, but they are so fused that only six remain on each side. In scutellation it is otherwise normal, and has 10-10 infralabials, 1-1 preoculars, 3-3 post- oculars, 19-19-17 scale rows, 149 ventrals and 74 caudals (male). The head shows no markings whatever. The body lacks all pat- tern characters save the middorsal stripe, which is narrower than in the type (in- volves only the edges of the paravertebral rows) and black-bordered; there are flecks of black on some lateral scales, and the nuchal blotches are indicated by black flecks on certain nuchal scales; the belly is unspotted. The Field Museum paratype has 167 ven- trals, 78 caudals, 7-7 supralabials, 10-10 in- fralabials, 1-1 preoculars and 19-19-17 scale rows. It is like the holotype, except that the pattern is less well defined (scales not spread apart as in holotype) ; the an- terior spots are fused together, reproducing more or less the lateral neck pattern of praeocularis; and the rows of spots on the belly are shorter and the spots smaller. Comparisons. This form appears to be most closely related to praeocularis, having 112 Zoologica: New York Zoological Society [XXVII: 17 a very similar body pattern. The most dis- tinctive pattern differences are found on the head. In praeocularis the dorsal sui'face of the head is dark, the lips are strongly barred, and the nuchal blotches are fused with the head color. The ventrals are a little less numerous than in cerebrosus, varying from 136 to 140 in four specimens (fe- males) . Whether there will prove to be aver- age differences in neck pattern cannot now be stated ; in praeocularis the first few body spots are single, instead of double and al- ternating, and cross the lateral light stripe; this pattern does not occur in the type of cerebrosus, but does in the single normal paratype. It is also possible that the number of supralabials in cerebrosus may average fewer than in praeocularis. Thamnophis sumichrasti fulvus (Bocourt) . This rather well-defined form is charac- terized by the indistinctness of the dorsal stripe, which is poorly defined and about one scale row wide (except sometimes on the neck) ; the head generally is light, and always with no or very discreet dark labial marks; dark marks on the belly, as in prae- ocular and cerebrosus, are completely lack- ing; scale rows 19-19-17 (a short row me- dially on each side, increasing the count to 21, in one specimen out of 32; one other specimen with 15 posteriorly) ; supralabials regularly 8-8; ventrals 136 to 153; caudals 58 to 76. The subspecies is known only from cen- tral Guatemala. Specimens examined are from Sierra Santa Elena, Tecpam, 9500 ft. ( FMNH Nos. 1926, 30432-3) and Chichi- vac, Chimaltenango (FMNH Nos. 20261-2, 20275). Slevin ( Proc. Calif. Acad. Sci., ser. 4, vol. 23, 1939, p. 397) records 26 speci- mens from Chichivac, in the vicinity of Tecpam (CAS Nos. 66983-98) ; Lake San Antonio, near San Antonio (CAS Nos. 66983-98) ; and Finca El Potrero, Volcan Agua (CAS No. 66973). The race may exist in Chiapas. Apparently it occurs only at relatively high elevations, while cerebrosus and praeocularis are lowland and foothill subspecies. Scale Counts Number Sex Scale Rows Ventrals 20261 $ 19-19-15 146 20262 $ 19-19-17 145 1926 2 19-19-17 136 20275 2 19-19-17 142 30432 2 19-19-17 144 30433 2 19-21-17 142 3 See Slevin {loc. cit.) for counts on the 26 specimens in the California Academy of Sciences. The closest relative of fulvus is not prae- ocularis or cerebrosus, which are closely as- sociated geographically, but more probably ' sumichrasti. With the latter fulvus agrees in head pattern (no dark labial marks), ventral pattern (no spots) and in scale counts; the only significant difference be- tween these two is in the presence of stripes (although dim except on neck) in fulvus, I the absence of them in sumichrasti. This ' difference, though slight, is very constant in the specimens examined. Thamnophis ordinoides errans subsp. nov. Holotype. USNM No. 46336, female, from Colonia Garcia, Chihuahua. Paratypes. \ Three topotypes, Nos. 46337-9, and two j from Coyotes, Durango (FMNH Nos. 1499A-B) . Diagnosis. A member of the ordinoides group, with 19-19-17 scale rows and maxil- lary teeth 17 or 18, the posterior not con- spicuously enlarged, although a little longer i than the anterior teeth; ventrals 155 to 166 I (155-156, females; 163-166, males) ; caudals 72 to 91 (72 to 82, females; 85 to 91, males) ; supralabials seven or eight; sixth labial large and not narrowed above ; stripes ) very poorly defined as a rule, median stripe generally covering but one scale row where visible; spots between stripes small, those of the outer row usually very poorly de- fined or absent; top of head dark, the color extending onto sides; supralabial region light, except for black posterior edges on most of the labials; a vaguely darker, longi- tudinal line through lower temporal region. Description of Holotype. Portion of ros- tral visible from above as long as inter- nasal; suture of one of latter with rostral as great as naso-rostral suture, or a little longer; one preocular; three postoculars; temporals 1-2-3; eight supralabials on one side, on the other side the two subocular labials fused; anterior and posterior edges of 6th labial parallel, inclined forward slightly; 6th labial a little higher than long; ten infralabials; posterior chinshields sepa- rated from each other, a little longer than anterior chinshields. of s. fulvus .3 Caudals Supral. Infral. Proc. Ptoc. 69 8-8 10-10 1-1 3-3 75 8-8 10-10 1-1 3-3 — 8-8 9-10 1-1 3-3 3-3 60 8-8 10-11 1-1 59 8-8 10-10 1-1 3-3 58 8-8 10-11 1-1 3-4 1942] Smith: Synonymy of Thamnophis 113 Scale rows 19-19-17, the fourth row drop- ping at the 104th ventral; ventrals 156; anal entire; caudals 82, total length 545 mm. ; tail 139 mm. Dorsal surface of head uniform dark brown, the color extending onto sides, with- out mai'ks; labial region cream; an area on posterodorsal edge of labials (except last) black; loreal, nasal, two lower postoculars and lower half of preocular suffused with cream; temporal region dark, nearly black at sutures with labials, abruptly differen- tiated from light labial color; posterior to temporal region and parietals, the brown color of head shades into a black nuchal patch extending posteriorly four scale lengths, this patch extending laterally a lit- tle below angle of mouth, and sharply dif- ferentiated posteriorly and below from the adjacent light color. Dorsal ground color brown, lighter brown below middle of third and on vertebral scale rows; on neck a sin- gle, poorly defined series of transversely elongate spots; this series breaks into two rows, and the lower of these rows dis- appears on anterior fourth of body; the median series of spots continues about to middle of body, there disappearing; on pos- terior third of body no markings whatever are visible, and only the lighter ground color below the middle of the third row indicates the position of the lateral light stripe. Tail uniform brown. Chin and gular area cream, unmarked ; otherwise entire ventral surface slate gray, the posterior edges of the scales light. Variation. In coloration the subadult male and one juvenile topotype are just like the holotype; the other juvenile topotype has the spots in the outer row more clearly defined, visible (though dimly) to the tail. All agree with the type in having the naso- rostral suture much smaller than the com- bined internasal-rostral sutures. The two paratypes from Coyotes, Du- rango, differ in coloration from the topo- types only in the greater distinctness of the dorsal stripe. In the larger the stripe is dis- tinct the full length of the body, and in- volves the inner halves of the paravertebral scale rows; it divides the nuchal blotches and reaches the parietals. In the smaller specimen the stripe is only one scale row wide, but it is distinct posteriorly as well as anteriorly. There are no markings below the vaguely indicated lateral light line. Remarks. These are the first specimens recorded from mainland Mexico (i.e., exclu- sive of Baja California) of the ordinoides group. They differ from all except ordi- noides, atratus, and hydrophila by having a maximum of 19 scale rows. Of these, ordi- noides has fewer ventrals and caudals; atratus has a distinct dorsal stripe and fre- quently has red in the dorsal color; and hydrophila regularly has eight labials, the sixth narrower above than below, and the combined internasal-rostral sutures less than a single naso-rostral suture (these characters from Fitch, Univ. Calif. Publ. Zool., vol. 44, 1940 ) . Of known Mexican species, the one most easily confounded with errans is eques eques, a form having similar scutellation at least in the southern part of its range. The latter may be distinguished by the presence of dark spots below the lateral light line ( on the scales); by the distinct lateral light line; and by the number and character of the maxillary teeth. Two eques eques exam- ined (from the states of Sonora and Mex- ico) have 24 and 26 maxillary teeth, and the last is much larger (two or three times) than the anterior and middle teeth. Two errans (one from each locality) have 17 and 18 maxillary teeth, and the posterior is but little larger than the middle and anterior teeth ; these maxillae resemble one extracted from a specimen of ordinoides vagrans from Rinconada, New Mexico (USNM No. 44361), with 20 teeth, the last but little enlarged. The relationships and phylogenetic posi- tion of errans are not clear, and cannot well be guessed until more specimens show the type of variation that exists in it. The re- semblance to eques is so close that the possi- bility of its derivation from the latter should be considered; if such is the case, then errans would have to be looked upon as the most primitive of the ordinoides group ; this is the view here considered the most probable. According to Fitch’s theories, Scale Counts in ordinoides errans. Number Sex Scale Rows Ventrals Caudals Supralabials Infralabials Preoculars Postoculars Total length (mm.) Tail length (mm.) 46336 46339 9 9 19-19-17 19-19-19 156 155 82 72 8 (7) -8 7-8 10-10 10-11 1-1 1-1 3-3 3-3 545 214 139 51 46337 46338 $ $ 19-19-17 19-19-17 166 163 91 85 7-7 7-7 10-10 10-11 1-1 1-1 3-3 3-3 437 242 117 68 1499 1499 $ S 19-19-17 19-19-17 156 158 90 91 7-7 7-7 10-10 10-10 1-1 79 3-3 1-1 610 2-2 159 300 114 Zoologica : New York Zoological Society [XXVII: 17 however (and Ruthven’s), errans probably would be considered one of the end forms of the artenkreis, paralleling the ordinoides section in its reduced scutellation and broad snout. Thamnophis sirtalis parietalis (Say). A specimen collected by Nelson and Gold- man at Casas Grandes, Chihuahua (USNM No. 46371), is the only specimen of this species known from Mexico. It is a female with 19-19-17 scale rows, 152 ventrals, tail incomplete, supralabials 7-7, infralabials 10-11, preoculars 1-1, postoculars 3-3. The spots in the upper row are fused together, and those in the lower row are confluent with the upper row but not with each other; the spaces between the latter are red. T hamnophis marcianus (Baird & Girard). This species is characterized by having a distinct lateral light stripe confined to the third row of scales; the only other having a similar disposition of the lateral stripe is ruthveni. In the latter, however, the mid- dorsal stripe involves only the vertebral scale row; in marcianus it covers the adja- cent halves of the paravertebral rows as well. In addition the ventrals in ruthveni are generally fewer. In body pattern, particularly in the neck region, this species as well as ruthveni shows a much greater similarity to the eques group than it does to megalops. The forty-seven specimens examined are from the following localities: Sonora : (No. 7235) ; Chihuahua : Ojos del Diablo, Santo Domingo Ranch ( No. 30837); 16 leagues north of Guerrero » (No. 46583) ; 10 miles north of Ciudad Delicias (No. 105293); Progreso ( near Casas Grandes) (Nos. 104634-41) ; 5 miles south of Cuidad Juarez Scale Counts Number Sex Scale Rows Ventrals 28653 2 21-21-17 150 5491 2 21-21-17 146 849 2 21-21-17 144 7235 2 21-21-17 153 26135 2 21-21-17 150 30837 2 21-21-17 150 46583 2 21-21-17 148 104635 2 21-21-17 157 104637 2 21-21-17 154 104639 2 21-21-17 156 104640 2 21-21-17 156 104641 2 21-21-17 156 105293 2 21-21-17 155 105295 2 21-21-17 154 15344 $ 21-21-17 151 95183 $ 21-21-17 153 104634 $ 21-21-17 162 104636 $ 21-21-17 165 104638 $ (19)21-21-17 160 (EHT-HMS No. 5418); Rio San Pedro, betw. Chihuahua City and Naica (EHT- HMS Nos. 5319-22, 5419, 5421-3). Du- rango: 5 miles north of Conejos (EHT- HMS No. 5420) ; between Lerdo and La Goma (No. 109295). Coahuila: 2-3 miles east of Torreon (EHT-HMS No. 4996) ; Santa Helena Canyon, Rio Grande (FMNH No. 26135). Nuevo Leon : Mamulique Pass (EHT-HMS No. 5287); Sabinas Hidalgo (EHT-HMS No. 28653); 8 miles W. of Monterrey (EHT-HMS No. 23615). Tamau- lipas: Matamoras (Nos. 861, 5491, 15344); Charco Escondido (No. 1849); Rancho El Plato, 38 miles southeast of Reynosa (No. 95183). Carman (Bull. Essex Inst., vol. 19, 1887, pp. 7-8) records the species from “San Luis Potosi,” but the record appears to be in error. Thamnophis ruthveni Hartweg & Oliver. Similar to marcianus, having the lateral stripe confined to the third scale row, but differing in having a narrow middorsal stripe covering only the vertebral scale row (one and two half rows in marcianus) , and generally by having fewer ventrals (147 to 154 in males, 141 to 150 in females; mar- cianus has 151 to 165 in males, 144 to 159 in females). It is known only from the Pa- cific side of the Isthmus of Tehuantepec, where it has been taken near Tehuantepec, at Chivela, and at San Mateo del Mar, state of Oaxaca. Two specimens examined are from Te- huantepec (No. 110802) and Chivela (No. 46364). Both are males, and respectively have 21-21-17, 21-21-16 scale rows; ven- trals 153 in both; caudals ?, 71 ; supralabials 8-8, infralabials 10-10, preoculars 1-1, post- oculars 4-4, a minute anterior temporal in each. Three other specimens (EHT-HMS Nos. 27562-4) are from Tehuantepec. of marcianus. Caudals Supral. Infral. Proc. Ptoc. 65 8-8 10-10 1-1 4-4 — 8-8 10-10 1-1 3-4 — 8-8 10-? 1-1 4-4 67 8-8 10-10 1-1 4-4 — 8-8 10-11 1-1 3-3 — 8-8 9-10 1-1 4-4 — 8-8 10-10 1-1 4-4 66 8-8 9-10 1-1 3-3 62 8-8 9-9 1-1 3-3 8-8 9-10 1-1 3-3 — 8-8 9-10 1-1 3-4 69 8-8 10-11 1-1 3-4 — 8-8 10-10 1-1 3-4 76 8-8 10-10 1-1 3-4 78 8-8 9-10 1-1 4-4 — 7-8 10-10 1-1 4-4 72 8-8 10-10 1-1 4-4 70 8-8 9-9 1-1 3-4 1942] Smith: Synonymy of Thamnophis 115 Thamnophis macrostemma macro - stemma (Kennicott). As stated previously (Smith, Zool. Ser. Field Mus. Nat. Hist., vol. 24, 1939, pp. 29- 30), this species cannot stand as megalops; if the latter name is to be used, it must be for a subspecies of macrostemma (since it is not based upon a distinct species) for Cope (Bull. U. S. Nat. Mus., No. 1, 1875, p. 41) chose the latter as the name for the species. It seems that macrostemma is composed of one or more populations differing to some extent from each other. Nevertheless I am not certain that the differences previously pointed out by me ( loc . cit.) between north- ern and southern specimens actually differ- entiate two populations, instead of merely expressing a lack of specimens from terri- tories intervening between the two geo- graphical extremes. Specimens from certain localities in such intermediate territory which have been studied more recently do turn out to ue intermediate, and thus is supported a possibility that the variation in ventral and caudal counts exhibits a defi- nite south-north trend independent of the variation in other characters, such as color, that may actually define several true popu- lations in the species. That color character- istics may define geographic races in this species has been emphasized by Ruthven (op. cit.) and more recently by Brumwell (Trans. Kans. Acad. Sci., vol. 42, 1939 [1940], pp. 423-429, pi. 1). The difficulty of analyzing color differences upon the basis of only pi'eserved specimens has made im- ! practical a further study of this aspect of the problem at the present time. However, despite certain doubts in my own mind of the validity of such a separa- tion, the recognition of a northern and a southern race of macrostemma based upon differences in ventral and caudal counts is still statistically sound. In caudal counts of males, 93% of to. macrostemma have 78 or less, while 91% of to. megalops have 79 or more; in females 94% of to. macrostemma have 68 or less, while 95% of to. megalops have 69 or more. In ventral counts of females 85% of w. macrostemma have 185 or less, while 54% of to. megalops have 159 or more; the over- lap in males is practically complete. In total counts of males, 86% of to. macrostemma have 242 or less, while 95% of to. megalops have 243 or more; in females of to. macros- temma, 79% have 225 or less, while in to. megalops 89% have 226 or more. Specimens recently examined add further data for the table of variation previously published. The following distributional records are available. Nayarit : Santa Teresa; Tepic. Jalisco: Atemajac; Chapala; 20 km. south of Guadalajara; Guadalajara; La Quemada; Magdalena; Ocotlan. Michoacan : Lake Cuit- zeo; Patzcuaro; Zamora; Tacicuaro. Mex- ico: Chaleo; Chimalhuacan ; Lerma; San Pedro Tultepec; Toluca; Nevado de Toluca; 7-10 miles west of Villa Victoria. Distrito Federal: Chapultepec; Coyoacan; La Viga Canal; Mexico City; Xochimilco. Puebla: Atlixco; Puebla; Tecamachalco. 4 Veracruz : Acateno; Mirador; Orizaba. Oaxaca: Mitla. The specimens from the vicinity of Lake Chapala and La Quemada are intermediate in character between the two races, and were not included in the comparisons given above. For data on variation in these speci- mens see Brumwell, op. cit. Thamnophis macrostemma megalops (Kennicott) . The following distributional records in Mexico are available. Chihuahua: Chihua- hua; Colonia Juarez; Jimenez; Minaca; San Andres; Rio San Pedro between Naica and Chihuahua City; Progreso, near Casas Grandes (on Rio Santa Maria) ; Casas Grandes; Colonia Garcia; Santa Rosalia; Sierra Madre. Sonora: Santa Magdalena (Tuscon Ariz. ?). Durango: Ada Magda- lena; Coyotes; Durango; El Salto; Rio Tunal, above Pueblito; Lerdo. San Luis Potosi : Hda. La Parada. Guanajuato : Guanajuato; Irapuato; Moro Leon; Sala- manca; Tupataro (this may be one of the towns of the same name in the state of Michoacan); Celaya. Hidalgo: La Vega, Mizquiahuala, Valle de Mezquital (Martin del Campo, Anal. Inst. Biol. Mex., vol. 8, 1937, p. 264). 4 Localities in this state probably incorrect. Variation in Scale Counts of Thamnophis macrostemma. Race Sex Caudals Ventrals Totals No. Range Av. No. Range Av. No. Range Av. Spec. Spec. Spec. to. megalops $ 23 77-89 82 32 159-172 165 20 242-255 248 to. megalops 5 22 68-89 74 30 149-164 157 18 222-260 235 w. macrostemma $ 43 65-81 74 53 129-169 164 42 226-245 237 to. macrostemma ? 33 61-71 66 48 150-171 158 33 210-228 221 116 Zoologica : New York Zoological Society [XXVII: 17 Thamnophis sauritus proximus (Say). Mexican specimens examined since 1938 ( Smith, Occ. Payers Univ. Mich. Mus. Zool., No. 388, 1938, pp. 5-7) are from 60 miles south of Matamoras, Tamaulipas (FMNH No. 27183) ; Hda. La Clementina, Tamau- lipas (No. 105305) ; and Acultzingo, Vera- cruz (Nos. 110803-4). These have 165, 160, 159, 160 ventrals, respectively; caudals 94+ ? ( 3 ) , 102+ ( 3 ) , 97(3), 98(2). The Acultzingo record must represent nearly the extreme southern edge of the range of the subspecies, as at about the same parallel chalceus occurs on the coast. Thamnophis sauritus chalceus (Cope). The range of this subspecies is from cen- tral Veracruz to Costa Rica. To the locali- ties mentioned by Dunn ( Herpetologica , vol. 1, 1940, pp. 192-3) and Smith (op. cit., pp. 5-6, pi. 1) can be added Jonuta (No. 110805) and Montecristo (No. 46584), Ta- basco; Puerto Morelos (No. 46530) and Cozumel Island (No. 13906), Yucatan ; Coba (FMNH No. 26972), Quintana Roo; San Geronimo (FMNH No. 1459), Oaxaca; and Potrero Viejo (EHT-HMS No. 5272), Veracruz. Thamnophis melanogaster melanogaster (Peters) . It seems not to have been generally recog- nized that eastern specimens of this very distinct species are the only ones with a broad, black area on the belly and tail; specimens from the western and northern parts of the range have a narrow, mid- ventral dai'k line frequently scarcely evi- dent. The lateral light stripes are seldom not visible (placed on the second and third scale rows) in eastern specimens, but sel- dom evident in the western. Young, sub- adults and sometimes even large specimens of eastern specimens have a narrow, mid- dorsal light line involving only the vertebral scale row, while western specimens show no middorsal light line in either young or old. Moreover eastern specimens generally (81%) have the second labial in contact with the nasal on both sides of the head (usually the contact is broad), while in western specimens this occurs in but 22%. Eastern specimens also usually (75%) have 3-3 or more postoculars, while western specimens usually have 2-3 or 2-2 (72%). Finally, the caudals in eastern specimens are usually 64 or less in males (84%), 54 or less in females (73%), while in western specimens they usually number 65 or more in males (69%), 55 or more in females (84%). The name melanogaster probably, al- though not certainly, is referable to the eastern specimens. Peters’ description is very brief and states only that a median black band is present on the belly, and that it is less distinct on the tail. The descrip- tion might well apply to western specimens, but such a name as melanogaster hardly fits them, while it is very descriptive of the eastern specimens. Accordingly I restrict Peters’ name to the eastern race, pending a study of the types that will definitely allo- cate the name. Tropidonotus mesomelanus Jan also is based upon the eastern race as shown by the excellent figures on plates 5 and 6, livraison 27, of the Iconographie Generate des Ophidiens. The pigmentation of the tail is less extensive than usual and the anal plate is unmarked in one of the figures (pi. 6), but I have seen a specimen from Lake Xochimilco which duplicates the pattern shown in the figure so exactly that I have no doubt in referring the specimen depicted to the eastern race. On plate 5 another speci- men, this from the Vienna Museum, is fig- ured; it is a very typical specimen of the eastern race, and is here designated lecto- type of Jan’s mesomelanus. A final name synonymous with melano- gaster is Tropidonotus Baronis Mulleri Troschel. This is synonymized with the typical race of melanogaster for several rea- sons. In the first place the name is not in binomial form, since the species name con- sists of two words. The name, thus written, cannot be accepted; had a hyphen been used, thus making one word of it, the form would satisfy requirements of binomial nomen- clature. Boulenger, in fact, referring the name to the synonymy of melanogaster (Cat. Snakes Brit. Mus., vol. 1, 1893, p. 226) hyphenated the name, thus giving it nomen- clatorial status. His description is based upon specimens of both races of melano- gaster; I restrict it to eastern specimens, referable to the typical race. Troschel’s description of T. Baronis > Mulleri, moreover, is not certainly indenti- | fiable to subspecies; the specimens are, of course, definitely melanogaster, but whether eastern or western is not made evident. By inference, since a comparison is made with T. grahamii and the belly is stated to have a midventral dark streak like that species, Troschel’s specimens may have been west- ern, for the belly streak in grahamii, when present, is quite narrow as in western melanogaster. The identification at present however cannot be definite ; fortunately this is not imperative since Troschel’s name is not available. Accordingly no name proposed in the past appears to be available for the widely-dis- tributed western race. It is named and de- scribed in the following. If, however, at some future date it develops that Peters’ 1942] Smith: Synonymy of Thamnophis 117 type of melanogaster is based upon western specimens, then his name must take pre- cedence over the one proposed here, and mesomelanus Jan will be available for the eastern race. Unfortunately it is impossible at present to offer a more assuredly perma- nent nomenclature for these snakes. In 31 specimens, 25 have the second labial in contact with the nasal on both sides, 4 on one side, 2 on neither side. In 44 specimens, the postoculars are 2-2 in six, 2-3 in five, 3-3 in thirty, 3-4 in three. The ventrals are 144 to 158 in males, average 148.3, in fe- males 140 to 153, average 145.2. The caudals are 58 to 68 in males, avei’age 62.3, in fe- males, 50 to 61, average 53.5. Specimens examined of Thamnophis me- lanogaster melanogaster are from the fol- lowing localities. Distrito Federal: Mexico City (Nos. 12726, 12729) ; Lake Xochimilco (EHT-HMS No. 5063). Mexico: Chaleo (FMNH Nos. 983, 1099, 2038); Chimal- huacan (Nos. 110793-8); Lerma (EHT- HMS Nos. 5044-62, 5075-8, 5076A, 15942-7 ) . The only record in the literature for a lo- cality not represented by specimens ex- amined and probably referable to m. melano- gaster, is Jicaltepec, Veracruz. These records indicate a range from cen- tral Mexico (state) to western central Vera- cruz; it possibly occurs also in northern Puebla and southern Hidalgo. Thamnophis melanogaster canescens subsp. nov. Holotype. EHT-HMS No. 5023, male, from Lake Chapala at Chapala, Jalisco, collected July 2, 1935, by H. M. Smith. Paratypes. Sixty-four, including EHT-HMS Nos. 4896, 4921, 4923-32, 4928A, 4934, 4936, 4939-41, 4942A, 4943, 4967-73, 4970A, 5020-2, 5024- 43, all topotypes, same date and collector; Scale Counts in melanogaster melanogaster. Museum Number Sex Scale Rows Ventrals Caudals Supral. Infral. Proc. Ptoc. FMNH 983 5 19-19-17 145 61 8-8 10-10 1-1 3-4 “ 2038 9 “ 153 55 8-8 10-10 1-1 3-4 “ 1099 9 “ 153 57 8-8 10-10 1-1 3-3 tt 1099 9 “ 147 — 8-8 9-10 1-2 2-3 USNM 12726 9 a 150 51 8-8 10-10 2-3 3-3 “ 12729 9 a 150 59 8-8 9-10 2-2 3-3 “ 110794 9 tt 150 — 8-8 10-10 1-2 2-2 “ 110797 9 “ 150 53 8-8 10-11 2-2 2-2 EHT-HMS 5044 9 “ 143 53 8-8 10-10 2-3 2-2 tt 5045 9 “ 143 51 8-8 10-10 2-2 3-3 tt 5046 9 “ 141 51 8-8 10-11 2-2 3-3 tt 5047 9 tt 141 52 8-8 10-11 2-2 3-3 tt 5048 9 a 142 53 8-8 9-10 2-3 3-3 tt 5049 9 “ 144 54 8-8 10-10 2-2 3-3 tt 5051 9 a 143 52 8-8 9-10 2-2 3-3 a 5052 9 a 141 55 8-8 9-10 2-2 3-3 it 5057 9 n 143 52 8-8 10-10 2-2 2-3 it 5058 9 tt 141 54 8-8 10-10 2-2 2-2 tt 5060 9 “ 143 52 8-8 10-10 2-2 3-3 n 5063 9 tt 145 57 8-8 10-10 2-2 2-2 tt 15943 9 u 140 51 8-8 10-11 2-2 3-3 tt 15944 9 tt 146 52 8-8 10-10 2-3 3-3 a 15945 9 “ 142 — 8-8 10-10 2-2 2-3 tt 15946 9 u 146 53 8-8 10-10 2-2 3-4 tt 15947 9 tt 147 50 8-8 11-11 2-2 3-3 FMNH 1099 8 “ 155 67 8-8 10-10 1-1 3-3 USNM 110793 8 tt 158 68 8-8 10-10 2-2 3-3 tt 110795 8 “ 152 68 8-8 9-10 2-2 2-2 “ 110796 8 n 154 59 8-8 10-10 3-3 3-3 tt 110798 8 “ 152 64 7-7 10-10 2-2 3-3 EHT-HMS 5050 8 “ 150 59 8-8 10-10 2-2 2-3 tt 5053 8 tt 145 61 8-8 9-10 2-2 3-3 tt 5054 8 tt 144 60 8-8 10-11 2-2 3-3 it 5055 8 “ 147 62 8-8 10-10 2-3 3-3 tt 5056 8 “ 144 64 8-8 10-10 2-2 3-3 tt 5059 8 a 147 64 8-8 9-10 2-3 3-3 tt 5061 8 tt 149 64 7-8 10-10 1-2 3-3 tt 5062 8 tt 145 61 8-8 9-10 2-2 3-3 a 5075 8 a 147 63 8-8 10-10 2-2 3-3 it 5076 8 tt 147 61 8-8 10-10 2-2 3-3 u 5077 8 tt 146 58 8-8 10-10 2-2 2-3 tt 5078 8 “ 146 62 8-8 10-10 2-2 3-3 tt 5076A 8 tt 145 59 8-8 10-10 2-2 3-3 tt 15942 8 tt 146 59 8-8 10-10 2-2 3-3 118 Zoologica : New York Zoological Society [XXVII: 17 EHT-HMS No. 5019, Magdalena, Jalisco; EHT-HMS Nos. 5064-5, Lake Cuitzeo, Michoacan; EHT-HMS Nos. 5066-71, three miles east of Celaya, Guanajuato; FMNH No. 1529(2), Ocotlan, Jalisco; USNM No 110800, Tacicuaro, Michoacan; USNM No. 110799, La Palma, Michoacan. Diagnosis. Like Thamnophis melanog aster melanog aster, except: belly with scattered spots, immaculate or, usually, a narrow, some- times broken longitudinal midventral dark line ; lateral light stripes rarely evident ; sub- caudal surface unmarked or with a narrow, irregular median streak; never a middorsal light line; second labial generally (78%) separated from nasal on one or both sides of head; postoculars generally (72%) 2-2 or 2-3; subcaudal scales generally (67%) 65 or more in males, 55 or more in females (84%). Description of Holotype. Head narrow, somewhat pointed, lores somewhat flaring; length of rostral visible from above two- thirds length of suture between internasals; latter equal to length of suture between prefrontals; frontal pentagonal, anterior edge straight, posterior edges forming a right angle, sides somewhat concave, very slightly narrower posteriorly than anteri- orly; length of frontal subequal to length of suture between parietals, slightly less than distance from frontal to tip of snout; sutures between nasals and rostral three- eighths width of posterior margin of rostral; nasal completely divided, anterior section subequal in size to posterior; loreal large, rectangular, longer than broad; preocular single on one side, but a lower part partly divided by two incomplete sutures; pre- oculars two on other side, upper much the larger; postoculars 2-3; temporals 1-2-2. Supralabials 8-8, second narrowly in con- tact with nasal on one side, narrowly sepa- rated on the other; seventh labial largest, fourth and fifth entering orbit; infralabials 10-10, five in contact with anterior chin- shields, two (fifth and sixth) with posterior chinshields; latter a little longer and broader than anterior, separated from each other throughout their length, divergent posteriorly. Dorsal scales in 19-19-17 rows, all except those in outer two rows keeled and truncate; Scale Counts in melanogaster canescens (Males). Museum Number Scale Rows Ventrals Caudals Supral. Infral. Proc. Ptoc. FMNH 1384 19-19-17 146 65 8-8 10-10 2-3 2-3 a 1384 it 149 67 8-8 10-10 1-1 3-3 “ 1500 tt 155 75 8-8 10-10 1-1 3-3 EHT-HMS *4896 ii 147 59 8-8 10-10 2-3 3-3 it *4921 a 149 60 8-8 10-10 2-2 2-2 a f4923 it 146 63 8-8 10-10 2-3 2-3 a $4924 it 147 64 8-8 10-10 2-2 3-3 tt f4926 tt 147 67 8-8 10-10 2-2 3-3 tt f4927 tt 146 65 8-8 10-10 2-2 3-3 a $4928 tt 148 63 8-8 10-10 2-2 2-3 tt $4931 tt 147 64 8-8 10-10 2-2 3-4 tt *4934 ti 148 65 8-8 10-10 2-2 2-3 a *4936 19-19-15 149 70 8-8 10-10 2-2 3-3 a *4939 19-19-17 148 67 8-8 10-10 2-2 2-2 it *4940 tt 149 70 8-8 10-10 2-2 2-3 it *4941$ tt 147 67 8-8 10-10 2-2 2-2 tt 4942A a 145 63 8-8 10-10 2-2 2-2 a 4943 tt 147 67 8-8 10-10 §1-1 2-2 tt 4970 n 148 — 8-8 10-10 2-2 3-3 a 4970A tt 148 — 8-8 10-11 2-2 2-2 n 4973 it 149 72 8-8 10-10 2-2 2-3 n 5023 a 148 69 8-8 10-10 1-2 2-3 tt 5024 a 148 68 8-8 10-10 §1-2 2-2 a 5027 a 145 69 8-8 10-10 2-2 2-2 it 5028 it 143 67 8-8 10-10 2-2 2-2 a 5029 n 144 65 7-7 10-10 2-2 2-2 a 5030 it 146 68 8-8 10-10 2-2 2-2 a 5032 a 148 67 8-8 10-11 2-2 2-3 a 5034 tt 146 63 8-8 10-10 2-2 2-3 a 5066 tt 151 71 8-8 10-10 2-3 2-2 tt 5071 a 152 73 8-8 9-9 2-2 2-3 USNM 23985 tt 144 — 8-8 10-10 3-3 3-3 it 23986 tt 149 — 8-8 10-10 2-2 3-3 it 23987 tt 145 61 8-8 10-? 2-2 2-3 it 46412 tt 154 69 8-8 10-10 2-3 3-3 a 46413 tt 149 70 7-7 9-10 2-2 2-2 n 46414 tt 150 71 8-8 10-10 2-2 3-3 * Young of No. 5038 t Young of No. 5021 t Anal divided § Loreal entering orbit. 1942] Smith: Synonymy of Thamnophis 119 scales in anal region with a central en- largement on the keel; ventrals 148; caudals 69; total length 525 mm., tail 124 mm.; male. Entire dorsal surface slate (scales shed) ; when scales are spread, a narrow, inter- rupted, pure white line on the skin and ad- jacent edges of the third and fourth scale rows is evident; it is distinct and almost continuous anteriorly, but becomes less dis- tinct posteriorly and is not evident on pos- terior fourth of body ; a series of small black spots placed in the middle of this line at every second or third scale ; these spots vis- ible nearly to tail ; a somewhat similar series of spots and light streaks, but much less evident, between the seventh and eighth scale rows; no dorsal or typical lateral light stripes (the light stripe described above is not the primary light stripe usually referred to in Thamnophis; it is an accessory mark- ing). Belly and tail light slate, chin cream; near the middle and posterior part of belly a very narrow, broken, longitudinal mid- ventral streak is discernible, but nowhere distinct; no dark marks on subcaudal sur- face; anal plate cream (not slate). Specimens Examined. In addition to the 63 specimens of the type series, I have ex- amined 16 others, including FMNH No. 1384(6), USNM Nos. 23985-9, Durango, and FMNH No. 1500, Coyotes, Durango; USNM Nos. 46411-4, La Parada, San Luis Potosi. Records in the literature referable to this race are from Guadalajara, Nevado de Colima, La Laguna and Colonia Brizuela, Jalisco; Tupataro and Guanajuato, Guana- juato; and Tangancicuaro, Michoacan. Variation. In 59 specimens, 13 have the second labial in contact with the nasal on Scale Counts in melanogaster canescens (Females). Museum Number Scale Rows Ventrals Caudals Supral. Infral. Proc. Ptoc. FMNH 1384 19-19-17 145 63 8-8 10-10 1-1 3-3 ii 1384 ii 146 61 8-8 10-10 2-2 3-3 a 1384 a 138 53 8-8 10-10 2-2 2-2 a 1384 a 143 55 9-9 10-10 2-2 2-2 a 1529 a 145 57 8-8 10-10 1-1 2-2 a 1529 a 143 57 8-8 10-10 2-2 2-2 EHT-HMS f4925 a 142 59 8-8 10-10 2-2 2-2 ii f4928A a 141 59 8-8 10-10 2-2 2-3 a f4929 a 140 60 8-8 10-10 2-2 3-3 a 44930 a 145 55 8-8 10-10 2-2 2-3 u 44932 a 142 56 8-8 10-10 2-2 2-2 a 4969 u 145 57 8-8 10-10 2-2 2-2 ii 4971 a 143 60 8-8 10-10 2-2 2-2 a 4972 a 145 59 8-8 10-10 2-2 3-3 a 5019 ii 137 56 8-8 10-10 2-2 3-3 a 5020 ii 142 55 8-8 9-10 2-2 2-2 a 5021 ii 146 — 8-8 10-10 2-2 3-3 a 5022 a 138 54 8-8 10-10 2-2 2-2 a 5025 a 139 60 8-8 10-10 2-2 2-2 u 5026 a 144 59 8-8 10-10 2-2 2-3 a 5031 a 138 55 8-8 10-11 2-2 2-2 a 5033 a 145 55 8-8 10-10 1-2 2-3 a 5035 a 139 55 8-8 10-10 1-2 2-2 a 5036 19-19-16 137 55 8-8 9-10 2-2 2-2 a 5037 19-19-17 141 51 8-8 10-10 2-2 2-2 a 5038 “ 141 56 8-8 10-10 2-2 2-2 a 5039 ii 143 54 8-8 10-10 2-2 2-2 a 5040 “ 140 58 8-8 10-10 2-2 2-2 u 5041 19-19-16 147 58 8-8 10-10 1-2 2-3 a 5042 19-19-17 140 — 8-8 10-10 2-2 2-2 a 5043 “ 143 55 7-8 10-10 1-2 2-2 a 5064 17-19-17 140 60 8-8 10-10 2-2 2-2 a 5065 19-19-17 141 — 8-8 10-10 2-2 2-2 a 5067 “ 147 61 8-8 10-11 2-2 2-3 a 5068 “ 147 58 8-8 10-10 2-2 2-2 a 5069 a 143 57 8-8 10-10 2-2 2-2 a 5070 “ 146 57 8-8 10-10 2-2 2-3 a 4984A a 139 — 8-8 9-10 2-2 3-3 a 4984B a 137 55 8-8 10-10 2-2 3-3 USNM 23988 a 140 52 8-8 10-10 1-2 2-2 ic 23989 a 142 55 8-? 11-? 2-? 1-1 u 46411 a 147 — 8-8 10-10 2-3 3-3 110799 19-19-16 143 — 8-8 10-10 2-2 2-2 110800 19-19-17 145 49 — — — — t Young of No. 5021 120 Zoologica : New York Zoological Society [XXVII: 17 both sides, 12 on one side, 34 on neither side. In 79 specimens, the postoculars are 2-2 in 39, 2-3 in 18, 3-3 in 21, 3-4 in one. The ventrals vary from 143 to 155 in males, average 148.3; in females they are 137 to 147, average 142.3. The caudals are 59 to 73 in males, average 66.4; in females they are 49 to 63, average 56.6. Other details of the variation in scutellation are given on the accompanying table. The largest specimen examined is a fe- male (No. 5069) measuring 751 mm. in total length, the tail 155 mm. The largest male (No. 4973) measures 622 mm. in total length, the tail 146 mm. In color there is considerable variation. There are two notable patterns. One, the more common, is exemplified by the holo- type. In this type there is no very promi- nent dorsal pattern; most clearly evident is a lateral series of small spots in the third and fourth scale rows. A female of this type (No. 5038) contained seven young (Nos. 4896, 4921, 4934, 4936, 4939-41), all with the same pattern with the exception of one that is completely melanistic (there is also one completely black adult, No. 5022). In the second pattern type, the spots in the two rows described in the type are much enlarged, although not with sharply de- fined or regular edges; the spots do not meet; the belly may have irregularly-placed spots near the midventral line, two or three on each ventral scale. One female (No. 5021) of this type contained eleven young, and all are marked dorsally like the mother, with very prominent spots ; some have scat- tered dark spots on the belly, others have a midventral line or no marks whatever. The constancy of pattern in the young suggests that there may be more than one form involved. There are five adults (Nos. 4969, 4973, 5021, 5030-31) that definitely belong to the distinctly-spotted group, 35 that belong to the nearly unicolor group, and one that is more or less intermediate (No. 5034). Since the spotted specimens form such a small proportion of the adult population, it might be expected that, if the variant were of a normal type, there would be an equally small proportion of the young with such a pattern. That all the young of one female is the same, the rarer, type is distinctly unusual. In no specimen is a middorsal light stripe present. Some show slight evidence of a lateral light stripe anteriorly, on the second and third scale rows. Remarks. There seems to be little doubt of the distinctness of the two races of me- lanogaster . In greater doubt is the alloca- tion of the names, and the identity of the spotted specimens from Chapala. The west- ern i-ace, m. canescens, ranges from eastern Guanajuato and Michoacan and southern San Luis Potosi westward to the Nevado de Colima and central Nayarit, northward to central Durango. It is possible that the Du- rango specimens may be distinguishable from the others, for 3-3 postoculars and 1-1 or 1-2 preoculars frequently occur in them. Specimens from that area have not been i'e- examined during the study of this species and are thus not included as paratypes. The Tepic specimens are typical, non-spotted. The Celaya specimens show some evidence of approach to m. melanogaster, as the mid- ventral stripe is somewhat broader than in others of m. canescens. Thamnophis rufipunctatus (Cope). The type of angustirostris (No. 959) has the appearance of a hybrid between melano- gaster and the species which has been known since 1908 (Ruthven. op.cit., pp. 120- 124) as angustirostris (= rufipunctatus). That it cannot be identified with the species long known by that name has already been noted by Taylor (Taylor & Knobloch, Proc. Biol. Soc. Wash., vol. 53, pp. 129-130, 1940). The essential differences between these two species may be contrasted as follows : rufipunctatus Dorsal pattern Upper row of spots distinct Ventral pattern Irregularly mottled Head pattern (juv.) Preocular Labials enter eye Dorsal scales Ventrals Caudals Range Labia strongly barred, head mottled Never one One 21-21 (23)-17 9 152-160 (9 counts) 9 65-68 (5 counts) Durango to Arizona melanogaster No upper row of spots, lower row poorly defined Nearly all black, or a midventral dark line Labia indistinctly barred, head not mottled Occasionally one Two 19-19(17)-17 9 137-153 (69 counts) 9 49-63 (60 counts) San Luis Potosi Mexico City to Durango and 1942] Smith: Synonymy of Thamnophis 121 The type of angustirostris has a very dis- tinct body pattern, with well defined spots in the upper row as well as the lower; the labia are somewhat more strongly barred than in typical melanogaster ; and the dorsal scales are in 21-21-17 rows. In these char- acters it is more nearly like rufipunctatus. It also has a nearly completely black belly, head nearly uniform except labial marks, one preocular, two labials entering orbit, and 149 ventrals. In these characters it is like melanogaster. In caudal count the type is exactly intermediate between the two species, having 64 (female). Since the type is from a locality intermediate between the known ranges of melanogaster and rufipunc- tatus there can be little question that it may be considered a hybrid.5 It may be an inter- grade ( i.e ., of a fertile hybrid population), but the fact that both rufipunctatus and melanogaster are distinguishable in one lo- cality (Coyotes, Durango, FMNH) indicates that the hybrids, if and when produced, are not fertile. Accordingly angustirostris gives no evidence in support of a view that rufipunctatus and melanogaster are sub- species. The proper name for the northern form is open to question. In general the proced- ure is to let stand for hybrids those names based upon them, and accordingly angusti- 5 The possibility that it represents a distinct species still remains, however ; only further specimens from the area represented by the type will show conclusively which interpretation is correct. rostris cannot be applied to either of the parent species. The next earliest name avail- able is rufipunctatus, which also unfortu- nately is based upon an atypical (partial albino?) juvenile. The type (No. 8600) now lacks all dorsal markings although Cope de- scribed it with red spots. In belly pattern it is typical. The head pattern cannot be dis- cerned. The scutellation is typical, with the exception that two labials enter the orbit (a rare occurrence, but seen in other speci- mens). There are 173 ventrals; 86 caudals ( S ) ; 8-9 supralabials; and 21-21-17 scale rows. It is not an anomalous Natrix valida, which has 139 to 154 ventrals (Van Den- burgh, 1922). A later name, Atomarchus multimaculatus Cope, is without question referable to this species (c/. Taylor & Knob- loch, loc. cit.) Twenty-six Mexican specimens examined are from the following localities. Chihua- hua: Colonia Juarez (FMNH No. 1096) ; Ahumada (FMNH No. 1259); San Andres (FMNH No. 1275a); Chihuahua (Nos. 14254, 14261, 14265, 14271, 14275, 14286, 14288); Rio Casas Grandes (No. 2659); Meadow Valley (No. 26592); Arroyo del Alamos, 70-74 kilometers south of Nueva Casas Grandes (Nos. 42874-5); Guadelupe y Calvo (No. 46368) ; Rio Papagochic, Guer- rero (No. 95607); Progreso (Nos. 104658- 61), Mojarachic (EHT-HMS No. 23015). Durango : Coyotes (FMNH No. 1501-2); Guanacevi (No. 46369). Sonora? (No. 21055). Scale Counts in rufipunctatus. Museum Number Sex Scale Rows Ventrals Caudals Supral. Infral. Proc. Ptoc. Enter Labials Eye EHT-HMS 23015 $ 21-21-17 156 64 + 9-9 12-12 3-3 3-4 i FMNH 1259 5 21-21-17 158 — 8-8 10-11 3-3 3-4 i USNM 14275 2 21-21-17 152 68 8-9 10-11 3-3 3-4 i “ 14288 2 21-21-17 160 67 9-9 10-11 3-3 3-4 i “ 26591 2 21-21-17 155 — 8-8 10-10 2-2 3-3 i U 26592 2 21-23-17 160 65 8-9 10-11 3-3 4-4 i U 104658 2 21-21-17 154 — 8-9 10-11 2-3 2-? i “ 104659 2 21-21-17 158 — 8-8 10-10 2-2 3-3 i FMNH 1501 2 21-21-17 156 68 8-8 10-10 2-2 3-3 2 “ 1096 $ 21-21-17 162 79 8-9 10-10 2-2 4-4 1 << 1275a $ 21-23-17 163 78 8-8 9-9 3-3 3-4 1 USNM 14254 $ 21-21-17 161 — 8-8 10-10 2-3 2-3 1 H 14261 $ 21-21-17 162 76 8-? 10-? 2-3 3-3 1 << 14261 $ 19-21-17 158 — 8-8 10-10 3-3 3-4 1 14265 $ 21-21-17 160 79 8-8 10-10 2-2 3-3 1 U 14271 $ 21-21-17 164 80 8-? 11-? 3-3 3-4 1 u 14286 $ 19-21-17 159 82 8-9 11-11 3-3 3-4 1 u 42874 $ 21-21-17 163 — 8-8 10-10 2-2 3-4 1 il 42875 $ 21-21-17 165 — 7-8 10-10 2-2 3-3 1 u 46368 $ 21-21-17 174 76 8-8 10-10 2-3 3-3 1 (( 46369 $ 21-21-17 177 78 8-9 9-10 2-3 3-4 1 u 95607 $ 21-21-17 166 77 8-8 10-10 2-3 3-3 1 u 104660 $ 21-21-17 159 — 8-8 10-11 2-2 3-3 1 FMNH 104661 $ 21-21-17 162 76 8-8 9-10 2-4 2-4 1 1502 $ 21-21-17 165 8-8 10-11 2-2 3-3 1 USNM 21055 $ 21-21-17 164 79 8-8 10-10 2-2 3-3 1 122 Zoologica: New York Zoological Society [XXVII: 17 Key to Mainland Mexican and Central American Thamnophis. 1. Lateral stripe involving third and fourth scale rows on anterior part of body ... 2 Lateral stripe not involving fourth row, indistinct, absent, or involving only the second or third row or both 5 2. Scale rows 21 at middle or on anterior third of body; anterior edges of ventrals black 3 Scale rows 19, maximum (except immedi- ately behind head) 4 3. Caudals 78 or less in males, 68 or less in females . . .macrostemma macrostemma Caudals 79 or more in males, 69 or more in females macrostemma megalops 4. Caudals 93 to 104, ventrals 155 to 169, totals 255 to 262 in females; 102 to 109, 158 to 172, and 260 to 277, respectively, in males sauritus proximus Caudals 83 to 96, ventrals 142 to 159, totals 229 to 254 in females; 92 to 99, 149 to 158, 245 to 256, respectively, in males sauritus chalceus 5. Scale rows 21 or more on anterior third of body 6 Scale rows less than 21 10 6. Preoculars two or more ; usually one labial entering orbit rufipunctatus Preocular single; usually two labials en- tering orbit 7 7. Usually anterior edges of ventrals black, and the entire belly clouded, but never a longitudinal dark line or entire belly black; chin and sometimes subcaudal surface cream, sharply differentiated from dark belly color 3 Not so; belly unmarked or with a longi- tudinal dark line, or nearly entirely black, or with scattered dark spots; chin and subcaudal surfaces not sharply dif- ferentiated from dark color of belly. . .8 8. Belly almost completely black (or perhaps a midventral black stripe) ; lateral light stripe not restricted to third row an- teriorly angustirostris Belly not with a median, more or less ex- tensive black stripe ; lateral stripe, when evident, restricted to third row anteri- orly 9 9. Vertebral light stripe one and two half scale rows wide, at least at base of tail marcianus Vertebral light stripe one scale row wide, or nearly indistinguishable, even at base of tail ruthveni 10. Preoculars two or more, rarely one; belly usually with a continuous, longitudinal black area on center; scale rows 19 on anterior third of body 11 Preoculars single; belly not with a con- tinuous black area (or if so, scale rows no more than 17) 12 11. Black on belly and tail very extensive, covering nearly all of ventral surface; vertebral and lateral light stripes evi- dent in all except large specimens; sec- ond labial generally (81%) in contact with rostral; postoculars usually (75%) 3-3 or more; caudals usually 64 or less (84%) in males, 54 or less (73%) in females . . . .melanog aster melanogaster Black on belly and tail entirely absent or restricted to a midventral line or a few scattered spots; rarely light lines evi- ; dent, never the middorsal ; second labial | usually not (78%) in contact with ros- tral; postoculars usually (72%) 2-2 or 2-3; caudals usually over 64 (69%) in males, 55 (84%) in females melanogaster canescens 12. Maximum dorsal scale rows 17, and no evidence whatever of a middorsal stripe on any part of body . . . chrysocephalus i Dorsal scale rows more than 17, or if only 17, a vertebral light stripe present or indicated on some part of body 13 13. A moderately dark area in middle of each parietal, sometimes fused with nuchal blotches, darker than most of the re- mainder of the dorsal surface of head (light) 14 No distinct dark mark on center of pari- etal ; head nearly uniform light or dark above 15 14. Large spots present on body, usually a single series extending completely across back phenax phenax Spots on body (except on neck) poorly defined or absent .... phenax halophilus 15. The spots in the two rows on each side, between the vertebral and lateral stripes, fused together over most of body (not just on neck 17 The spots in the two rows on each side, be- tween the vertebral and lateral stripes, not fused together except on neck; spots disappearing posteriorly or not 16 16. Scale rows anteriorly 17° 17 Scale rows anteriorly 19 19 17. Dark color of dorsal surface of head and nape with a clearly defined, dark-edged indentation a little posterior to angle of mouth, extending anterodorsally toward midoccipital region; scale rows usually 19 on some part of body; spots usually distinct on most of body 18 Dark color of dorsal surface of head and nape lacking any clearly defined, dark- edged light indentation behind angle of mouth; scale rows seldom over 17 on any part of body; no spots distinct on body posterior to nape, dorsum nearly uniform brown scalaris godmani 18. Scale rows posteriorly usually 17 (94%) ; anterior scale rows usually 19 (88%) scalaris scaliger Posterior scale rows usually less than 17 (70%) ; anterior scale rows usually less than 19 (93%) scalaris scalaris 19. Spots in upper row of the two series be- tween vertebral and lateral light lines fused together and usually with the spots in the outer row, the latter spots remaining distinct from each other; ground color red between spots of outer row (above lateral stripe) ; middorsal stripe well-defined .... sirtalis parietalis Spots not fused in such a manner, al- though sometimes poorly defined or cor- responding spots of the two rows fused together; ground color not red; mid- dorsal stripe distinct or not 20 6 Some care must be used to determine whether the scale rows are reduced to 17 at any point on the anterior part of body. 1942] Smith: Synonymy of Thamnophis 123 20. Middorsal stripe with continuous, straight edges, covering one and two half scale rows; rounded dark spots on anterior edges of ventrals on sides of at least part of belly; ventrals seldom over 157, caudals seldom over 76 21 Middorsal stripe usually covering only the vertebral row, or indistinct, or absent; if broader, not straight edged and ven- trals and caudals more numerous; belly not spotted 22 21. Supralabials strongly barred; head dark above, the color fused with that of nape; ventrals 136 to 140 sumichrasti praeocularis Supralabials not or weakly barred ; head light above, its color sharply differenti- ated from that of nape; ventrals 144 to 167 sumichrasti cerebrosus 22. No median light stripe on any part of body, its place occupied by a median series of dark spots 27 A median light stripe at least anteriorly; no median series of dark spots 23 23. A very distinct median light stripe throughout length of body; and dark spots present on the scales below the lateral light line at least anteriorly, visible without spreading the scales; ventrals seldom less than 149 24 Median light stripe indistinct or absent posteriorly; or, if distinct posteriorly, no dark spots on the scales below the lateral light line 25 24. Ventrals 167 or more in males, 163 or more in females eques cyrtopsis Ventrals 166 or less in males, 162 or less in females eques eques 25. Ventrals 155 to 166 . . . . ordinoides errans Ventrals fewer, 136 to 153 26 26. Median light stripe scarcely distinguish- able on any part of body, and nowhere more than one scale row wide eques postremus Median light stripe covering one and one half to three scale rows anteriorly, but disappearing completely on posterior part of body sumichrasti fulvus 27. Ventrals 139 to 157 in males, in females 139 to 147; caudals 58 to 72 sumichrasti sumichrasti Ventrals 155 to 162 in males, 149 to 156 in females; caudals 78 to 89. . . .vicinus 1942] Funkhouser : Membracidae from British Guiana 125 18. Membracidae (Homoptera) from British Guiana.1 W. D. Funkhouser University of Kentucky. [This is one of a series of papers dealing with the collections made at Kartabo, Brit- ish Guiana, at the Field Station of the De- partment of Tropical Research of the New York Zoological Society, under the direction of Dr. William Beebe. During the eight years of the maintenance of the station, ex- tensive collections and ecological studies were made within an area of jungle one- quarter of a mile square, which may be called the Guiana Jungle-zone. For details concerning this area of intensive study, see Zoologica, Vol. VI, No. 1, 1925. Miss Maud D. Haviland reported on the first collection of Membracidae from Kartabo in Zoologica, Vol. VI, No. 3, 1925. The pres- ent paper is the result of Dr. W. D. Funk- houser’s study of a second collection.] Introduction. Through the courtesy of Dr. William Beebe, the writer has had the privilege of examining a very interesting lot of Mem- bracidae collected chiefly in British Guiana. In fact, Dr. Beebe states that practically all of the specimens were taken within a quar- ter of a square mile of jungle at Kartabo. This small area must certainly have a rich insect fauna since it has yielded nearly fifty different species of one small family. The collection is particularly valuable, however, in its contribution to our knowl- edge of the geographical distribution of the Membracidae since a very large number of the listed species represent new locality rec- ords. Included in the collection are two spe- cies taken at Guanoco, Venezuela, and one in Trinidad. The species represented in the collection are here reported as follows: 1. Membracis fusca DeGeer. DeGeer, Ins., Vol. Ill: p. 208, No. 10, Tab. 32, fig. 14. (1773). Eight specimens from Kartabo and six from Bartica. A large foliaceous bizarre form, originally described from Peru but is 1 Contribution No. 637, Department of Tropical Re- search, New York Zoological Society. common throughout South America. It has been previously reported from Kartabo by Miss Maud Haviland (Mrs. H. H. Brindley) in 1925. 2. Membracis foliata Linn. var. c-album Fairm. Linnaeus, Syst. Nat., Vol. II: p. 705, No. 2. (1767). Fairmaire, Revue Memb., p. 244, No. 4. (1846). Ten specimens from Kartabo. Another of the well known foliaceous forms with a wide distribution in South and Central America, its known range extending from Brazil across the northern part of South America and through Centi'al America to Mexico. Fairmaire described the variety c-album from British Guiana. 3. Membracis provittata Buckt. Buckton, Mon. Memb., p. 42, PI. 3, figs. 6, 6a. (1903). Twenty-two specimens from Kartabo and one from Bartica. This species has been con- sidered a synonym of tectigera but is now known to be distinct. Buckton described it from Surinam and we have a long series in our collection from British Guiana. These are the only known localities, for the species is apparently rare and is seldom seen in collections. 4. Membracis fasciata Fabr. Fabricius, Syst. Ent., Vol. II: p. 2092, No. 54. (1767). One specimen from Kartabo. A striking species very common in Brazil and with a rather confused taxonomic history since it has been redescribed under a large number of synonyms. Miss Haviland reported it from British Guiana in 1925. 5. Membracis arcuata DeGeer. DeGeer, Ins., Vol. Ill: p. 206, No. 9, Tab. 32, fig. 10. (1773). One specimen from Kartabo. Common throughout South and Central America and previously reported from British Guiana by Miss Haviland. 126 Zoologica : New York Zoological Society [XXVII: 18 6. Enchenopa lanceolata Stoll. Stoll, Cic., Tab. 27, fig. 166. (1780). Four specimens from Kartabo and ten from Bartica. One of the commonest of the neotropical membracids. Reported from all of the northern countries of South America and across Panama into Central America. Previously reported from British Guiana by Miss Haviland. 7. Enchenopa serratipes Buckt. Buckton, Mon. Memb., p. 49, PI. 5, fig. 9. (1903). Two specimens from Kartabo. This spe- cies has often been confused with E. albi- dorsa Fairm., but is distinct. E. albidorsa is abundant throughout South America but we have undoubted records of serratipes only from Brazil, British Guiana and Colom- bia. Miss Haviland reported this species as albidorsa from Kartabo. 8. Campylenchia hastata Fabr. Fabricius, Mant. Ins., Vol. II: p. 263, No. 9. (1787). Two specimens from Kartabo. A widely distributed species, showing some variation in size and color in different regions but found in one phase or another in most parts of South America, Central America and Mexico. One of the common forms of this species was described as nutans and was so reported from British Guiana by Miss Haviland. 9. Leioscyta rufidorsa Godg. Goding, Journ. N. Y. Ent. Soc., Vol. XXXVI: p. 37. (1928). Two species from Kartabo. Goding de- scribed this species from Ecuador and it has not been reported in the literature from any other country but we have specimens in our collection from Peru, Colombia and Brazil. This is the first record from Brtiish Guiana. 10. Erechtia punctipes Buckt. Buckton, Mon. Memb., p. 53, PI. 7, figs. 5-5b. (1903). One specimen from Kartabo. This is ap- parently a very rare insect and is seldom seen in collections. Buckton gave only “South America” as the locality and it was a long time before it was finally recognized. Finally Goding identified it from Brazil and the writer secured a good series from Peru. This British Guiana specimen gives us a new locality record. 11. Erechtia brunneidorsata Funkh. Funkhouser, Can. Ent., Vol. XLVI: p. 357, PI. 24, fig. 1. (1914). Three specimens from Kartabo; one from Bartica. Previously known only from Peru but the British Guiana specimens agree en- tirely with the type material. A new locality record. 12. Pterygia uropigii Buckt. Buckton, Mon. Memb., p. 72, PI. 12, fig. 3. (1903). One specimen from Kartabo. A bizarre form described from Brazil and reported from British Guiana by Miss Haviland in 1925. The writer has a few specimens from Peru. These are the only locality records known. 13. Pterygia cerviceps Fowl. Fowler, B. C. A., p. 24, No. 3, Tab. 2, figs. 12, 12a. (1894). Two specimens from Kartabo. The center of distribution of this species is certainly Central America where it is quite common but we have seen specimens from Colombia and from Venezuela so it apparently ranges down into South America. This is the first record from British Guiana. 14. Umbonia spinosa Fabr. Fabricius, Syst. Ent., p. 675, No. 4. (1775). Eighteen specimens from Kartabo. One of the largest, commonest, most widely dis- tributed and most easily recognized of the neotropical Membracidae. Reported from al- most every country in South America and Central America. Previously reported from Kartabo by Miss Haviland. 15. Aconophora marginata Walk. Walker, List. Horn. B. M., p. 540, No. 16. (1851). Eleven specimens from Kartabo. This species is common in Mexico and Central America and has been reported from Pan- ama, Brazil, Ecuador, Peru and Bolivia but this is the first record from British Guiana. 16. Aconophora pallescens Stal. Stal, Hem. Fabr., Vol. II: p. 35, No. 12. (1869). Two specimens from Kartabo. Another widely distributed species, closely related to the preceding, and reported from the same general regions but never before recorded from British Guiana. 17. Aconophora fusiformis Fowl. Fowler, B. C. A., p. 69, No. 19. Tab. 5, figs. 18, 18a, 19. (1895). One specimen from Kartabo. This is an- other Central American form which is only rarely reported from South America. The writer has material from Brazil and Bolivia but the Kartabo specimen gives us the first record from British Guiana. 18. Aconophora projecta Funkh. Funkhouser, Journ. N. Y. Ent. Soc., Vol. XXV: No. 2, p. 160. (1927). One specimen from Kartabo. Since this species has been recorded from Ecuador, Bolivia, Colombia and Brazil, it would be ex- pected to occur in British Guiana but has 1942] Funkhouser : Membracidae from British Guiana 127 not been previously reported from that country. Another new locality record. 19. Darnis partita Walk. Walker, his. Saund., p. 75. (1858). Five specimens from Kartabo and one from Bartica. When Miss Haviland reported this species from Kartabo in 1925 it was the first record from South America but since then we have secured specimens from Peru, Ecuador, Panama, Canal Zone and Colombia. It would therefore seem that its range extends along the northern coast of South America. 20. Darnis latior Fowl. Fowler, B. C. A., p. 52, No. 2, Tab. 4, figs. 16, 16a. (1894). One specimen from Kartabo. This is an- other new locality record but the species has been found in Peru, Ecuador and Colom- bia so its distribution is probably about the same as that of partita. 21. Stictopelta fraterna Butler. Butler, Cist. Ent., Vol. II: p. 340, No. 9. (1878). One specimen from Kartabo and one from West Bank. Again a new locality record. It was described from Peru and both Coding and Fowler reported it from Mexico. We have specimens in our collection from Co- lombia and Bolivia. These localities with this British Guiana record represent its present known distribution. 22. Nassunia binotata Fairm. Fairmaire, Revue Memb., p. 291, No. 3. (1846). Three specimens from Kartabo. A very rare species, almost never seen in collec- tions and previously known only from Brazil. A new locality record. 23. Heteronotus tridens Burm. Burmeister, Rev. Silb., Vol. I: p. 229. (1833). Seven specimens from Kartabo. The larg- est species of its genus, very grotesque in appearance, strikingly marked and very well known. Common in northern South America but this is the first time it has been re- ported from British Guiana. Another new locality record. 24. Heteronotus spinosus Lap. Laporte, Ann. Soc. Ent. France, Vol. I: p. 96, PI. 3, fig. 7. (1832). Four specimens from Kai'tabo. Closely related to the preceding and showing consid- erable variation so that it has been de- scribed under several synonyms. Its distri- bution seems to be about the same as that of H. tridens of which it may be merely a variety. Miss Haviland reported it from British Guiana under the name Heniconotus armatus. 25. Cymbomorpha dorsata Fairm. Fairmaire, Rev. Memb., p. 293, No. 6. (1846). One specimen from Kartabo and one from Guanoco, Venezuela. Both the British Gui- ana and the Venezuela records are new. It has previously been reported only from Brazil, Peru and Costa Rica. 26. Horiola picta Coq. Coquebert, III. Ic., p. 78, Tab. 18, fig. 10. (1801). One specimen from Trinidad. The com- monest species of its genus, found in most parts of South America, Central America and the West Indies but not previously re- corded in the literature from Trinidad so that this also is a new locality record. 27. Horiola lineolata Fairm. Fairmaire, Rev. Memb., p. 492, No. 2. (1846). Three specimens from Kartabo. Closely related to the preceding species and with about the same range. It is commonly found in Brazil and Colombia but has not been previously reported from British Guiana. 28. Ceresa sallai Stal. Stal, Hem. Mex., p. 70, No. 421. (1864). A good series of thirty-one specimens from Kartabo. This species was described from Mexico where it is abundant and it is common in Central America. It is reported only rarely from South America, the only known records being from Brazil and Bo- livia. This is the first record from British Guiana. 29. Ceresa vitidus minor Fowl. Fowler, B. C. A., p. 103. (1894). Five specimens from Kartabo. This is the small variety of C. vitulus Fabr. which seems to occur with vitulus wherever vitulus is found. C. vitulus was reported from Brit- ish Guiana by Miss Haviland but this is the first record of the variety from that coun- try. It is found throughout South America. 30. Micrutalis pattens Fowl. Fowler, B. C. A., p. 118, No. 4. (1895). One specimen from Kartabo. This is a common species in Mexico and Central America but has not been reported in the literature of the family from South Amer- ica. However we have two specimens in our collection from Venezuela so it is not sur- prising to see it from British Guiana. This is a new record from British Guiana and the first published record from South America. 31. Boethoos reticulata Fabr. Fabricius, Syst. Rhyng., p. 29. (1803). Two specimens from Kartabo. Apparently abundant in Brazil, Colombia and Vene- zuela and previously reported from British Guiana by Miss Haviland. 128 Zoologica : New York Zoological Society [XXVII: 18 32. Vanduzea triguttata Burm. Burmeister, Rev. Silb., Vol. IV : p. 183, No. 4. (1836). Two specimens from Kartabo. The domi- nant species of its genus in southwestern United States, Mexico and Central America. The South American forms which are found in Colombia and Venezuela are slightly dif- ferent in minor respects from the northern ones but we believe that they are all of the same species. This is the first record from British Guiana. 33. Stylocentrus ancora Perty. Perty, Del. Anim., p. 179, PI. 35, fig. 15. (1834). One specimen from Kartabo. Very abun- dant in Brazil and recorded from Ecuador, Colombia and Venezuela. This is the first record from British Guiana. 34. Cyphonia clavata Fabr. Fabricius, Mant. Ins., Vol. II: p. 264, No. 17. (1878). Eight specimens from Kartabo and one from West Bank. One of the commonest membracids in South America and found in all of the northern and central areas of that continent. Reported from British Guiana by Miss Haviland in 1925. 35. Cyphonia proxima Guer. Guerin, Ic. Reg. Anim. Ins., p. 365, Tab. 59, fig. 3. (1838). Two specimens from Kartabo. A less abundant species than the preceding but with a range extending from Mexico through Central America and Panama to northern South America. This, however, is the first record from British Guiana. 36. Antianthe expansa Germ. Germar, Rev. Silb., Vol. Ill: p. 245, No. 1. (1835). Three specimens from Guanoco, Vene- zuela. One of the best known and most widely distributed of the American Mem- bracidae. It has been reported from prac- tically every country in South and Central America and extends northward into Flor- ida, California and Arizona. 37. Amastris antica Germ. Germar, Mag. Ent., Vol. IV: p. 16. (1821). Two specimens from Kartabo. A very rare species previously known only from Brazil and Peru. It is quite distinct from all other species of the genus and is easily recognized. A new locality record. 38. Amastris obtegens Fabr. Fabricius, Syst. Rhyng., Vol. II: p. 25. (1803). Seven specimens from Kartabo. The best known and probably the most abundant of the species of the genus but not previously reported from British Guiana. The speci- mens most often seen in collections are from Brazil, Colombia, Ecuador and Peru. A new locality record. 39. Amastris compacta Walk. Walker, List Horn. B. M. Suppl., p. 140. (1858). Two specimens on one pin from Kartabo. A well known and rather widely distributed species, fairly common in Brazil, Peru, Ecuador and Colombia but this is the first record from British Guiana. 40. Amastris minuta Funkh. Funkhouser, Journ. N. Y. Ent. Soc., Vol. XXX: No. 1, p. 30, fig. 6. (1922). One specimen from Kartabo. This species has never been mentioned in the literature of the family since its original description and so far as we know this is the second specimen ever to be found since the type was a single specimen from Peru. Of course this is a new locality record. 41. Hille exaltata Walk. Walker, List Horn. B. M. Suppl., p. 140. (1858). Two specimens from Kartabo. This spe- cies has been reported in the literature only from Brazil but we have specimens in our collection from Peru and Colombia. This is the first record from British Guiana. 42. Hille herbicola Hav. Haviland, Zoologica, Vol. VI: No. 3, p. 255, PI. 4, figs. 1, la. (1925). Three specimens from Kartabo. This spe- cies was described from Kartabo and has never been reported from any other locality. 43. Tynelia hirsuta Funkh. Funkhouser, Journ. N. Y. Ent. Soc., Vol. XXX: No. 1, p. 25, PL 3, fig. 1. (1922). One specimen from Kartabo. Described from Peru and we have seen material from Nicaragua. This is the first record from British Guiana. 44. Rhexia pallescens Fabr. Fabricius, Syst. Rhyng., Vol. II: p. 28, No. 8. (1803). One specimen from Kartabo. A species which has been described under a number of different specific names in at least five differ- ent genera. It is quite common throughout northern South America and shows consid- erable variation in color. It seems to be most abundant in Brazil, Surinam, Colombia and Ecuador. This is a new record from British Guiana. 45. Stegaspis insignis Buckt. Buckton, Mon. Memb., p. 59, figs. 6, 6a. (1903). Three specimens from Kartabo. One of the peculiar “dried leaf” forms which al- 1942] FunJcliouser : Membracidae from British Guiana 129 ways attract attention and are often seen in collections. Most of the material which we have seen has been from Brazil, Peru, Ecuador and Colombia. This is the first rec- ord for British Guiana. 46. Stegaspis laevipennis Fairm. Fairmaire, Rev. Memb., p. 527, No. 14. (1846). Four specimens from Kartabo. Another of the brown, foliaceous species of this genus with about the same distribution as the preceding. It was reported from British Guiana by Miss Haviland in 1925. 47. Lycoderes hippocampus Fabr. Fabricius, Syst. Rhyng., p. 20, No. 22. (1803). One specimen from Kartabo. Apparently a very rare species. It was reported from British Guiana in 1925 by Miss Haviland but was known previously only from Brazil and has never been recorded from any other countries. 48. Bocydium globulare Fabr. Fabricius, Syst. Rhyng., p. 16, No. 3. (1803). Two specimens from Kartabo. A species often figured to illustrate a most bizarre type of pronotal development, especially be- cause of the globules on the bar above the head. However it is not common. It was de- scribed from Brazil and we have specimens from Peru. It has not been previously re- corded from British Guiana. 49. Tolania opponens Walk. Walker, List Horn. B. M. Suppl., p. 159. (1858). Three specimens from Guanoco, Vene- zuela. One of the few membracids which completely lack a posterior process. Very abundant throughout South America, Cen- tral America and Mexico. 1942] Breder: Evolutionary Hypotheses 131 19. A Consideration of Evolutionary Hypotheses in Reference to the Origin of Life. C. M. Breder, Jr. New York Zoological Society (Text-figure 1). Introduction. The living organisms of today are thought by most biologists to have evolved from more or less dissimilar ancestors. In fact, since the time of Charles Darwin biologists have addressed themselves, in the main, to establishing evidence to support or demolish various propositions concerning the mode of evolution, as set up by him or others who followed him to a greater or lesser extent. It seems not to have occurred to many such students that the importance to be attached to interpretations of the results of their studies must vary with our ideas concerning the origin of life. Since we are in no position to make any postulates, it may be useful to consider the various ideas that have arisen from time to time in reference to their bear- ing on the acceptability of any particular evolutionary hypothesis. We may go along with Lotka (1925) in his policy of resignation concerning a defi- nition of life without in any way encounter- ing difficulty in an attempt to discuss the possibilities of origin. Many of the attempts to explain the origin of life have come from non-scientific groups and most of these may be dismissed with a few words. All are in- cluded in this discussion since their many interlocking ideas make it essential at least to mention each one. The author is grateful to Dr. Richard T. Cox of New York University for checking the physical parts of the discussion, to Prof. Albert E. Parr of the American Museum of Natural History, for helpful criticism, to Dr. G. E. Hutchinson of Yale University, for valuable suggestions, and to Mr. J. W. Atz for editorial assistance. Concepts of the Origin of Life. Although superficially numerous, see Woodruff (1936), all the ideas that have been put forth concerning the possible origin of life are reducible to two basic pat- terns. These may be discussed separately for the purpose of considering their tenable contents in reference to the adequate inter- pretation of evolutionary experiments and arguments. 1. Planted life forms. The idea that life may have been planted on the earth from some exterior source has always been at- tractive, but today it seems less reasonable than ever before. The idea divides into two parts. a. Normal cosmic behavior. One idea has been that simple, very resistant particles drifting through space from other planets or similar bodies may revive to an active life when happening to land on a body of suitable environment. Arrhenius (1908, 1911) by in- genious reasoning even went so far as to suggest that thermophilic bacteria rained on the earth from possibly Venus, being im- pelled by the radiant energy of the Sun. This is mentioned in this connection merely to indicate that such possibilities have been given consideration by thoroughly serious persons. Certainly modern experiments on the effects of low pressure and temperatures have shown nothing tending to make this an impossibility as has been indicated, for in- stance, by Goertz (1928). Venus itself as a source of living substance would seem un- likely on the basis of the view of Wildt (1940) who argues for a negligible amount of water and an atmosphere of formalde- hyde. See also Jeans (1942). Such a condi- tion would not seem likely as a life source, in spite of the fact that certain molds and dipterous larvae appear to thrive on rela- tively low concentrations of formaldehyde in aqueous solution. Since Venus evidently rotates, although its speed has not been de- termined, it would appear that its nearness to the Sun, surface temperatures as high as 60 °C having been recorded, would also stand in the way of the continued existence of protoplasm at any place on its surface. The effects of unfiltered solar or other radi- 132 Zoologica : New York Zoological Society [XXVII: 19 ation in space, moreover, may represent a very real hazard. Oparin (1938) on this basis presents a very convincing argument against the possibility of survival under such radiation. In effect, this idea would have inter- stellar space occupied by viable motes pass- ing from place to place, along with the mat- ter and radiations already known to be there. Such a condition would provide for a polyphyletic origin of terrestrial life, if evo- lutionary capabilities be ascribed to such organisms. Lipman (1932) thought he had succeeded in reviving bacteria from the in- terior of stony meteorites. Oparin (1938) dismisses this study by expression of the belief that contamination was responsible. Actually, while such may have been the case, Oparin gives no foundation for so assum- ing, other than his well-reasoned argument against organisms surviving the radiations of interstellar space. He wrote, “The or- ganisms which he [Lipmanl succeeded in iso- lating were identical with the bacterial forms existing on the Earth. This makes it very pi-obable that, in spite of all his pre- cautions, Lipman did not succeed in pre- venting earth bacteria from contaminating the meteorites while they were ground to a powder. Even in different regions of our planet there are different forms of micro- organisms, and it would be extremely strange if exactly the same bacterial forms found on the Earth were present also on some remote planets.” Exception is taken to this last remark, for if Lipman is right, such results, on the contrary, are exactly what one should expect. Since meteorites are continually falling to earth, one then should also expect a continual seeding of forms capable of surviving such transit. These then would be expected to be common on the Earth in proportion to its ability to support them in active life. This, indeed, also might well lead to a spotty distribution dependent on their needs. Beutner (1938b) also re- jects such a possibility but submits no fur- ther closely reasoned arguments. Jones (1940) treats the subject in a similar fashion. b. Special events. The planting of or- ganisms on earth, on the other hand, may be thought of as a special event, more or less unique in cosmic history. Any- thing that could be conjured up to induce such an event would be classified here. The creation of life by decree would come here and would be practically the equivalent of the planting of a culture by some space- traveling, superior animal. Space travel is one of the most recent serious aspirations of man and as yet is seriously handicapped, to say the least, on technological grounds. If extra-terrestrial beings have developed such an art, we certainly have no evidence of it, but the fact that there have been no visits to earth establishes nothing. It might even be imagined that life on earth is the remnant of some ill-fated expedition from another body from which only food animals survived. The possibilities of this sort are limited only by one’s imagination and have been well covered by the writers of scien- tific fiction, and need not longer detain us here. Other special events in the nature of a cosmic accident happening but once could make a single planting. Such an accident would provide for a monophyletic origin or a polyphyletic one to the extent of the va- riety of organisms in this single seeding. One thinks of the cluster of asteroides be- tween Mars and Jupiter that supposedly represents a disrupted planet of consider- able size. Its disintegration must have scat- tered material far and wide. Fantastic as these notions may sound, there is but one other basic pattern of source that has been suggested which, as will be seen later, also has its fantastic side. ; The idea that living entities were created by fiat is a widespread one among theologists’ mysticisms, which, with minor variations, is the common property of many theologies. Such ideas could belong in the category of special events, if not mere figments of wish- ful thinking. 2. Spontaneous generation. The long- discredited belief in spontaneous genera- tion, experimentally unsupported, is never- theless the idea which most scientific men seem to think best accounts for the origin of life. By the simple expedient of pushing the event far enough back into geologic time to prevent any kind of experimental ap- proach and postulating an unknown but suitable environment, many seem to have eased their minds. This interpretation of life origin likewise divides into two parts i parallel to those of item number one. a. Chemical evolution. Given a sphere;! such as the Earth, chemists, physicists, as- i tronomers and geologists have shown that a i chain of events must follow due to the inter- 1 action of forces involved. Radioactive degra- dation, the salt concentration of the ocean, : 1 and so on, come to mind, which it is un- necessary to discuss at length here. See, for example, Clarke (1924), Fairchild (1938) ! and Jones (1940). The point of this is that! chemical evolution proceeds systematically 1 according to the second law of thermody- I namics and is very definitely identifiable as1 1 a kind of inorganic orthogenesis. Thus we I have a world stage in which the inorganioj props and scenery are changing by an ortho- genetic or “built in” process. Oparin (1938) ! devotes nearly an entire book to visualizing* how such chemical changes took place. He gives by far the most careful analysis and the most plausible picture of a cooling 1942] Breder: Evolutionary Hypotheses 133 Earth. Whether all his details are sound and whether events lead automatically to living entities or not, it is hard to deny the basic soundness of his chemical evolution. Beut- ner (1938) arrives at a very similar conclu- sion, differing chiefly in the order of chemi- cal events but lacking the close reasoning of Oparin. Riddle (1939) sketches his similar views with extreme brevity. Jones (1940) gives a presentation of the theoretical and observational data of astronomy, indicating what is known and fairly inferential of the conditions on other planets in reference to the basic needs of life. If the idea is valid that in this process there automatically comes a time and a con- catenation of events which grade from the strictly inorganic to the organic and the sentient, then we have life beginning as of necessity as a part of a general “ortho- genesis.” As such an event would hardly oc- cur simultaneously throughout the world, presumably there would be started various orthogenetic series, the first advanced in de- velopment as compared with the later. This should go on until the period of its possi- bility is passed by the general orthogenesis of the whole system, repetition becoming impossible when the stage is sufficiently changed. This point would be passed pre- sumably when there were no longer any large sterile areas. Oparin believes that early sterility is absolutely essential for a life origin of this sort. The pre-living or- ganic components would simply be broken down on earth today, long before they reached anywhere near the living state, by the activities of living beings, according to his views. Lichtig (1938), on the other hand, disre- gards this point and supposes the transition from lifeless to living matter to be taking place more or less continuously, indicating a widespread polyphylogenesis. Herrera (1942) proposes an interesting and startling theory which would derive life from the sublimation of volcanic emana- tions, after years of close study of the life- resembling behavior and physical appear- ances of a tremendous number of chemical substances. As he indicates, this, too, would imply a continual synthesis of life, unless . some peculiar but necessary factor was pres- ent but once in the Earth’s history. b. Rare accident. If, on the other hand, we do not consider the origin of life as defi- nitely pai't of the general increasing en- tropy, then we may consider it as an “acci- dental” event repeatable as often as the needful circumstances repeat. This, of course, is pure, spontaneous generation in the simple sense, and no doubt, when re- ferred to known physical law, apparently leads such students as Smith (1932) to re- fer somewhat poetically to life as “an eddy in the second law of thermodynamics.” See Pike (1939) for a recent discussion of vari- ous views on the subject. Both these views may lead to polyphyleti- cism, the first definitely orthogenetic and the second not necessarily so by virtue of questioning the full and simple operation of the second law in living systems. If, as thought by various students, the origin of life occurred by unique “accident” another element is intruded. This view arose and gained force by the failure of Pasteur et al. to produce life by simple means, and the large evidence that life, as we know it, is produced only by pre-exist- ing parents. The rare accident view is peril- ously close to the fiat creation of theologians, differing largely in use of words, especially since it is doubtful if any sect ever believed that the creation of human beings was an “accident.” This rare accident view should make for a monophyletic origin of life and leave the way open for any evolutionary process what- ever. Judging from the literature of evolu- tion, it would seem that most workers in that field imply that they are interpreting their results on the assumption of an origin of this sort. Either the rare accident or the straight chemical evolution could lead at first, at least, only to an orthogenesis. Paired species, parallel specializations, and parallel mutations certainly suggest at least a “charge” of similar potentialities as a re- current phase in any group of sufficiently numerous organisms. These views are con- ditioned by whether one considers life as a highly probable or improbable result of the operation of statistical mechanics. 3. Interaction of methods. Of these sev- eral views, it will seen that evolution could conceivably proceed by any method in all but one. That one, as a part of general chemical evolution, is definitely linked with orthogenesis. All those items in which life would arise by the ordinary working out of natural processes (la and 2a) would be ex- pected to lead to a polyphyletic condition. A special planting accident (lb) could be either mono- or polyphyletic, while a rare accident of spontaneous generation (2b) would be monophyletic. This interaction is especially pertinent in the case of chemical evolution (2a) and natural seeding from outer space (la). Since sterility is perhaps necessary for the first, the second presum- ably could not successfully be brought into play. If extensive seeding did occur early enough, it could, on this basis, check chemical development of the basic organic compounds. Before proceeding to a consideration of some current biological ideas in reference to the preceding, it may be best to discuss cer- tain theoretical considerations basic to the establishment of life systems. 134 Zoologica : New York Zoological Society [XXVII: 19 Non-protoplasmic Systems. Active life processes as we know them are remarkably limited in their temperature range. It is evident that it must be below the coagulating point of essential proteins (these may range from about 35° to 70°C), and above the freezing point (scarcely be- low 0°C). Many various forms are able to survive protracted periods far below this, but are in a state of suspended activity until the temperature is raised again. See, for ex- ample, Goertz (1938). Warm-blooded ani- mals may, of course, be active below this temperature of environment, but their life processes are going on at a higher tempera- ture by virtue of their very pretty trick of operating an internal heating plant. Since the chief solvent of body fluids is water which exists in the liquid phase only between 100° and 0°, here are essentially the broad limits, 100° never being reached because protein coagulates at a lower point and 0° being slightly passed on the down side because of the lowered freezing point of the mildly saline body fluids. These remarks, naturally, refer to surface pressures. Actually many aquatic organisms exist at much greater pressures, as, for ex- ample, the abyssal fish of great depths with pressures of tons per square inch. Here tem- peratures may be such as would cause the fluids to change phase at surface pressures, but this is merely an expression of the pres- sure-temperature relationship. Extremely light pressures, on the other hand, are lim- ited by the greatly lowered boiling point, re- sulting in a rapid vaporization and loss of fluids. Actually, in nature, active life is lim- ited in the higher atmosphere by low tem- perature and apparently by low oxygen con- tent, but considering what low concentra- tions of available 02 some aquatic animals thrive on, this in itself might not be insur- mountable. However, unlike the combination of high pressure and low temperature, the combinations of low pressure and low tem- perature, with the resultant change of phase of H20, at a fairly low level of chemical ac- tivity may be enough to check active life. Nevertheless, extreme conditions in this di- rection have been survived for short per- iods. Insects have been placed under the in- fluence of highly efficient vacuum pumps and rapidly bi'ought to a vacuum comparable to that of interstellar space and lived to sur- vive an immediate return to normal surface conditions, Lutz (1929). Obviously this could only be a transient phenomena because the great moisture loss would quickly result in death. It is evident, however, that while terres- trial animals well cover the gamut of tem- perature range in their normal activities, most live much closer to their minimum pressure threshold than to their maximum. The truth of this is constantly attested by the pressures under which caisson workers survive and the relatively slight altitudes that foi'ce aviators to don masks or pressure suits. Anaerobic organisms are all of a small size and are so presumably of necessity be- cause of the low combustion rate to which their particularly limited metabolism is restricted. Thus, life, as we know it, is very sharply restricted in a number of directions; by temperature, the effects incident to pres- sure, and oxygen, either freely supplied or broken out of the substrate. After these comes a host of others going to make up the milieu necessary for the survival of any par- ticular form. All such life is concerned with a single though slightly variable gel-like compound — protoplasm, or as Beutner (1938) expresses it, “life, Carbon’s out- standing property.” The actual basic auto- catalytic activity, as is pointed out by Alex- ander (1939), is properly only referable to “. . . the simplest self duplicating units — chromosomes and their constituent genes, mitochondria and possible subcellular sym- bionts.” If there is analogous activity going on or possible in other systems of chemical and physical combination, we are not cogni- zant of such. Baldwin ( 1937 ) writes in italics . . that the conditions under which cell life is possible are very restricted indeed and have not changed substantially since life first began.” With the extensive knowl- edge of life forms now available, it would seem that the terrestrial existence of non- protoplasmic life forms is exceedingly un- likely. Henderson (1913) discusses at length the unique position of carbon in this regard and Jones (1940) makes much of the tetra- valence of carbon atoms. If we consider such matters in a broader way, the question naturally arises as to the possibility of equivalent activity in physical systems covering quite other ranges of tem- perature. Obviously, what we call life is either an unique phenomenon of a very tiny i temperature range, or it is not. Many doubt- I less would consider it mere foolishness to speculate on other possibilities. However, a consideration of various features of it, at the very least, makes it possible for us to ob- tain a better understanding of the peculiari- ties of life activities in the range in which we know it exists. As a starting point, for a basic require- ment, there must be solids, liquids and gases present in some specific temperature range if the inhabitants are going to be suf- ficiently like known life forms to be called equivalent or analogous. Thus the higher one goes on the temperature scale, the fewer solids and liquids and the more gases there will be found until that point is reached 1942] Breder: Evolutionary Hypotheses 135 where only a state of highly excited ions exist— a state of complete incandescence. Descending, more and more solids are found and less and less liquids and gases, until absolute zero is reached where no chemical activity is possible. If the melting and boiling points of the elements are plotted, they form a regularly ascending series from absolute zero. If the phases of the elements are expressed in terms of percentages, a chart, as in Text-fig- ure 1, may be constructed. It will be noted that our life range is very close to absolute zero as compared with the whole scale. These remarks consider only the elements. A study of the possible compounds, which would be of great importance, is a task be- fore which a corps of physicists and chem- ists might well quail. However, it would ap- pear that the various compounds possible at the different temperature levels, because of the statistical nature of the known and ex- pected possible combinations of elements, would all go to produce curves not dissimilar in form to those obtained for the elements as shown in Text-figure 1. In connection with a consideration of the distribution of phases represented in vari- ous temperature ranges, the relative abun- dance of the separate elements enters as an item of considerable importance. Since oxy- gen, nitrogen and hydrogen occur on Earth in considerable quantities, it is possible for large amounts of water and an atmosphere comprised mostly of oxygen and nitrogen to be present when the suitable temperature range and other features have appeared. It would, on the other hand, be impossible to have an atmosphere in which large quanti- ties of krypton were present. If an organic system were to be based on an environment of some of the less abundant elements it would be seriously restricted. Two condi- tions can be conceived. One would be such that the important materials were widely spread but highly dilute, placing the need of great effort and consequent strain on any energy exchange system so based and in this manner limiting it sharply. The other would be that the needful materials were concen- trated at some focal point and there be rela- tively abundant. This would restrict such a system on a geographical or spatial basis. Until some hypothetical conditions at an- DEGREES ABSOLUTE Text-fig. 1. Phases of the elements at temperatures from 0° to 4200°. The tempera- tures, expressed in terms of absolute degrees, are grouped in classes of 100° each and the index figures read as the higher limit. Thus 100 should be read as from 0° to 100° absolute. The phases are expressed in terms of percent, of the total number of elements. Due to certain chemical features such as sublimation and the fact that complete data are not available for every element, there are some minor difficulties in the construction of such a chart, but these in no way destroy the basic nature of the curves as shown. See text for an explanation of “life range” as here indicated. 136 Zoological New York Zoological Society [XXVII: 19 other temperature range have been postu- lated in some detail, to which are to be fitted some equally hypothetical creatures, it is pointless to pursue the item further. The studies of Goldschmidt (1937) on the known distribution of elements in the Earth’s crust would seem to be fundamental to any such consideration. See also Russell (1941). While it is sufficiently obvious from the above that the quantities of individual ele- ments present would have a limiting influ- ence on energy exchange systems of various hypothetical kinds, there could be conceived very definite mechanisms. Although such basic building blocks to known life as nitro- gen, carbon, oxygen and hydrogen are suf- ficiently abundant, it should be remembered that many of the less common elements are needful to the present life on Earth. Iodine, for example, an important element in the living processes of many forms, is not an abundant substance. Furthermore, accord- ing to Clarke (1924), the only elements that are represented by more than one percent, of the terrestrial elements are oxygen, sili- con, aluminum, iron, magnesium, calcium, sodium and potassium, in the order named. For the first listed he gives 49.2%, descend- ing to the last with 2.4%. All are found in organisms but most are not present in a massive sense. On the other hand, carbon in this list shows only 0.08% and nitrogen, 0.03%. It is thus clear that in the present life system some of the basic materials are comparatively rare. The discussion of Gold- schmidt (1937) on the mechanics of con- centrating influences on the rarer elements is especially pertinent in this connection. See also Lotka (1925) for further discussion of the relative abundance of elements, their distribution and especially their availability, concentration and circulation. Thus far we have not mentioned that other important element in the maintenance of living systems — radiant energy. It is hard to conceive of a hypothetical system without a primary or secondary source of some radiant energy as a prime mover. Since astronomers have given us a great amount of spectrographic data on radiant energy, not only in the Milky Way but from far off nebulae, for our purposes, at least, it is safe to speculate on the gamut of wave lengths sprayed out into the universe. Their relative similarity is perhaps their most striking feature. The effects on other sys- tems, for example, the transparency and florescence of various substances to differ- ing wave lengths and related matters, is too complex to be entered upon casually and would not be sufficiently significant at this point. With these items in mind, we can make a few tentative propositions of what life- systems, if they are to be analogous, would require at any temperature-pressure range. 1. The lower the temperature the greater the variety of solids available for body- building, while the higher the fewer and more limited these would be. 2. Inverse to this would be the variety of fluids and gases for the seas, atmos- pheres, and body fluids, with an increase in kinds and complexities with increased temperatures. 3. The range of the temperature-pres- sure relationship would have to be such as not to change the phase of the body struc- tures except in a manner analogous to that for protoplasmic life. 4. Radiant energy would have to be such as to maintain a source of prime moving force and the systems would have to be able to utilize it directly or indirectly without self-destruction. 5. Some system of energy exchange by continuous chemical activity would have to be possible in any milieu imagined. 6. With these elements the origin of life forms would be presented with the same basic problems already discussed for life on Earth. Speed of development and evolution (chemical activity) would be slower in each successive lower temperature range. Thus the length of time required for each cooler period would have to be progressively longer for complex entities to develop. It should be clear that transfers from one temperature level to another, in an evolu- tionary sense, could hardly be conceivable. In connection with this, it is apparent that protoplasmic life would be nearly the last, or the last, of such a hypothetical series. In other words, our temperature range is close to the lowest at which it is easy to conceive of such activity (see Text-figure 1). This is for two reasons: (1) so many substances are locked up as solids, limiting the possible liquid and gaseous environ- ments, and (2) the level of energy is so close to complete entropy that the activity is necessarily of a low grade in a purely chemi- cal sense. Stated another way, the foregoing should make it clear that life is either a unique event not far preceding a full heat death, or it is near the end of a succession of simi- lar phenomena that occur in a series of focal points as non-connected modes along a descending temperature scale. It may be em- phasized that each range of such activity would have no bearing on the next lower one, any more than the previous chemical combination of a substance bears on any further combinations it may enter as im- pelled under physical changes. Since the Earth supposedly existed at a temperature too high for protoplasmic life prior to its present condition, the question naturally arises as to the existence of a 1942] Breder: Evolutionary Hypotheses 137 previous similar activity at a slightly higher temperature, and if such may have existed, what are the chances of finding some evi- dence of it. Since we have no idea as to what range such a thing might have occupied, for sake of illustration we might take the range of 500° to 600° absolute. This is equal in span of temperature to our life range, but not far above it — perhaps not far enough to be fair to the thesis. It has been selected at random for purposes of illustration but, nevertheless, an examination of the condi- tion of the elements is of interest. Instead of about seventy-eight percent, of the elements being solids, about sixty percent, would be solids. There would be about twenty-seven percent, liquids as against our fifteen per- cent. and gases would be about thirteen per- cent. as against our seven percent. For ex- ample, the following solids in our life range would be liquids: selenium, lithium and sulphur, while iodine and bromine would al- ways be gases. Various elements between the range of 500° and 600° would change phase as various points were passed. For example, cadmium, thallium and bismuth would melt and freeze while certain forms of phosphorous would boil and condense. Lead would occur as a liquid in very hot regions, for it melts at 600.4°, and in the presence of other substances (fluxes) below that value. In a world so conceived, what chance of fossil survival would its “organic” remains stand on a drop to present temperatures? What is known of igneous rocks, their crys- talline structures, consolidation and meta- morphosis would certainly seem to preclude the survival of any structural entity frozen to our relative frigidity. Our earliest fossil- bearing rocks are so clearly of present tem- perature ranges that we simply cannot look hopefully to such sources for evidence. The great variety of carbon compounds known to man and the theoretical possibility of stupendous numbers of others leads Beut- ner (1938b) to consider life a peculiar prop- erty reached by the proper combinations of that element. Silicon has been frequently suggested as a conceivable alternate, within our life temperature range, largely because of its comparable ability to build a large number of compounds. This has been most recently discussed by Jones (1940). Silicon, it may be noted, is not so involved in energy exchange systems of which we have knowl- edge. While not being disposed to debate the unique position of carbon, we may submit that a mere multiplicity of “building blocks” does not necessarily mean that an endless variety of structures will actually be built nor that a comparatively few types of build- ing blocks preclude construction. Actually there are relatively few of the many known carbon compounds to be found in living en- tities although surely there are many more present not as yet recognized. Beutner himself seems to be largely con- cerned with the ingenious devices of Butschli, Traube, Leduc, Herrera, Jennings and Crile, many of which show simulated lifelike activity not involving organic com- pounds. Although these structures are of ex- treme interest in various connections, it does not seem that they are likely to throw much light on the question of the origin of life, possibly excepting the work of Herrera. Since living entities are clearly controlled by the same physical conditions that control non-living units, it should not be surprising that various arrangements can be made of non-living substances that react in ways similar to some of those observed in living entities. The elaborateness of some of these reactions is distinctly interesting and should be of aid in explaining behavior, morpho- logical detail and pattern, but certainly not origins. The striking nature of these con- trivances would seem to be limited only by the ingenuity of the experimenter and were it not for their spectacular nature would not be thought of in connection with a dis- cussion of the origin of life any more than would be experiments which show that ani- mal heat is compai'able to inorganic combus- tion or that HC1 in the stomach acts in a manner identical to HC1 in a test tube. The models here under consideration can be seen to move or grow and are comparable to studies in locomotion or growth in a man- ner similar to the above suggested chemical comparisons and probably useful to an equivalent degree. It should also be clear that primarily the right chemical compounds must be obtained for “livingness” to begin. The form or movements come later, and obviously these physical aspects may be ex- tremely various whereas the chemical na- ture of life, as we know it, is limited to one type of chemical system of energy exchange. Rosett (1917), in describing his particu- lar “design” of an artificial osmotic cell, gives an excellent evaluation of the over- emphasis and under-emphasis that has been accorded such studies by various schools of thought. Beutner (1938a) recognizes the separate nature of morphology and “the power of self-reproduction” and writes, “Obviously, then, living organisms, according to our present conception, have two distinct gen- eral characteristics : “1 — The power of chemically transform- ing the material of their environment (or food material) into their own substance; “2 — The ability to develop diverse forms.” The first item is a definition for an auto- catalytic enzyme and the second is clearly not a basic item in the nature of life but is 138 Zoologica : New York Zoological Society [XXVII: 19 a historical statement based on what has ap- parently happened and which condition may be needful for long continued survival. He considers viruses as having no mor- phology when he writes, “A filterable virus possesses the first general property to its fullest extent, but nothing of the second one.” It might be equally well argued that its morphology is on a submicroscopic basis as indeed it would have to be, and that it possesses morphology in the sense that any molecule does, especially since the entire atomic theory is based on structure and ar- rangement (morphology). Evidences of Modern Science. If we examine the evidences of modern science and not the more or less generally accepted speculations based thereon, we have to dig rather deep to obtain anything of real significance. For the theologists’ mysticism there is none at all, except wishful thinking and sen- timentalism. This is not to be taken as a mere derogatory dismissal, for until the idea of extra-sensory perception can be fully disposed of, simple dismissal cannot have the full assurance, formerly possible. However, if such a concept is to be even con- sidered, it would automatically reduce to the consideration (lb) (page 132). For the idea of life as planted, there is present-day evidence of the great resistance of spores and other living objects to ex- treme low temperatures and of their pres- ence in the substratosphere. Coupled with this is the known force of radiant solar en- ergy as a possible propulsive force. How- ever, there is also the known destructive force of unfiltered radiant energy in inter- planetary space, Lewis (1934). Planting as a cosmic accident would thus seem to have nothing whatever any longer to support it. Carbon found in meteorites has been thought perhaps to represent the remains of extra-teri’estrial life forms. The nature of impact of these objects is such as to make such a source exceedingly unlikely. Elemen- tal life forms have been claimed to be re- covered from the interior of stone meteor- ites by Lipman (1932), but the true source of their origin is still undetermined. See Oparin (1938). For the concept of spontaneous genera- tions, several ideas have been brought for- ward. Born in the biological ignorance of early man, to explain his failure to under- stand the appearance of small animals, a mode long disproved, it nevertheless lingers in a modified form as a possibility. The part of chemical and physical changes in a sys- tem of mixed substances gives a point of de- parture on the assumption that the world is a cooling sphere. Under the influence of solar radiation with the Earth acting like a gigantic Soxlett extractor, it is easy to imagine life as a spontaneous event on the reaching of a certain state. Today, with the virus situation as it is, the differentiation between living and non-living seems to be largely academic, Stanley (1937, 1938a and b), Rivers (1939), Hunt (1939) and Martin and Fisher (1942). It is to be emphasized, however, that viruses or even unicellular units may not in the least be primitive, but may just as well, so far as present evidence goes, be derived from the products of living metazoan bodies, Breder (1936). Even if Oparin’s broader views are accepted, viruses would not necessarily have to considered as primitive. However, this may be, whether viruses came first or last, or all the way along the line, there is no really objective line of demarcation. It is not necessary for us in the present discussion to go into a close consideration of the steady state which life represents or the electric fields which surround such units of electro-chemical activity. The work of Northrop & Burr (1937) may be noted, in passing, as indicative of this field of ap- proach to the basic nature of livingness. No matter what restrictions or modifica- tions must be made, the terrestrial water circulation acts basically like a Soxlett ex- > tractor and strongly influences the present chemical composition of the ocean, Lotka (1925). That the blood plasma of animals is not very divergent from it in a chemical sense, and protoplasm itself is clearly re- lated, makes only a small leap of imagina- tion necessary to assume a relationship. It has been argued most recently by Macallum (1926) and Beutner (1938b) that the first land animals may have simply carried some of their earlier aquatic environment along with them. This would seem to be too pat a statement, for it would seem that if oceans and organisms are all part of one evolving chemical system, it would be quite natural for both to partake of considerable chemical similarity. Redfield (1934) gives some very suggestive data in this connection. This re- semblance of organisms would naturally be to a fluid in which there was considerable matter in solution rather than to the solid substrate or to fresh water. The osmotically low value of fresh water, if nothing else, would hardly be friendly to the consolida- tion of a primitive gel. See Pantin (1931) and Baldwin (1937). Croneis & Krumbein (1936) discuss the hypothesis of Chamber- lin to the effect that life may have originated in soil, largely on the basis of the dispersive effects of oceanic conditions. Under water, I subsoil conditions would seem to obviate the principal objection. Perhaps the primordial gel was interstitial in the sands of early beaches, an arena of life now occupied by many specialized forms, and which only re- cently has begun to receive the attention of biologists that it undoubtedly deserves. 1942] Breder: Evolutionary Hypotheses 139 The above remarks apply equally well to either a normal chemical evolution or a rare “accident.” Effects on Scientific Thought. The main purpose of the present paper is to discuss the effect of these various hypoth- eses on scientific thought. The different ideas that have been advanced to explain the methods of evolution are usually given and discussed as though the origin of life had no bearing on the subject. This is certainly not the case but in evaluating the plausi- bility of any of them we must necessarily consider their merits against a background of what origin of life is presupposed. Even in cases where there are alternates, the one selected causes certain strictures or pro- duces certain effects implied in all reason- ing derived therefrom. The primary forms of organic transfor- mation that have been suggested by biolo- gists may be reduced to four basic concepts : Cl) Inheritance of acquired characteris- tics (Lamarckianism), now discarded because of the failure of experimental evidence. (2) Natural selection in the Darwinian and neo-Darwinian sense and more or less under critical appraisement. (3) Orthogenesis, perhaps, not suscep- tible of scientific investigation by direct methods. (4) Mutation in the DeVriesian sense or in the modern version of Goldschmidt (1940) involving violent and sudden trans- formations. All other views seem to be variants or combinations of the above four, more fre- quently differing in terminology than in content. See Dobzhansky (1940). Before discussing these varying views in detail, the number of points of origin of life, which also has a marked bearing on the whole question of organic relationship, may best be examined. Considering monophyleticism versus poly- phyleticism, it should be clear that if the latter is implied by the nature of the origin assumed, it is pointless to attempt to build phylogenetic trees that try to tie all forms together. Since most present-day biologists, tacitly at least, imply a monophyletic origin, it must mean that some of the previously discussed ideas of life genesis have been dis- carded by them. It is evident that while a monophyletic origin may be considered for all concepts, since life on earth must have commenced at least once, such is not neces- sarily true of every concept. Stated another way, polyphyleticism is possible only if life origin is part of a normal process (planted or spontaneous). While this could also be monophyletic, the assumption of a rare “improbable” chemical “accident” or planting would almost certainly preclude polyphyleticism. With the “normal” processes of chemical evolution, an orthogenetic basis of evolu- tion would certainly be expected — simply as a mere continuing of a spontaneous activity. With the planted processes, either repeated or rare, or with a rare chemical accident, it might or might not be expected. In other words, any of the conceivable sources of life could carry with it the possibility of some orthogenetic scheme “built into” organisms. Under such a scheme there would be nothing to prevent the possible occurrence of muta- tions in the Goldschmidt sense. In fact, it might well be that such should be expected and thought of roughly as somewhat analog- ous to the change of phase in inorganic “orthogenetic” systems occurring when cer- tain points of development have been reached. This concept is, of course, not un- like that of the emergent evolution of C. L. Morgan (1923). Natural selection in the coarsest sense would no doubt operate from the first. Whether it carries over into the refinements generally expected of it is another matter, but one on which the present discussion may be suggestive. If organisms have some such orthogenetic “program” built into their be- ings, it is certainly not surprising that neo- Darwinians have much trouble in trying to explain natural selection on a micrometric basis. Adaptation, in the teleological sense, these thoughts do not tend to support, since the organic units must by random or other means find themselves in environments in which they could continue as such. Here again only the coarse effects can be argued for with any strength. All others could be just as well accounted for by some straight process of orthogenesis in which the pri- mary organisms either found themselves in an environment sufficiently suitable or per- ished if they did not. It is of more than passing interest in this connection that Spencer & Melroy (1942), on a basis of their results on exposing bac- teria, protozoa and flat worms to carcino- genic agents through many generations, wrote, “The biological generalization that certain environments may be ontogeneti- cally harmless but phylogenetically lethal is suggested.” This should not be interpreted to mean that organisms are not affected by their en- vironment except to survive or perish. How- ever one may care to try to account for the interesting associations of highly specialized organisms of today with their environment, these refinements certainly appeared at a relatively late date in phylogeny unless one wishes to assume that this feature of evo- lution is an inherent property of organisms. 140 Zoologica: New York Zoological Society [XXVII: 19 If it is, it is rather amazing how many ani- mals have lost the faculty of transmitting environmental effects (acquired characters) at least quickly, and how well they manage without it. This thought is, of course, re- lated to the complete inability of any one to establish satisfactorily the inheritance of acquired characters. The well-known habit of flounders of quickly matching their background in con- siderable detail may be considered a matter of individual behavior. This operates, and presumably without conscious effort, through sense organs (the eyes) but is in truth no more remarkable than the fact that we stand on two legs without conscious effort. Very possibly a flounder on a wrong back- ground is as uncomfortable as we are when out of plumb. Since any creature necessarily is limited to the equipment it is endowed with, it normally makes use of such as it has and survives or not, according to the results obtained. This, of course, is a far cry from the implications of the adaptationists who for long have certainly over-played the nice- ness of fit between organism and environ- ment and have failed conspicuously to ex- plain the development of complex organs, such as eyes, for example. That it is at least possible for flounders to reach sexual ma- turity without benefit of the usual color matching changes has been shown by Breder (1938). His fish may have survived on a purely random chance or it may have con- sistently kept its conspicuous coloration out of sight by more adequate burial than is usual to the species. All this obviously implies that what an animal does with its equipment is consider- ably superior to what particular kind it has, in a functional sense. The divergence of the forms of organisms can certainly be used to support this view as well as the essential basic similarity of what they all do. What it does is more closely associated with the re- strictions of environment than what it has. Stated another way, a Mammoth can be frozen in the Arctic tundra for generations and essentially retain its form outside of temperature ranges at which it can do any- thing. Or again, an insect might have its form changed (and be ruined) by being caught in the beak of a bird. It could be saved from this by flying away (if it had wings), crawling between grass blades (if it had none), or even just by sitting very still. This is merely fitting its activity to its morphological limits to suit its environ- ment. That these morphological limits may be exceedingly various is evident from the above — but what it does must fit its needs much more closely; thus a long-legged and a short-legged insect may find shelter beneath the same leaf. It is little wonder that the “fitness” and lack of “fitness” of animals to environment as based on morphology have given rise to endless discussion. Returning to the basic argument, how- ever, it should be clear from the preceding, without laboring the point, that before such activity can take place at all the primitive organic entity must first fit its environment in a purely physiological sense. The tem- perature thresholds that hem in our kind of life are clearly first in physiological impor- tance and these must be substantially main- tained long before adjustments between one animal and another can even be conceived. The losses of species due to glaciation and other major changes so far overshadow micro-adjustments that one is forced to wonder at their alleged importance. If an orthogenesis could be established, all the ob- served results could be reasonably explained, whereas the reverse cannot be said for the usual concepts of pure natural selection and other similar ideas. Finally, the thoughts here expressed lean strongly to the conception of life origin on Earth as a physico-chemical evolution of the sort conceived of by Oparin (1938). This inevitably carries with it a strong bias to orthogenesis, places emphasis on the funda- mental limiting factors of life, on the im- portance of what an organism does with its equipment rather than what equipment it has, and questions the micrometric func- tioning of selective processes. Furthermore, in an evolving physico-chemical system in which one reaction follows another, no one would think of referring to the disappear- ance of some compound, as the result of this activity, as natural selection. Only when such become sufficiently complicated and in- dividuated does this word put in its appear- ance (biological literature) and in this sense would seem to be reducible to the status of a convenient phrase to hide behind. Lotka’s (1925) chapter on “the statistical meaning of irreversibility” is very suggestive in this connection as is the physico-chemical deter- minism of Alexander (1939). The function of reproduction in connec- tion with this problem is far from clear. Oparin (1938), in spite of his interesting discussion of coazervate systems, becomes somewhat vague at this point. Since it is one of the generally obvious basic differ- ences between the animate and inanimate, it is in serious need of further critical study from new approaches. Perhaps here, after all, lies the key to the secret of life and what may be responsible for the effects that have given biologists so much trouble of inter- pretation. Since viruses apparently do not have need for the elaborate mechanisms of reproduction to be found in the larger life forms, it may well be that the entire busi- ness of both sexual and asexual reproduc- tion is a means of circumventing the re- 1942] Breder: Evolutionary Hypotheses 141 striding influences of elaborate morphologi- cal structure, a need that does not arise so long as the naked enzymes have no more than molecular structure with which to deal. If we consider life as a retardation of a general increasing entropy it follows that by the peculiar nature of living processes we have in effect a building up of a certain amount of back pressure to the degradation of energy. This, so far as we know, always oc- curs in individuated clumps of heterogeneous matter — the individual of the biologist. This goes on for a time and finally the individual disintegrates. From the first, all such indi- viduated organizations are fighting a losing battle against the general downward trend of available energy, which is evidently bound to culminate in death. This places re- production in the role of casting off new colonies of such aggregates of material — like a long relay race up a down-moving escalator. Whether this necessarily culmi- nates in the eventual loss of the race and the generally accepted heat-death is still not clear, at least for our little niche in the space-time frame. A priori, there seems to be no particular reason for things working just that way, for surely it is conceivable that some sort of energy exchange system could be imagined that would not necessarily exhaust itself in a short time, passing on only a tiny fragment saved from the forces of destruction, if it were not running a breathless race up a down-escalator. These ideas lead naturally to a piece of pure speculation that it is tempting to in- dulge in and which really is at the heai't of the perhaps seemingly unconnected items that comprise this paper. If we visualize all known items within our universe as a sim- ple expression of physico-chemical activity, as a part of an increasing entropy, it is clear from the preceding that we are very near the end, in a cosmic sense. Life, as we know it, appears well along to the end and gives the appearance of a manifestation that seems to give back pressure to the common lowering of energy levels. Up to the appear- ance of autocatalysis all activity would run along well enough according to the straight- forward operation of the Second Law. Fi- nally the autocatalysists emerge into ex- tremely complex individualized units pos- sessing tremendous urges to run counter to the forces of their surroundings, which urges are more and more forceful as that strange element “consciousness” appears. Finally we reach the stage of present man with his deliberate, elaborate and more or less successful attempts to modify his en- vironment. With this has come careful, though blundering, thought processes at- tempting to understand what it is all about, purely emotional at first but grading finally into an intellectual and objective plane. One wonders if this drive is a new one directed against the continuing entropic drift simi- lar to but superimposed and of a later order than the fundamental sex drive that seems to be the final difference between the inert drift downward of the non-living and the hold-back of living entities. Looked at this way, there is little wonder that such opera- tions are a source of trouble and a general preoccupation. This, of course, is as far as we can see, as it is the level of activity at which we now cut off. What then of the future? Two pos- sibilities seem to appear. First that the life activity merely peters out sooner or later, and it is in truth just one hold-back on a descending energy scale (and there may have been others, as already suggested) and will pass on with the universe to complete entropy. Secondly, it may be that life ac- tivity is of more importance than reason dares permit us to postulate. The present activity of man one may speculate upon as part of the physico-chemical evolution now going on but inducing a progressive retar- dation in the otherwise increasing entropy. This could conceivably result in one of two fashions. One would be an increasing slow- ing of the process as entropy approaches the absolute. In other words, the evolution of such a physico-chemical system could be expressed by a curve reaching an asymptote. The other would be a complete reversal of trends, perhaps brought about by the “in- telligence” of some groups of individuated bits of matter in which case the curve might be something in the form of a parabola, a hyperbola or, fantastically, even a closed figure. 142 Zoologica : New York Zoological Society [XXVII: 19 Bibliography. For recent discussions of and full bibliog- raphies on the subject of life origin, see espe- cially Lippmann (1933) and Oparin (1938). Alexander, Jerome 1939. Physico-chemical determinism in biol- ogy. Trans. N. Y. Acad. Sci., Ser. II, 2 (1) : 14-20. Arrhenius, Svante A. 1908. Worlds in the making. The evolution of the universe, xiii + 229 p. Harper & Bros., New York. 1911. Das Schicksal der Planeten. 55 p. Aka- demische Verlagsgesellschaft, Leipzig. Baldwin, E. 1937. An introduction to comparative bio- chemistry. 112 p. Cambridge Univ. Press. Beutner, R. 1938a. The independent nature of morpho- genesis and self-reproduction and its significance for the cosmic develop- ment of life. Biodynamica, (38) : 1-7. 1938b. Life’s beginning on earth. 222 p. Williams & Wilkins, Baltimore. Breder, C. M., Jr. 1936. Tissue culture and explantation in nature. A review of certain experi- ments and possibilities. Zoologica (N. Y.), 21 (10) : 115-124. 1938. An unusual aberrantly colored pleuro- nectid. Zoologica (N. Y.), 23 (20): 393-396. Clarke, F. W. 1924. The data of geochemistry. Bull. U. S. Biol. Survey (770): 1-841. (5th ed.). Croneis, C. & W. C. Krumbein 1936. Down to earth, xviii + 501 p. Univ. Chicago Press. Dobzhansky, T. 1940. Catastrophism versus evolution. Sci- ence, 92 (2390) : 356-358. Fairchild, H. L. 1938. Selenology and cosmology. Science, 88 (2294) : 555-562. Goertz, A. 1938. The living state of matter in the range of low temperature. Science, 88 (2291): 501-502. (Abstract). Goldschmidt, Richard 1940. The material basis of evolution. 436 p. Yale Univ. Press. Goldschmidt, V. M. 1937. The principles of distribution of chem- ical elements in minerals and rocks. Journ. Chemical Soc., London: 655- 673. Henderson, L. S. 1913. The fitness of the environment. xi + 317 p. Macmillan, New York. Herrera, A. L. 1942. A new theory of the origin and nature of life. Science, 96 (2479) : 14. Hunt, W. R. 1939. “Whence cometh life?”. Scientific Monthly (June) : 550-554. Jeans, J. 1942. Is there life on other worlds? Science, 95 (2476) : 589-592. Jones, H. S. 1940. Life on other worlds. 229 p. Mac- millan, New York. Lewis, Gilbert N. 1934. The genesis of the elements. Physical Rev., 2nd ser., 46 (10) : 897-901. Lichtig, I. 1938. Die Entstehung des Lebens durch stetige Schopfung. xx+371 p. Uit- gevens Maatsch, Amsterdam. Lipman, C. B. 1932. Are there living bacteria in stony meteorites? Amer. Mus. Novitates, (58) : 1-19. Lippmann, E. O. von 1933. Urzeugung und Lebenskraft: zur Ge- schichte dieser Probleme von den altester Zeiten an bis zu den Anfangen des 20 Jahrhunderts. 135 p. Springer, Berlin. Lotka, Alfred J. 1925. Elements of physical biology. xxx+ 460 p. Williams & Wilkins, Baltimore. Lutz, Frank E. 1929. Experiments with “Wonder Crea- tures.” Natural History, 29 (2) : 160-168. Macallum, A. B. 1926. The palaeochemistry of the body fluids and tissues. Physiol. Rev., 6 (2) : 316-357. Martin, G. J. & Fisher, C. V. 1942. On the nature of virus adaptations. Science, 95 (2476) : 602-603. Morgan, C. Lloyd 1923. Emergent evolution. The Gifford lec- tures. xii + 313 p. Henry Holt, New York. Northrop, F. S. C. & H. S. Burr 1937. Experimental findings concerning the electro-dynamic theory of life and an analysis of their physical meaning. Growth, 1 (1): 78-88. Oparin, A. I. 1938. The origin of life. 270 p. Macmillan, New York. (Trans, by S. Morgulis). Pantin, C. F. A. 1931. The origin of the composition of the body fluids of animals. Biol. Rev., 6 (4) : 459-482. Pike, F. H. 1939. On a thermodynamic concept of the environment. Anat. Rec., 75 (4, suppl.) : 95. (Abstract). Redfield, A. C. 1934. On the proportions of organic deriva- tives in sea water and their relation to the composition of plankton. James Johnstone Mem. Vol. : 176-192. Univ. Liverpool. Riddle, Oscar 1939. Epic of life. Scientific Monthly (June) : 530-533. Rivers, T. M. 1939. Viruses and virus diseases. Stanford Univ. Pub. Univ. Ser. Med. Sci., 4 (1) : 1-133. 1942] Breder : Evolutionary Hypotheses 143 Rosett, J. 1917. Observations on a new type of arti- ficial osmotic cell. Plant World, 20 (2) : 37-57. Russell, H. N. 1941. The cosmical abundance of the ele- ments. Science, 94 (2443) : 375-381. Smith, Homer W. 1932. Kamongo. 167 p. Viking, New York. Spencer, R. R. & Melroy, M. B. 1942. The mechanism of species adaptation to carcinogens. Science, 95 (2476) : 592-595. Stanley, W. M. 1937. Isolation and properties of virus pro- teins. Erg. Physiol., biol. Chemie u. exp. Pharm., 39: 294-347. 1938a. The nature of viruses. Trans. N. Y. Acad. Sci., Ser. II, I (2) : 21-24. 1938b. The reproduction of virus proteins. Amer. Nat., 72 (739): 110-123. (Re- printed in Ann. Rept. Smithsonian Inst, for 1938 : 499-509). WlLDT, R. 1940. On the possible existence of formalde- hyde in the atmosphere of Venus. The Astrophysical Journ., 92 (2) : 247- 255. Woodruff, L. L. 1936. Foundations of biology, xiv+583 p. Macmillan, New York. (5th ed.) See Chapter 16. 1942] Index to Parts 1-Jf 145 INDEX. Names in bold face indicate new genera, species or varieties; numbers in bold face indicate illus- trations; numbers in parentheses indicate the paper containing the Plate numbers immediately following. A Aconophora fusiformis, 126 marginata, 126 pallescens, 126 projecta, 126 Albula, 43 vulpes, 44 Amastris antica, 128 compacta, 128 minuta, 128 obtegens, 128 Amazona mercenaria mercenaria, 50 Ameiurus nebulosus, 89 Ammobia, 31 Ammophila, 32 Ampulex minor, 29 sagax, 29 Ampulicidae, 29 Anabacerthia striaticollis montanus, 51 Anabates montanus, 51 ochrolaemus, 51 Anchoviella sp., 44 Anguilla, 91 Anguispira alternata, 81 Anoa depressicornis, 19 Anoptichthys, 7, (3) Plates I-III jordani, 7, 14 Antianthe expansa, 128 Antilope cervicapra, 19 Aphriza virgata, 49 Aratinga mitrata mitrata, 50 Arremon frontalis, 53 Artibeus jamaicensis, 10 Artiodactyla, 17 Astyanax, 8 mexicanus, 11, 14 Atherina sp., 44 Atlapetes brunnei-nucha brunnei-nucha, 53 Atomarchus multimaculatus, 121 Automolus ochrolaemus ochrolaemus, 51 B Bison bison, 19 Blennius cornutus, 65 Bocydium globulare, 129 Boethoos reticulata, 127 Bos taurus, 19 Box salpa, 65 c Cambarus blandingii cuevachicae, 11 Campylenchia hastata, 126 Capito glaucogularis, 51 versicolor glaucogularis, 51 Capra hircus, 19 Caprimulgus ocellatus, 50 Carassius auratus, 89 Ceresa sallai, 127 vitulus minor, 127 Cervus axis, 19 canadensis, 19 Ceyx dillwynni, 56 enopopygius, 56 erithacus, 5? erithacus captus, 58 erithacus erithacus, 55 erithacus macrocarus, 56 erithacus motleyi, 56, 57 erithacus vargasi, 58 innominatus, 58 melanurus, 55 rufidorsa, 58 rufidorsus, 55, 57 rufidorsus jungei, 59 rufidorsus rufidorsus, 58 Chamaeza brevicauda olivacea, 51 olivacea, 51 Chanos, 45 Charadrius Winterfeldi, 49 Chelidonichthys cupensis, 65 Chlorion, 30 (Ammobia) brasilianum, 31 (Ammobia) funestum, 31 (Ammobia) ichneumoneum, 31 (Ammobia) melanopum, 31 (Ammobia) neotropicum, 31 (Ammobia) singularis, 31 (Isodontia) aztecum, 30 (Isodontia) costipennis, 30 (Isodentia) dolosum, 30 (Priononyx) striatum, 30 Cissopis leveriana leveriana, 52 minor, 52 Clinus anguillaris, 65 capensis, 65 cottoides, 65 superciliosus, 65 taures, 65 Clupea harengus, 75 Cnesterodon decemmaculatus, 85 Columba frenata, 50 infuscata, 53 meloda, 50 plumbea delicata, 53 Conurus mitratus, 50 rupicola, 50 Cyclops, 13 Cymbomorpha dorsata, 127 Cyphonia clavata, 128 proxima, 128 Cyphorhinus thoracicus, 52 Cyprinus, 89 carpio, 89 D Darnis latior, 127 partita, 127 Dendrocolaptes chunchotambo, 51 Dixonina, 43, (8) Plates I, II nemoptera, 43 pacifica, 43 Dorosoma, 45 E Elaenia viridiflava, 51 Elapomorphus mexicanum, 37 Enchenopa albidorsa, 126 lanceolata, 126 serratipes, 126 Endosphaera engelmanni, 66 Entomodestcs leucotis, 52 Erechtia brunneidorsata, 126 punctipes, 126 Esox lucius, 89 masquinongy, 71 niger, 71, 89 Euglandina rosea, 81, (15) Plate I Exelencophis, 33 146 Zoologica : New York Zoological Society [XXVII F Fundulus, 85 G Gasterosteus, 90 Geagras redimitis, 33 Gobiosoma bosci, 61 molestum, 61 robustum, 61, 63, (11) Plates I, II H Helix, 82 Heniconotus armatus, 127 Heteronotus spinosus, 127 tridens, 127 Hille exaltata, 128 herbicola, 128 Homalocranion deppei, 36 praeoculum, 38 Homalocranium atriceps, 34 bocourti, 34 boulengeri, 38 deppei, 36 jani, 36 melanocephalum, 37 miniatum, 36 moesta, 38 Horiola lineolata, 127 picta, 127 Huro salmoides, 78 i Icterus brevirostris, 53 Iridosornis analis analis, 52 Isodontia, 30 J Jenkinsia, 77 Jenynsia lineata, 85 L Lebistes reticulatus, 86, 88, (16) Plates I, II Leioscyta rufidorsa, 126 Lepomis auritus, 76 Leptopogon superciliaris, 52 superciliaris superciliaris, 52 Leucolepis thoracica thoracica, 52 Liguus, 82 Limaces, 82 Limia tricolor, 86, 88, (16) Plate II Lophius piscatorius, 89 Lycoderes hippocampus, 129 M Macrobrachium jamaicensis, 11 Membracidae, 125 Membracis arcuata, 125 fasciata, 125 foliata, 125 fusca, 125 provittata, 125 Mesomphix cupreus, 81 inornatus, 81 Micronycteris megalotis mexicanus, 11 Micrutalis pallens, 127 Mollienisia latipinna, 86, 88, (16) Plate I sphenops, 86, 88, (16) Plates I, II Molothrus bonariensis occidentalis, 53 Mormoops megalophylla senicula, 10 \/ Mormyrus, 88 Morone labrax, 78 Mugil, 3 capito, 65 N Nassunia binotata, 127 Natalus mexicanus, 10 Natrix valida, 121 Notemigonus, 89 Nyctiphrynus ocellatus ocellatus, 50 o Odocoileus virginianus, 19 Opisthonema, 45 Opsanus beta, 62 tau, 62 Oreopeleia bourcieri frenata, 50 bourcieri subgrisea, 50 Ortalis guttata adspcrsa, 49 Orthalicus, 82 Ortygonax rytirhynchus rytirhynchus, 53 Ovis aries, 19 canadensis, 19 musimon, 19 tragelaphus, 19 p Penelope adspersa, 49 Pholeomyia indecora, 11 Pimelodus clarias, 85 Pionus tumultuosus, 50 Pipra chloromeros, 52 Platypoecilus, 8 variatus, 85, 86, (16) Plates I, II Pneumatophorus grex, 75, (14) Plate I Podium, 29 brevicolle, 30 (Dynatus) nigripes, 29 gorianum, 30 (Trigonopsis) abdominale, 30 Poeciliids, 85, (16) Plates I, II Poephagus grunniens, 19 Polygra septemvolva, 82 volvoxis, 82 Priononyx, 30 Pristopoma bennettii, 65 Pseudopleuronectes americanus, 89 Psittacus mercenarius, 50 tumultuosus, 50 Psychoda sp., 11 Pterygia cerviceps, 126 uropigii, 126 Ptilogonys leucotis, 52 Pungitus, 89 Pyrrhura rupicola, 50 R Rallus caesius, 53 Rhexia pallescens, 128 Sceliphron, 30 fistulare, 30 Scolecophis fumiceps, 38 Scorpaena, 65 Signalosa, 45 Sphecidae (Sphecinae), 29 Spheroides maculatus, 65, 71 testudineus, 65 Sphex, 31 abbreviatus, 31 melanarius, 31 muticus, 31 opulentus, 32 Squilla hancocki, 44 Stepaspis insignis, 128 laevipennis, 129 Stictopelta fraterna, 127 Stizostedion, 47 Stylocentrus ancora, 128 Symphodus, 65 T Tanagra analis, 52 frugilegus, 52 Tantilla, 33 armillata, 36 atriceps, 33, 34 bimaculata, 35 bocourti, 33, 34 brevissima, 33 1942] Index to Parts 1-U 147 calamarina, 33, 35 canula, 33, 35 coronata, 33 cuniculator, 33, 35 deppei, 33, 36 depressa, 33 deviatrix, 33, 36 eiseni, 33, 41 fumiceps, 33 fusca, 37 gracilis, 33 hobartsmithi, 33, 36 jani, 33, 36 kirnia, 38 lintoni, 33 martindelcampoi, 33, 37 melanocephala, 36 mexicana, 33, 37 miniata, 33, 38 miniator, 33, 38 moesta, 33, 38 moesta cuniculator, 35 nelsoni, 33 nigriceps, 33 nigriceps fumiceps, 38 nigriceps nigriceps, 38 phrenitica, 33, 39 planiceps, 33, 41 rubra, 33, 40 rubricata, 33 schistosa, 39 striata, 33, 40 taeniata, 37 trilineata, 37 trivittata, 37 utahensis, 33 virgata, 37 wagneri, 33 wilcoxi, 33, 40 wilcoxi rubricata, 40 wilcoxi wilcoxi, 40 yaquia, 33, 41 Tantillita, 33 Tarpon, 45 Taurotragus oryx, 19 Thamnophilus luctuosus, 53 melanochrous, 53 Thamnophis, 97 angustirostris, 120 chrysocephalus, 104 eques cyrtopsis, 105, 108 eques eques, 104, 106 eques postremus, 109 macrostemma macrostemma, 115 macrostemma megalops, 115 marcianus, 114 melanogaster canescens, 117 melanogaster melanogaster, 116 ordinoides errans, 112 ordinoides vagrans, 113 phenax halophilus, 100 phenax phenax, 99 rufipunctatus, 120 ruthveni, 114 sauritus chalceus, 116 sauritus proximus, 116 scalaris, 100 scalaris godmani, 101 scalaris scalaris, 100 scalaris scaliger, 103 sirtalis parietalis, 114 sumichrasti cerebrosus, 110, 111 sumichrasti fulvus, 110 sumichrasti praeocularis, 110, 111 sumichrasti sumichrasti, 110 vicinus, 104 Thraupis bonariensis darwinii, 52 Tolania opponens, 129 Torpedo occidentalis, 25, 26-28 Trichodina, 65 blenni, 65 chelidonichthys, 65 clini, 65 fariai, 65 halli, 65, 66. 67, (12) Plates Mil labrorum, 65 mitra, 69 mugilis, 65 patellae, 71 pediculus, 70 renicola, 70 scorpaena, 65 spheroidesi, 65, 66, 66, 68, 69, (12) Plates I-III steinii, 69 scoploplontis, 70 Trigla, 65 Triodopsis albolabris, 81 Tropidonotus Baronis Mulleri, 116 grahamii, 116 intermedius, 99 mesomelanus, 116 natrix, 99 Tynelia hirsuta, 128 Tyranniscus viridiflavus, 51 u Umbonia spinosa, 126 Urceolaria korschelti, 71 Vanduzea triguttata, 128 Vauchomia nephritica, 70 Vorticella nebulifera, 70 x Xiphophorus hellerii, 73, 74, 85, 88, (16) Plate I Xiphorhynchus chunchotambo chunchotambo, 51 z Zenaida asiatica meloda, 50 NEW YORK ZOOLOGICAL SOCIETY General Office: 630 Fifth Avenue, New York City OFFICERS Fairfield Osborn, President Alfred Ely, First Vice-president Laurance S. Rockefeller, Chairman, Executive Committee & Second Vice-president Harold J. O’Connell, Treasurer Cornelius R. Agnew, Treasurer SCIENTIFIC STAFF General William Bridges, Editor and Curator of Publications Jean Delacour, Technical Adviser John Tee-Van, Executive Secretary Zoological Park Lee S. Crandall, Curator of Birds Leonard J. Goss, Veterinarian Claude W. Leister, Curator of Mammals John Tee-Van, Acting Curator of Reptiles Charles M. Breder, Jr., Director Christopher W. Coates, Aquarist Myron Gordon, Research Associate in Genetics Ross F. Nigrelli, Pathologist G. M. Smith, Research Associate in Pathology Homer W. Smith, Research Associate in Physiology Department of Tropical Research William Beebe, Director Jocelyn Crane, Research Zoologist Henry Fleming, Entomologist Aquarium William K. Gregory, Associate Gloria Hollister, Associate John Tee-Van, Associate Mary VanderPyl, Associate Editorial Committee Fairfield Osborn, Chairman William Beebe Charles M. Breder, Jr. William Bridges Jean Delacour John Tee-Van ZOOLOGICA SCIENTIFIC CONTRIBUTIONS of the NEW YORK ZOOLOGICAL SOCIETY VOLUME XXVIII 1943 Numbers 1-23 Published by the Society The Zoological Park, New York 60, N. Y. NEW YORK ZOOLOGICAL SOCIETY General Office: 630 Fifth Avenue, New York 20, N. Y. OFFICERS Fairfield Osborn, President Alfred Ely, First Vice-president Laurance S. Rockefeller, Second Vice-president Harold J. O’Connell, Secretary Cornelius R. Agnew, Treasurer William Bridges, Editor and Curator of Publications Zoological Park Lee S. Crandall, General Curator & Curator of Birds Leonard J. Goss, Veterinarian Claude W. Leister, Associate, Mammals Harry C. Raven, Associate, Anatomy John Tee-Van, Associate, Reptiles Grace Davall, Assistant to General Curator Charles M. Breder, Jr., Director Christopher W. Coates, Aquarist Ross F. Nigrelli, Pathologist Myron Gordon, Research Associate in Genetics G. M. Smith, Research Associate in Pathology Homer W. Smith, Research Associate in Physiology SCIENTIFIC STAFF General John Tee-Van, Executive Secretary Jean Delacour, Technical Adviser Claude W. Leister, Education Donald Marcy, Associate, Education Aquarium Department of Tropical Research William Beebe, Director Jocelyn Crane, Research Zoologist Henry Fleming, Entomologist William K. Gregory, Associate Gloria Hollister, Associate John Tee-Van, Associate Mary VanderPyl, Associate Editorial Committee Fairfield Osborn, Chairman William Beebe Charles M. Breder, Jr., William Bridges Jean Delacour Claude W. Leister John Tee-Van CONTENTS Part 1. May 4, 1943. 1. Biometry of Puffers and Their Parasites. By Ross F. NlGRELLi & James W. Atz 1 2. The Genetics of Melanoma in Fishes. VI. Mendelian Segregation of Melanophore Reaction Types In Embryos of a Melanomatous Mother. By Myron Gordon & Fred Flathman. (Text-figures 1-3) 9 3. Pattern and Color in the Cichlid Fish, Aequidens tetramerus. By William Beebe. (Plate I) 13 4. A Revision of the Genera and Species of the Family Pycnonotidae (Bulbuls). By Jean Delacour. (Plates I & II; Text-figures 1-13) 17 5. Two New Subspecies of Pycnonotus cafer. By Jean Delacour. (Text-figure 1) 29 6. Eastern Pacific Expeditions of the New York Zoological Society. XXXI. Uca schmitti, a New Species of Brachyuran Crab from the West Coast of Central America. By Jocelyn Crane. (Plate I; Text-figure 1) 31 7. Crabs of the Genus Uca from Venezuela. By Jocelyn Crane. (Plate I; Text-figure 1) 33 8. Observations on the Electric Discharge of Narcine brasiliensis (Olfers) . By R. T. Cox & C. M. Breder, Jr. (Text-figures 1-4) 45 Part °2. September 15, 1943. 9. Physical Factors in the Ecology of Caripito, Venezuela. By William Beebe. (Plates I & II; Text-figures 1-5) 53 10. On the Locomotor and Feeding Behavior of Certain Postlarval Clupeoidea. By C. M. Breder, Jr., and Louis A. Krumholz. (Plates I & II; Text-figures 1-3) 61 11. A Revision of the Subfamily Estrildinae of the Family Ploceidae. By Jean Delacour. (Text-figure 1) 69 12. The Relationship Between Weight and Body Form in Various Species of Scombroid Fishes. By Sidney Shapiro. (Text- figures 1-12) 87 13. A Record of the Successful Breeding of the Quetzal ( Pharo - machrus mocinno costaricensis) in Captivity. By Charles CORDIER 105 14. The Occurrence of Leeches, Ozobranchus branchiatus (Menzies), on Fibro-epithelial Tumors of Marine Turtles, Chelonia mydas (Linnaeus). By Ross F. Nigrelli and G. M. Smith. (Plates I-III) 107 Part 3. December 6, 1943. 15. On the Relationship of Some Common Fishes as Determined by the Precipitin Reaction. By Douglas G. Gemeroy. (Text-fig- ures 1-7) 109 16. Spiders of the Families Lyssomanidae and Salticidae (Magoninae) from British Guiana and Venezuela. By JOCELYN CRANE. (Text-figures 1-4) 125 17. Notes on the Breeding of the Empress of Germany’s Bird of Para- dise in Captivity. By Prince K. S. Dharmakumarsinhji of Bhavnagar. (Plate I) 139 18. Evidence of Healed Hungerosteomalacia (Late Rickets) in a Green Monkey ( Cercopithecus sabaeus) . By Joseph Pick. (Plate I) 145 19. Eastern Pacific Expeditions of the New York Zoological Society. XXXII. Mollusks from the West Coast of Mexico and Central America. By Leo George Hertlein & A. M. Strong. (Plate I) 149 Part 4. December 31, 1943. 20. Chemical Sensory Reactions in the Mexican Blind Characin. By C. M. Breder, Jr., & Priscilla Rasquin. (Plates I-III; Text- figures 1-9) 169 21. The Avian Genus Zosterops in Siam, with Description of One New Race. By H. G. Deignan 201 22. Causes of Diseases and Death of Fishes in Captivity. By ROSS F. Nigrelli. (Plates I-VI) 203 23. Display, Breeding and Relationships of Fiddler Crabs (Brachyura, Genus Uca) in the Northeastern United States. By Jocelyn Crane. (Text-figure 1) 217 Index to Volume XXVIII 225 ZOOLOGICA SCIENTIFIC CONTRIBUTIONS of the NEW YORK ZOOLOGICAL SOCIETY VOLUME XXVIII Part 1 Numbers 1-8 Published by the Society The Zoological Park, New York May 4, 1943 CONTENTS 1. Biometry of Puffers and Their Parasites. By Ross F. Nigrelli & James W. Atz 1 2. The Genetics of Melanoma in Fishes. VI. Mendelian Segregation of Melanophore Reaction Types In Embryos of a Melanomatous Mother. By Myron Gordon & Fred Flathman. (Text-figures 1-3) 9 3. Pattern and Color in the Cichlid Fish, Aequidens tetramerus. By William Beebe. (Plate I) 13 4. A Revision of the Genera and Species of the Family Pycnonotidae (Bulbuls). By Jean Delacour. (Plates I & II; Text-figures 1-13) 17 5. Two New Subspecies of Pycnonotus cafer. By Jean Delacour. (Text-figure 1) 29 6. Eastern Pacific Expeditions of the New York Zoological Society. XXXI. Uca schmitti, a New Species of Brachyuran Crab from the West Coast of Central America. By Jocelyn Crane. (Plate I; Text-figure 1) 31 7. Crabs of the Genus Uca from Venezuela. By Jocelyn Crane. (Plate I; Text-figure 1) 33 8. Observations on the Electric Discharge of Narcine brasiliensis (Offers) . By R. T. Cox & C. M. Breder, Jr. (Text-figures 1-4) 45 Nigrelli & Atz: Biometry of Puffers and Parasites 1 1. Biometry of Puffers and Their Parasites. Ross F. Nigrelli & James W. Atz New York Zoological Society Introduction. In connection with other investigations of the New York Aquarium, studies were made of fishes and fish parasites indigenous to Sandy Hook Bay, N. J. The fishes of this area have been thoroughly studied by Bre- der (1923-1939) and his collaborators. Among other species, the northern puffer or blowfish, Spheroides maculatus (Bloch & Schneider) , is found there in large numbers from June to September and is collected in these months for exhibition purpose. A routine check for parasites is made on all fishes that eventually will become part of the Aquarium’s exhibition. In the course of these investigations it was evident that puf- fers were the foci for infestations that were responsible for mortalities of other and more valuable species of fishes. For this reason, it was decided that a survey of their parasitic fauna would yield additional in- formation in this regard and that perhaps such a study would suggest methods for their control. Samplings. A random selection of puffers was made when the specimens were brought into the Aquarium from the pound nets in Sandy Hook Bay. A total of 294 fish was taken, varying in size from 2 to 24 cm. In Table I the number of individuals collected in the months from June to October, inclusive, for the years 1938, 1939 and 1940, are shown. The largest number of puffers was caught in September for the three years. Table II gives the numbers of each sex collected dur- ing the same period. 49% of these were males; 44% were females; 7% were im- mature. The males varied in size from 10-20 cm.; females from 11-24; immature puffers measured up to 12 cm. In this col- lection, fish 21 cm. or more were females. This distribution is shown in Table III and agrees with the data given by Welsh & Breder (1922) for puffers caught off At- lantic City, N. J. The only difference noted is the slight extension for the mature forms in both the upper and lower size limits, the result of a larger sample. Analysis of the Intestinal Contents of Puffers. The intestinal contents of puffers were examined by Linton (1901, 1905) from Woods Hole, Mass., and Beaufort, N. C., and by Welsh & Breder (1922) from Atlantic City. These investigators found that the puffer’s diet consisted of minute crusta- ceans, crabs, shrimps, bivalve and univalve mollusks, annelids and sea-weeds. In addi- tion to such items, fish was also part of the diet of Sandy Hook puffers. It was Ward (1908) who stated that “The parasitic fauna of any animal is primarily a function of its habitat” and, one may add, of its eating habits. A knowledge of the food cycle is very important, often revealing the source of internal parasites. Thus, for the five trematodes reported in this paper (see Table IV), the infective stages ai'e to be found in mollusks, crustaceans and fishes. Martin (1939) showed that in the case of Stephanostomum tenue, one of the intestinal flukes of the puffer, the cercariae develop in the snail, Nassa obsoleta, and the metacer- cariae are found in the spearing or common silverside, Menidia menidia notata. The spearing is a small fish occurring in great abundance along our coast and is usually preyed upon by larger fish. It is not surpris- ing, therefore, to find this species of fluke living in a large variety of definitive hosts. In the life-histories of Lecithaster and Podocotyle (see Hunnien & Cable, 1941 a and b), the cercarial stages appear in cer- tain snails and the metacercariae or infec- tive forms occur unencysted in copepods and amphipods. Since these minute crustaceans are part of the diet of many marine fishes, it is to be expected that no host-specificity is shown by these trematodes. The cycles of Bianium and Lintonium are still unknown but they are almost certain to follow one of the patterns indicated above. 2 Zoologica: New York Zoological Society [XXVIII: 1 Parasites of the Puffer. A. External Parasites. Insofar as could be determined, about 18 species of parasites have been recorded from the puffer. Thirteen of these were found in Sandv Hook Bay fish and are listed in Table IV. The small opercular opening, characteris- tic of puffers, encourages the concentration of ectoparasites in the gill chambers. In regard to pathogenicity, ectoparasites of fishes are the most important ones to con- sider for invariably they cause death of the host if sufficient numbers are present. Table I. Number of Puffers Collected, According to Dates. Month 1938 1939 1940 Total June 7 6 12 25 Julv . 15 69 84 August 20 20 19 59 Septem j:t 39 54 30 123 October 1 2 — 3 Total 67 97 130 294 The dinoflagellate, Oodinium ocellatum Brown, is recorded from only two fish in this report but in previous years (1935- 1937 ) it was the cause of mortality of vari- ous marine fishes (tropical and temperate) kept in captivity. The life-history of Oodin- ium was demonstrated by Nigrelli (1936) and it was shown that the epidemic in the Aquarium was centered in puffers and spiny boxfish brought in from Sandy Hook Bay. The two species of Trichodina were de- scribed recently by Padnos & Nigrelli (1942). Heavy infestations of these cili- ates will cause death. Nigrelli (1940) showed that puffers were the source for Trichodina found on other fishes in the Aquarium. That these peritrich ciliates are not limited to Sandy Hook fish is indicated by the fact that in 1939 they were also present on the gills and skin of 27 out of 32 puffers examined from Orient Point, L. I. The protozoa reported by Linton (1940) as Cyclochaeta domerguei Moroff (p. 23) may be Trichodina. If this is true, it appears that the puffers are infected with the organ- ism throughout its known range. Tables V and VI show the relationship of the intensity of Trichodina infection to sex of the host and to the dates the puffers were taken. Fish showing an exceptionally light infection may be considered negative for it is suspected that some of the gills probably were contaminated at the time of the examinations. As will be seen later, the host-parasite relation shown here is oppos- ite to that demonstrated for the metazoan parasites of the puffers. There is a ten- dency towards an increase infection in males. The intensity of the infection in all fish decreases towards the latter part of the summer. The effect of temperature has not been definitely decided. Our records show that the temperature in the bay off the Battery reaches a maximum in August and begins to drop in the latter part of the month. Table VI shows that there may be a correlation, and this is to be expected, between the temperature and growth of the ciliates. The Table also indicates that the optimum is around 70° F., for it is at this temperature that the largest number of puffers were infected. This interpretation is further corroborated by the fact that most of the reproductive stages (conjuga- tion) of this organism were obtained from the gills of puffers caught in July and Sep- tember. A single myxosporidian cyst was found on the gills. The spores are bivalve and more or less oval in shape, indicating that this form should be allocated to the family Myxo- somatidae. No specific identification was made because of insufficient material. How- ever, since the mode of transmission is direct, such organisms must be considered as potential parasites of real importance for under epidemic conditions they could cause considerable damage to tissues (Ni- grelli & Smith, 1938) and eventually result in death. The copepod, Pseudochondracanthus dice- raus Wilson (see Wilson, 1932), is another important gill parasite of puffers. As many as 140 specimens were found attached to the gills of a single fish. The method of trans- mission is simple and direct. The fertile Table II. Number of Puffers Collected, According to Sex. Month M 1938 F I M 1939 F I M 1940 F I M Total F I Grand Total June 2 5 — 2 4 — 4 8 — 8 17 25 July — — — 7 8 — 40 29 47 37 — 84 August 2 4 14 10 7 3 13 6 — 25 17 17 59 September 18 21 — 28 24 2 18 12 — 64 57 2 123 October — — 1 1 1 — — — — 1 1 1 3 Total 22 30 15 48 44 5 75 55 — 145 129 20 294 1943] Nigrelli & Atz: Biometry of Puffers and Parasites 3 Table III. Collection of Puffers: Relation of Size to Sex. Size in cm. M F I Total 2 — 1 1 3 — — 2 2 4 — — 2 2 5 — — 1 1 6 — — 2 2 7 8 9 — — 4 4 2 2 10 4 — — 4 11 10 4 2 16 12 7 9 4 20 13 13 5 — 18 14 18 4 — 22 15 21 7 — 28 16 25 7 — 32 17 20 14 — 34 18 10 18 — 28 19 14 18 — 32 20 3 16 — 19 21 — 10 — 10 22 — 8 — 8 23 — 8 — 8 24 — 1 — 1 Total 145 129 20 294 eggs hatch into infective stages (copepodid) which attach themselves immediately to the i gills. It is obvious that if sufficient numbers are present on the delicate gill filaments in- terference with the normal respiratory func- tion will occur. B. Internal Parasites. 1. Trematodes. All the trematodes recovered from the puffers are digenetic. Of the five species found in the Sandy Hook puffers, Lintonium vibex (Linton) (Stunkard & Nigrelli, 1930) and Bianium plicitum (Linton) (Stunkard, 1931) are the only ones found in appreciable numbers and with any degree of constancy. They occur in the digestive tract and were present, respectively, in 66% and 67% of the fish examined. Lintonium vibex appears to be species- specific. It has not been recorded from any other fish occurring naturally in the North Atlantic. Linton (1940) reported an infes- tation in a queen triggerfish. Batistes vetula, which had straggled northward to the Woods Hole region during the summer. The triggerfish probably became infected when it migrated into the northern waters. Fur- ther, this species, examined for internal parasites from its natural habitat, has not been reported as a host for Lintonium. Close relatives of the puffer, such as the rabbit fish or smooth puffer, Lagocephalus laevi- gatas, and the common spiny boxfish, Chilo- myeterus schoepfii, taken at Sandy Hook Bay at the same time that puffers were seined, were free of this trematode. Both of these fish have feeding habits similar to the common puffer. Bianium plicitum is usually present in puffers in large numbers. The parasite was originally described as an unidentified spe- cies by Linton in 1898 from two specimens found in the smooth puffer and the tom cod, Microgadus codus, and again in 1929 as Psilostomum plicitum from the intestine of the herring gull, Larus argentatus. It was shown by Stunkard (1931) that the two forms were identical with a species he had described in 1930 as Bianium concavum, a common intestinal parasite of the puffer and for this reason was designated B. plicitum. This important discovery shows definitely, what often has been suspected, that para- sites which are normally found in one host may persist for a time in its predators. The presence of Bianium in the smooth puffer and the tom cod may be accounted for in this manner. From the record so far pre- sented, it seems that Bianium plicitum also is a specific puffer parasite. Podocotyle olssoni Odhner, Lecitliaster confusus Odhner and Stephanostomum tenue (Linton) have been described from a variety of North Atlantic fishes. Their life-histories were discussed above, and although together they were found in 8% of the puffers, indi- vidually there were never enough present at one time to make them significant. Cymbephallus vitellosus (Linton) is an- other intestinal fluke reported for puffers from Woods Hole region. It is closely re- lated to Podocotyle and like this form is found in a large number of fishes in this locality. 2. Cestodes. The striking feature of the present studies is the dearth of cestodes. A single, un- identified scolex was found. Linton (1924) reported more than 24 encysted and free scoleces from about 20 puffers examined from Woods Hole. From this material he was able to identify Rhynchobothrium bul- bifer, R. tumidulum and Tetrarhynchus bi- sulcatus. These identifications are incom- plete because, as pointed out by Nigrelli (1938 ) , the names Tetrarhynchus and Rhyn- chobothrium have no generic status. They are group names. Tetrarhynchus bisulcatus Linton has since been redesignated Nybelinia bisulcata (Linton) Dollfus. Larval cestodes occur in numerous teleost hosts and al- though their life-cycles are not completely understood, it is known that they become sexually mature in the spiral valve regions of the intestines of elasmobranchs. 3. Nematodes. Thirty-six per cent of the puffers were infested with nematodes belonging to the genera Contracaecum and Porrocaecum. No 4 Zoologica: New York Zoological Society [XXVIII: 1 Table IV. List of the Parasites Collected from Puffers of Sandy Hook Bay, N. J. Parasite Site of Infestation M F I Total Ratio of Infect. % 1. Oodinium ocellatum (Dinoflagellate) Gills and Skin — 2 — 2 .7 2. Trichodina spheroidesi 3. Trichodina halli (Ciliate) Gills and Skin ) Gills and Skin j 126 113 6 245 83.0 4. Myxosporidian (sp. ?) (Cnidosporidia) Gills — 1 — 1 .3 5. Lintonium vibex (Trematode) Pharynx 101 82 — 183 66.0 6. Bianium plicitum (Trematode) Intestine 98 98 — 196 67.0 7. Podocotyle olssoni (Trematode) Intestine 9 5 — 14 5.0 8. Lecithaster confusus (Trematode) Intestine 3 1 — 4 1.4 9. Stephanostomum tenue (Trematode) Intestine 1 4 — 5 2.0 10. Tapeworm scolex (sp. ?) Intestine 1 — — 1 .3 11. Contracaecum sp. (Nematode) Body Cavity 29 30 — 59 20.0 12. Porrocaecum sp. (Nematode) Intestine 24 23 — 47 16.0 13. Pseudochondr acanthus diceraus Gills 106 98 — 207 70.0 (Copepod) specific identification was made because they are difficult to distinguish from any one of a number of species reported from marine fishes. Both forms are larval in character. Contracaecum is invariably found coiled in delicate cysts in the retroperitoneal regions of the liver, intestine and other organs. Sex- ually mature stages are known to occur in piscivorus birds and mammals. Porrocaecum was found free in the intestine of puffers. These probably become mature in elasmo- branchs, marine turtles or seals. The acanthocephalan, E chinorhynchus acus, reported by Linton (1901) from the pharynx of a single puffer, was not present in our material nor has it been recorded from puffers since that time. This must be an accidental infestation. Effects of Internal Parasites on the Host. There was no evidence to indicate any real pathological effects in instances of severe infestations with the internal parasites just described. The presence of large numbers of flukes must without doubt elicit some physiological response not often obvious in the course of autopsy. In certain instances it has been noted that intestinal flukes in fishes are capable of causing an enteritis (Nigrelli, 1940), excess production of mu- cus and a superficial necrosis of the mucus coat. It is interesting to point out that in in- festations with Bianium, all of the indi- viduals appeared to be at about the same stage of development (as indicated by size), suggesting that puffers obtain a single dose. On the other hand, individuals of Lintonium were found varying in size from minute, hardly perceptible forms to large, sexually mature specimens. This shows that puffers become infested continuously with Linton- ium and that an initial infestation does not confer any resistance to the host. Whether or not the one infestation with Bianium is sufficient for puffers to develop an immunity against a further attack of this parasite is not known. That the infestation is less stable than that of Lintonium is demon- strated by the observation that puffers sub- jected to a period of captivity lose most or all of the Bianium while Lintonium may persist indefinitely. Occurrence of Parasites of Puffers According to Dates of Capture. Table VII summarizes the data regarding the number of parasites collected from the 294 puffers examined. As may be seen, the intensity of infestation reaches a peak in July and sharply drops off in August. For the first two months, 109 fish were examined in the three years from which 5,022 speci- mens of the three species were counted. In the last three months 185 fish examined yielded only 1,328 parasites. Although 25% 1943] Nigrelli & Atz: Biometry of Puffers and Parasiftes 5 Table V. Intensity of Trichodina Infection in Relation to Sex. Intensity of Infection" Male Female Immature Total % Exceptionally Heavy 7 8 15 Very Heavy 5 10 1 16 Heavy 18 18 36 23.00 Medium 9 6 15 Light 29 29 1 59 Very Light 19 20 3 42 39.00 Exceptionally Light 39 22 1 62 Negative 19 16 14 49 38.00 Total 145 129 20 294 100.00 * Exceptionally Heavy, too many in the field to count; Very Heavy, about 100 ciliates in each field; Heavy, about 50 in each field; Medium, about 25 in each field; Light, about 12 in each field; Very Light, 1 or 2 in each field; Exceptionally Light, 1 or 2 in an occasional field. Note : Some of the last mentioned may be contamination. more fish were examined during the latter period, the parasite fauna was 60% less. It should be indicated here that during June and July all females examined were gravid while most fish caught in August and Sep- tember had spawned. The great decrease in the numbers of parasites, during the latter period, as shown below, may be attributed partly to the spawning activities of the fish. Occurrence of Parasites in Relation to Size and Sex of Puffers. Table VIII shows the total number of parasites, excluding the protozoa, in rela- tion to host size. It will be seen that the number of parasites increases rapidly with the size and age of the fish. Conversely, puffers up to 10 cm. in length were par- ticularly free of parasites. A total of 6,520 parasites were counted. 66% of these were taken from female puffers, although there were fewer females than males in the total catch. This does not necessarily mean that females are more susceptible because of their larger size. Table IX shows the rela- tion of the infestation to sex, according to dates the fish were caught. In this series a total count of the three major parasites was made. The figures show that the fe- male puffers are susceptible to heavier in- festations. This is partly corroborated by the data given in Table X showing fish infected simultaneously with the three major parasites. It will be noted that in puffers measuring from 16-19 cm., inclu- sive, the number of males and females col- lected was about equal (33 and 32 respec- tively), yet the females had at least 50% more parasites than the males. The greater susceptibility of females to parasitic infes- tation may be attributed in part to their more intensive feeding preparatory to spawning. This Table also shows that Bian- ium is present in larger numbers. Puffers Simultaneously Infested with the Three Major Parasites (Excluding Protozoa). An analysis was made to determine whether or not a heavy infestation of one kind of parasite would confer on puffers a certain amount of resistance to other types. Table X shows that this does not happen. About 45% of the fish were simultaneously infested with Pseudochondr acanthus, Lin- tomum and Bianium. Furthermore, this 45% was infested with about 84% of these parasites. Table VI. Intensity of Trichodina Infection in Relation to Dates. Month Av. Temp.* Heavy to Except. Medium to Light Except. Light Total Heavy Infect. Infect. to Neg. Infect. June 63. OF. 10 10 5 25 July 68.0 23 33 28 84 August 72.0 8 20 31 59 September 70.0 25 53 45 123 October 66.0 1 — 2 3 Total 67 116 111 294 * Average temperature in Bay off Battery Park for years 1938, 1939, 1940. 6 Zoologica: New York Zoological Society [XXVIII: 1 Table VII. Number of Parasites Collected (Three Species Only), According to Dates. Month Copepod Lintonium Bianium Total June 219 129 1108 1456 July 917 433 2216 3566 August 195 192 263 650 September 161 248 268 677 October — 1 — 1 Total 1492 1003 Table VIII. 3855 6350 Size of Host and the Total Number of Parasites Collected from Puffers (Excluding Protozoa). Size in cm.* Male Female Immature Total 2 — — — — 3 — — — — 4 — — — — 5 — — — — 6 — — — — 7 — — — — 8 — — — — 9 — — 2 2 10 62 1 — 63 11 151 1 1 153 12 111 55 4 170 13 132 198 — 330 14 175 65 — 240 15 199 92 — 291 16 265 38 — 303 17 327 190 — 517 18 230 698 — 928 19 493 784 — 1277 20 86 573 — 659 21 — 713 — 713 22 — 442 — 442 23 — 331 — 331 24 — 101 — 101 Total ♦Standard length 2231 4282 7 6520 Frequency Distribution of Parasites of Puffers, Excluding Protozoa, (Considered in Groups of 25). Table XI shows the frequency with which certain numbers of parasites occur in puf- fers when considered in groups of 25. As may be seen 11% of the puffers were free of infection. 63% of the fish had from 1-25 parasites. Fish with more than 100 para- sites were all females. However, they com- prise only about 3% of the catch. immature, although they may reach a length of 12 cm.; males measured from 10-20 cm. ; females from 11-24 cm. Puffers 21 cm. or more in length were all females. 4. Food eaten by puffers consists of cope- pods, amphipods, crabs, shrimps, bi- valve and univalve mollusks, annelids, fish and sea-weeds. The life-cycle of internal parasites, particularly the flukes, involves any one of these as intermediate host. Summary and Conclusions. 1. The parasitic fauna of puffers from Sandy Hook Bay, N. J., were studied. 2. Two hundred and ninety-four fish were collected in the months from June to October inclusive, for the years 1938, 1939, 1940. The largest number of fish were caught in September for the three 5. Thirteen kinds of parasites were found : 4 species of protozoa, 5 species of di- genetic trematodes, 1 tapeworm, 2 ne- matode species and 1 copepod. The protozoans and the copepods are gill parasites, heavy concentrations of which will cause death of puffers and other marine fishes. years. 3. The size range of the host extended from 2-24 cm. Fish 9 cm. or less were 6. In reference to numbers, the important metazoan parasites of the puffer are the flukes, Lintonium vibex and Bian - 1943] Nigrelli & Atz: Biometry of Puffers and Parasites 7 ium plicitum, and the copepod, Pseudo- chondr acanthus diceraus. 7. Both Lintonium vihex and Bianium pli- citum are specific to puffers of the North Atlantic coast of the United States. 8. Fishes related to the puffers (rabbit fish and queen triggerfish) migrating into this area may become infested with these flukes. However, other relatives of puffers, such as the spiny boxfish, living in the same area and apparently with the same feeding habits, seem to be resistant to Lintonium and Bianium. 9. From observations, it appears that puffers are infested continuously with Lintonium while in the case of Bianium only one infestation occurs at a time. 10. In captivity, the puffers lose their Bianium after a short period while Lintonium may persist indefinitely. 11. Other species of flukes recovered from Sandy Hook Bay puffers have also been reported from a large variety of fishes of the North Atlantic. In some cases, these parasites may be accidental infes- tations. 12. Although a number of species of ces- todes have been described from puf- fers, only a single unidentified scolex was recovered from Sandy Hook Bay fish. 13. The two species of nematodes found were larval in character and presum- ably sexually mature forms come up in piscivorous hosts. Table IX. Number of Parasites Collected (Three Species Only), According to Sex. Total for the Three Years Month Male Female Total June 374 1082 1456 July 1267 2299 3566 August 221 429 650 September 283 394 677 October 1 — 1 Total 2148 4204 6350 Table X. Puffers Infested Simultaneously with the Three Major Parasites (Excluding Protozoa). Size in No. of Inc cm. in each 10 4 11 16 12 20 13 18 14 22 15 28 16 32 17 34 18 28 10 32 20 19 21 10 22 8 23 8 24 1 M Copepod Lintonium. Bianium Total 280 Number of Parasites 0 1-25 26-50 51-75 76-100 101-125 126-150 151-175 176-200 201-225 226-250 251-275 276-300 Total 3 — — 43 4 15 5 — • - 27 26 51 2 4 67 34 31 3 4 59 8 100 4 — 48 45 42 6 3 58 42 105 8 1 66 23 93 8 5 90 72 261 7 11 89 135 513 10 15 205 190 697 3 9 83 72 424 0 8 85 33 565 6 189 30 205 6 46 34 202 1 5 1 93 61 71 1160 749 3397 Table XI. of Parasites of Puffers (Excluding Protozoa). ; In Groups of 2 Male Female Immature Total Percenta; 7 7 17 34 10.54 105 76 3 184 62.65 25 21 46 15.64 6 9 15 5.10 2 8 10 3.43 0 2 2 .68 0 0 0 .00 0 2 2 .68 0 2 2 .68 0 1 1 .33 0 0 0 .00 0 0 0 .00 0 1 1 .33 145 129 20 294 100.06 Zoologica : Neiv York Zoological Society 14. It has been shown that the intensity of infection in Sandy Hook puffers reaches a peak in July and sharply drops off in August. 15. It has also been found that the number of parasites increases rapidly with size and age of the fish. 16. Conversely, young and sexually imma- ture puffers are particularly free of parasites. 17. Female puffers are more susceptible than males to the metazoan parasites. Since the heaviest infestation occurred in June and July, it is assumed that this may be attributed to their more intensive feeding preparatory to spawn- ing. 18. There is no evidence to indicate that a heavy infestation of any one species of parasite confers resistance to the host against other species. It has been shown that about 45% of the catch was simultaneously infested with Pseudo- chondr acanthus, Lintonium and Bian- ium and that this 45% harbored 84% of these parasites. 19. About 63% of the puffers were infested with 1-25 parasites. All puffers har- boring more than 100 parasites were females. About 11% of the fish were free of parasites. 20. It is suggested, as a preventative meas- ure, that young puffers (measuring up to 12 cm.) or those collected late in the season (September) be used for exhibi- tion purpose since such fish are com- paratively free of detrimental para- sites which may spread to other and more important fishes. References. Breder, C. M., Jr. (and Collaborators). 1923-1939. Fish Notes from Sandy Hook Bay, N. J. Copeia, 1923-1939. Hunnien, A. V. & R. M. Cable. 1941 a. The Life History of Podocotyle ato- mon (Rudolphi) (Trematoda: Ope- coelidae). J. Para., 27:12. 1941 b. Studies on the Life History of Le- cithaster confusus Odhner (Trema- toda: Hemiuridae). J. Para., 27:13. Linton, Edwin. 1898. Notes on Trematode Parasites of Fishes. Proc. U. S. Nat. Mus., 20:507- 548. 1901. Parasites of Fishes of the Woods Hole Region. U. S. Fish. Com. Bull, for 1899, pp. 405-492. 1905. Parasites of Fishes of Beaufort, North Carolina. Bull. Bur. Fisheries, 34:321-428. 1924. Notes on Cestode Parasites of Sharks and Skates. Proc. U. S. Nat. Mus., 64:1-114. 1928. Notes on Trematode Parasites of Birds. Proc. U. S. Nat. Mus., 73:1-36. 1940. Trematodes from Fishes Mainly from the Woods Hole Region, Massachu- setts. Proc. U. S. Nat. Mus., 88:1-172. Martin, W. E. 1939. Studies on the Trematodes of Woods Hole. II. The Life Cycle of Stephano- stomum tenue. Biol, Bull., 77:65. Nigrelli, Ross F. 1936. The Morphology, Cytology and Life- History of Oodinium ocellatum Brown, a Dinoflagellate Parasite on Marine Fishes. Zoologica, 21:129-164. 1938. Parasites of the Swordfish, Xiphias gladius Linnaeus. American Mus. Nov., (996) :1-16. 1940. Mortality Statistics for Specimens in the New York Aquarium, 1939. Zoo- logica, 25:525-552. Nigrelli, Ross F. & G. M. Smith. 1938. Tissue Responses of Cyprinodon varie- gatus to the Myxosporidian Parasite, Myxobolus lintoni Gurley. Zoologica, 23:195-202. Padnos, Morton & Ross F. Nigrelli. 1942. Trichodina spheroidesi and Trichodina halli spp. nov. Parasitic on the Gills and Skin of Marine fishes, with Special Reference to the Life-History of T. spheroidesi. Zoologica, 27 : 65-72. Stunkard, H. W. 1930. Another Trematode with Two Anal Openings. Anat. Rec., 47 :363. 1931. Further Observations on the Occur- rence of Anal Openings in Digenetic Trematodes. Zeitschr. f. Parasit., 4:713-725. Stunkard, H. W. & Ross F. Nigrelli. 1930. On Distomum vibex Linton, with Spe- cial Reference to its Systematic Posi- tion. Biol. Bull., 58:336-343. Ward, H. B. 1908. Internal Parasites of the Sebago Sal- mon. Bull. Bur. Fisheries, 37:1153- 1194. Welsh, W. W. & C. M. Breder, Jr. 1922. A Contribution to the Life-History of the Puffer, Spheroides maculatus (Schneider). Zoologica, 2:261-276. Wilson, C. B. 1932. The Copepods of the Woods Hole Re- gion, Massachusetts. Bull. 158, U. S. Nat. Mus., 623 pp. Gordon & Flathman: Genetics of Melanoma in Fishes 9 2. The Genetics of Melanoma in Fishes. VI. Mendelian Segregation of Melanophore Reaction Types In Embryos of a Melanomatous Mother.1 2 3 * Myron Gordon" & Fred Flathman3 (Text-figures 1-3). The experimental production of mela- nomas in hybrid fishes, by genetic methods alone, was first discovered, not by geneticists nor oncologists, but by aquarists in 1912 and 1913. The early history of their discoveries has been pieced together by noting the de- tails presented in the minutes of their aquarium society meetings. In their early efforts to develop new and more colorful varieties of fishes for their home aquaria these fish-breeders crossed the black-spotted platyfish, Platypoecilus maculatus, with the swordtail, Xiphophorus hellerii. The aquar- ists were successful in breeding strikingly beautiful hybrids but the black-spotted ones developed melanotic tumors. The fish hybri- dizers have continued making new combina- tions. One of the more recent developments has been the black-banded hybrid with either yellow or bright red back, popularly spoken of as the “tuxedo swordtail.” The junior author bred some of these in his pri- vate aquarium; the mating record is given in the diagrams. One of the hybrids that de- veloped a melanoma was turned over to the senior author for further study. The Melanomatous Female and Its Brood. When the black-banded, yellow-backed fe- male swordtail hybrid was about 20 months old and 65 mm. long, it developed a mela- noma measuring 12 mm. X 7 mm. X 4 mm. on her left side. A smaller tumor appeared on the other side. At the time of its death it was gravid and when it was dissected, 40 embryos were found. Their age is estimated to be about 25 days, assuming that the aver- age gestation period is thirty days. 1 The authors wish to express their thanks to the De- partment of Birds of the American Museum of Natural History for the use of their laboratories in the Whitney Wing- where genetic and correlated studies of fishes are being continued. 2 Research Associate in Genetics, New York Zoological Society. ■5 Private Aquarist, Woodhaven, Long Island , New York. Under a binocular microscope, the em- bryos were sorted out into four genetic classes, as follows: 10, St N, gray-backed and well developed black bands. 9, st N, light colored backs and poorly de- veloped black bands. 12, St n, completely gray, like the wild ' type of swordtail. 9, st n, extremely light in color over all. This typical backcross ratio of 1 : 1 : 1 : 1 was expected on the basis of two indepen- dent factors : N for the black band, and St for many tiny black pigment cells which make the fish appear gray. Bellamy (1928) showed that the black-banded pattern of the platyfish may be referred to a sex-linked dominant gene, N, for “nigra.” Gordon (1931) confirmed this and added that St was a dominant autosomal factor for many small melanophores, micromelanophores, which have a “stippling” effect. When a fish has both dominant factors, St N it is black- banded and gray-backed. When the fish is recessive for the stippling factor, st N, it is black-banded and gold-backed. The black band is composed of many large melano- phores, macromelanophores, while the stip- ple pattern is made of micromelano- phores. The smaller are not the young stages of the larger but are independently developed. The Early Development of the Black Band. It is significant that the four phenotypic categories, St N, st N, St n, st n, can be dis- tinguished in hybrid fishes in embryonic stages. Gordon (1931) pointed out in his study of the development of genetic pat- terns in the platyfish that in pure species, the black-banded type, N, could not be dis- tinguished until the birth of the fish and even at this comparatively late stage, the N pattern is represented by only one or two discrete macromelanophores. It is therefore 10 Zoologica : New York Zoological Society [XXVIII: 2 Platyfish StSt NN aa bb Golden Swordtail Golden Swordtail X stst nn AA BB Stst Nn Aa Bb f st n A B F2 Back-cross hybrids Text-fig. 1. The genetic history of the four types of embryos. A black-banded, gray- backed platyfish female, StSt NN aa bb, was mated to a golden swordtail male, stst nn AA BB. These are shown on the top. One of their black-banded, gray-backed sons, Stst Nn Aa Bb, was mated to a golden swordtail female, stst nn A A BB; the second mating is shown on the second line. The offspring of the second mating were of four types: st n A B, golden; st N A B, black-banded, golden-backed; St n A B, gray; and St N A B, black-banded, gray-backed. The four types are shown on the third row and the figures represent dorsal views of the adults. The mother of the embryos (shown in another Text-figure) is the black-banded, golden-backed type, st N A B; the father is the gray-type, St n A B. The father and mother are shown within the brackets in row three. The offspring of this mating may be seen in Text-fig. 3 on the opposite page. 1943] Gordon & Flathman : Genetics of Melanoma in Fishes 11 Text-fig. 2. First generation hybrid. This hybrid is the product of mating a female black-banded Platypoecilus maculatus and the wild-type Xiphophorus hellerii. In Text- fig. 1 it is represented on the second row to the right, under the legend Fi hybrid. F3 25 day old embryo hybrids st N A B St n AB St NAB Text-fig. 3. Four genetic types of embryos. These embryos represent the offspring of the third generation, st n A B = golden; st N A B = black-banded, golden-backed; St n A B = gray; St N A B — black-banded, gray-backed. The black band is much better developed in the St N A B type than in the st N A B type, indicating an inter- action of St and N genetic factors. (The authors wish to thank Mr. Jack Beckenstein for his aid in the preparation of these charts.) quite apparent that the rate of growth of the macromelanophores is much accelerated in the hybrids. This marked increase in the rate of development is brought about by two genetic modifiers, A and B of the swordtail. Kosswig suggested this interpretation from his study of adult black-banded neoplastic hybrids. The change of the macromelano- 12 Zoologicci: New York Zoological Society phore habit of growth from the normal to the pathological has also been analyzed by Gordon in his study of the Sp gene which is closely related to the N factor: Sp brings about the spotted pattern in pure platy- fish and macromelanophore overgrowths in platyfish-swordtail hybrids. The Influence of the St Gene Upon N. The N gene is influenced by the micro- melanophore factor St in addition to A and B and this effect can only be detected in the embryos. For instance, in the gold-backed, black-banded hybrid embryos, st N, the macromelanophores are far less numerous, and form a much weaker black band, than in the gray-backed, black-banded type St N. This is illustrated by the diagram. A similar situation of interaction of fac- tors has been described by Gordon (1928) in two varieties of the platyfish: the gold, spotted platyfish, st Sp, had far fewer macro- melanophores than the gray, spotted type, St Sp. In this instance the differences per- sist throughout life whereas in the black- banded hybrid types mentioned above, and in black-banded pure platyfish, the adult st N type has as strong a black band as the St N type. Another instance of factor interaction in- volving the melanophore ratios may be found in the two genetic phases of the Mo gene in hybrids. The Mo is characterized by the presence of many macromelanophores arranged in lateral lines; usually the body is orange-red in color. Gordon (1938) has pointed out that the St Mo type is outstand- ingly more heavily spotted with macromela- nophores than the st Mo form; indeed, the st Mo has but one or two spots near the head. In platyfish-swordtail hybrids, macromel- anophores alone, st Sp, are able to evoke melanomas. Micromelanophores, alone, St sp, cannot do this. Yet micromelanophores ex- ert a definite force in intensifying the se- verity of the neoplasm when these cells are present together with macromelanophores. For instance, hybrids of the constitution St Sp develop their tumors earlier and faster than st Sp. This is also true in N hybrids : st N hybrids develop melanomas but the St N develop them first. An insight into the changing relationships between the two types of melanophores of a melanotic hybrid may be had by comparing the melanophores in st N embryos and in adults. In the st N embryo some micromelanophores are found along the dorsal ridge of the back and ex- tend over the meninges of the brain; the number of micromelanophores found is, of course, far less than in the St N embryo. In the st N adult, practically all the micromel- anophores have been eliminated by the hy- pertropic activity of the macromelanophores in the making of the black band. It appears that as the st N hybrid develops, its macro- melanophores usurp all the melanin-pro- ducing substances. Summary. The primary genic function of St is the formation of thousands of micromelano- phores. This clothes the platyfish or the swordtail with a grayish coat of pigment. The secondary effect of the St gene is to ac- celerate the growth-promoting forces of the macromelanophores both in normal and neoplastic stages of their development. For example, more macromelanophores appear at birth of a hybrid in a St Sp than in a st Sp ; more in a St A than in a st N. More macromelanophores are developed by the adult St Sp than by a st Sp platyfish ; more in the St Mo than in the st Mo swordtail derivative. Thus the St gene has a dual effect: first, in the production of many micromelano- phores and second, in the stimulation of the production of macromelanophores when the gene Sp, N or Mo is present. The growth promoting effect of St on Sp, N or Mo is limited; st does not have the power to shift the macromelanophore growth-pattern in pure species from the normal to the neo- plastic. The neoplastic activity of the Sp or N gene in hybrid fish combinations is con- trolled by factors A and B which are con- tributed by a foreign species. Perhaps at a future time, some definite physico-chemical process may be suggested in the activity of genes St, A and B, but at present such data are lacking. References. Bellamy, A. W. 1928. Bionomic studies on certain teleosts (Poeciliinae) . II. Color pattern inheri- tance and sex in Platypoecilus. Gene- tics, 12:226-232. Gordon, Myron 1928. Pigment inheritance in a Mexican kill- ifish: Interaction of factors in Platy- poecilus maculatus. Journal of He- redity, 19:551-556. 1931. Morphology of heritable color pat- terns in the Mexican killifish, Platy- poecilus. American Journal of Cancer, 15:732-787. 1938. The genetics of Xiphophorus hellerii: Heredity in montezuma, a Mexican swordtail fish. Copeia, 1938 (No. 1) ; 19-29. Kosswig, C. 1931. Zur Frage der Geschwulstbildungen bei Gattung-bastarden der Zahnkar- pfen Xiphophorus und Platypoecilus. Ztschr. f. ind. Abstamm. u. Vererb., 59:61-76. Beebe: Pattern and Color in Aequidens tetramerus 13 Pattern and Color in the Cichlid Fish, Aequidens tetramerus.' William Beebe Director, Department of Tropical Research, New York Zoological Society. (Plate I). [This is a contribution from the Forty-third or Venezuelan Expedition of the Department of Tropical Research of the New York Zoological Society made under the direction of Dr. William Beebe. The expedition was sponsored by grants from the Committee for Inter-Amer- ican Artistic and Intellectual Relations and from four trustees of the Zoological Society, George C. Clark, Childs Frick, Laurance S. Rockefeller and Herbert L. Satterlee, and by invaluable assistance from the Standard Oil Companies of New Jersey and Venezuela.] As a preliminary to the study of some Venezuelan fishes and especially their re- actions to the dry and the wet seasons, I wish to present a comparison of the pattern and color descriptions of one species of cichlid taken from ichthyological literature, with the twenty-four hour variation of a single individual of the same species. This fish is Aequidens tetramerus, origin- ally described by Heckel ( Acara tetramerus, Heckel, 1841, Ann. Wiener Mus., II, 341). It has a wide distribution, from Venezuela and the Guianas, on throughout the Amazon- ian drainage system to eastern Ecuador, and south to Brazil and Paraguay. Eight Descriptions Selected from Literature. “Color in alcohol, at the present time: in general chocolate brown; toward the back of the head almost black, lighter toward the breast; each scale with a light spot at its base. A black spot below the eye ; the lateral spot below the 9-1 1th scales of the lateral lines, and the tail spot encircled by bright white dots which also appear farther below on the base of the fin. The soft portion of the vertical fins has several diagonal rows of black dots. “Color in life: upper half dark greenish- brown becoming blue-gray below and violet toward the throat; sides of the head dark gray-brown, yellow-brown below ; each scale, especially on the lower half of the body, 1 Contribution No. 651. Department of Tropical Re- search, New York Zoological Society. lighter in the center than the ground color; the three lateral spots black and the spots surrounding the tail spot pure white. The spinous dorsal same color as the back; the soft rays and the caudal fin brownish-yellow ; the two last as well as the end of the anal fin dark greenish-brown, the pectorals brownish-yellow. The iris dark brown-yel- low, lighter below with some brown shad- ing.” (Translated from the German of Heckel : Wiener Museum D. N aturg eschichte Annalen, 1841, Vol. 2, p. 342). “Coleur de l’animal conserve dans la liqueur, d’un brun chocolat avec une tache ocellee sur la dos a la base de la nageoire dorsale; le ventre parait avoir ete plus clair et les opercules dores.” The plate figure differs considerably from this description, showing the iris bright yellow, body in gen- eral dark greenish-brown with an elongate ocellus on the upper part of the end of the peduncle. (Castelnau, 1855, Anim. Amer. Slid. Poiss., p. 15, pi. VI). “Brown, with a black spot on the middle of the sides and with a black ocellus edged with whitish superiorly at the root of the caudal fin, the two spots sometimes united by a band; a dark spot below the orbit; a whitish streak between the eye and the nos- tril.” (Gunther, 1862, Cat. Fishes Br. Mus., IV, p. 277). “Color brown, several darker cross shades across the dorsal region ; a large, black ocel- lus on lateral line. No other spots. Fins dusky. Lower lip yellow.” (Cope, 1870, Proc. Amer. Phil. Soc., XI, 570). “Olive, with jointed, unpaired fins yellow. A large black spot before middle of side on lateral line, a vertical shade at base of cau- dal. As compared with A. dimeras, its closest ally, this fish differs in . . . lack of cross- bars.” The figure shows no markings at all, except lateral spot. (Cope, 1871, Proc. Acad. Nat. Sci. Phila., XXIII, 255, pi. XI, fig. 4). “Olivatre avec une tache noire sur le cote et une autre bordee de blanc en haute de l’origine de la caudale; ces taches parfois 14 Zoologica: New York Zoological Society [XXVIII: 3 reunies par une ligne foncee. Un point noir sous l’orbite et une ligne clair entre la narine et l’oeil. Nageoires impaires habituellement ponctuees.” (Pellegrin, 1903, Mem. Soc. Zool. Fi ance, XVI, 171). “Olivaceous, with 5 dark cross-bars, the first crossing the anterior 7 or 8 scales of the lateral line, the second the 9th or 10th, and 3 succeeding scales of the lateral line, below which it bears a thickish blotch ; a dark spot below the posterior part of eye; a dark ocellated spot on the upper part of the base of caudal ; usually a dark longitud- inal band from lateral blotch to caudal spot, sometimes continued forward to the eye; vertical fins dusky, usually spotted.” (Regan, 1905, Ann. Mag. Nat. Hist., (7) XV, 332). “The pond specimens deeply colored. They show faintly the vertical color bands and a lateral line from the eye to the caudal basis. This interrupted by the lateral spot. Color notes made in the field on the Puerto Ber- mudez specimens, show the dorsal half of the body deep blue, shading to gray ventral- ly. Ocellated spot on the peduncle, bordered with yellow; irregular blue-black markings on the caudal fin; fins washed with yellow; five or six longitudinal stripes of yellow on the head from the snout to the eye and cheek. In alcohol a longitudinal color band can be seen in many.” (Eigenmann & Allen, 1942, Fishes of Western South America, p. 392). Comments on Aquarium Specimens. This fish is fairly well known to aquarists and we find descriptions such as the follow- ing : “Color changing; upper half green to greenish brown, bluish gray to reddish be- neath, throat violet; sides with about six cross-bars, one of which intensified into a spot below the lateral line; sometimes a dark band from eye to an ocellated spot at the base of the upper caudal rays. Vertical fins spotted or barred ; dorsal in its spinal part and anal dark greenish brown ; soft dorsal and caudal yellowish. The color markings of the body become obscure with age, those of the fins more intense.” (Rachow, 1937, Trop- ical Aquariafish, p. 111). “Color and markings vary greatly with age and the influence of local conditions but usually it is light green to olive brown, often with a golden or brassy tone. The sides are yellow to slate and the belly is light gray with a rose tint. A black horizontal stripe extends from the eyes, through the middle of the sides, to the caudal spot on a yellow field. Three or more, light to dark brown vertical bands run through the eyes and upper section of the body, enhancing the large area behind the eyes and another under the dorsal fin. The anterior parts of the dorsal and anal fins are dark olive and these fins as well as the caudal, are peppered with dark spots and short lines. The body markings of young specimens are quite pro- nounced. More mature specimens lose the body markings and those in the fins are more pronounced.” (Stoye, Tropical Fish for the Home, p. 168). “No one phase seems to give an adequate representation. A slight change in light pro- duces an entirely different effect, and as we all know, the various Cichlids have a tre- mendous range of color pattern, due to the influences of emotion, temperature and health.” (Innes, 1934, The Aquarium, II, p. 277). Comparison of the Eight Descriptions with the Color Phases of a Single Individual. In these eight selected descriptions, ex- tending from Heckel in 1841 to Eigenmann and Allen, 1942, we have a cross section of pattern and color as observed and recorded by these authors. We find the following re- lative mention of five elements of the pat- tern : Lateral ocellus 8 times, caudal ocellus 7, longitudinal stripe 4, vertical bands 3, sub- ocular spot 4 times. It is interesting to compare this with the seven phases of a single individual in the accompanying plate, figure one of which is diurnal, and all the others nocturnal. Here, the proportion is, lateral ocellus 7, caudal ocellus 3, longitudinal stripe 4, vertical bands 3, and sub-ocular spot 2 times. Color Description of Plate. It has been possible to reproduce the ac- companying plate only in black and white, so a few notes on the lost colors are neces- sary. There are seven figures, all drawn from the same living fish within a period of twenty-four hours. This fish was taken from an almost dried-up mud hole at the edge of the jungle at Tenth Kilometer, Caripito, Venezuela, March 21, 1942. Dep’t. Tropical Research Cat. No. 30,006; standard length 75 mm. Figure 1 shows the typical diurnal pattern and coloration. This is dark brown above and turquoise on the ventral surface. The lateral ocellus is very black. There are five or six, greenish-brown, narrow, longitudinal stripes along the body, alternating with equally narrow stripes of grass green. The opercles are densely, the sides of the body more sparsely, dotted with turquoise blue. The belly and abdomen and the first pelvic ray are solid turquoise, and the sides of the head are alternately streaked and dotted with the same color. The fins are brownish, the ver- tical fins faintly barred and spotted with darker. The edge of the entire dorsal fin is warm orange. 1943] Beebe: Pattern and Color hi Aequidens tetramerus 15 Figures 2 to 7 show various phases of the pattern as observed at night. Brown is the dominant color, the only exceptions being the black of the ocelli, and the turquoise spots, of which traces remain on the cheeks and opercles, and in all cases the blue of the pelvic ray. All other colors of the diurnal phase are absent. Pattern in Five Preserved Specimens. Of five other specimens collected in Ven- ezuela, two came from the dried pool at Tenth Kilometer, two from a small muddy creek known as Rio San Pablo, and one from the clear waters of Rio Caripe. In death all are almost identical, although one was killed in alcohol, two leaped out of their aquarium the first night and died slowly, and the fifth was killed after a week in captivity. All show the heaviest, most relaxed type of pigmentation, closely approximating Fig- ure 5 on the plate. The whole of each fish, however, is much darker, the vertical bands considerably wider and blacker, the lateral ocellus being almost buried in its particular band. As in the fifth pictured phase, there are nine vertical bands, with a consistent branching of the eighth (counting from the caudal). In two fish, the displaced nuchal band is continuous with the rest of the ninth through eye and opercle. The caudal ocellus is indicated only by a narrow vertical mark, which I call band number one. The lateral ocellus is always in the sixth band from the tail. The boundary of the lateral ocellus is clearly marked in all the fish by a circle of turquoise dots. Others can be seen scattered anteriorly over the side of the body, and many on the opercles. No hint of the blue pelvic ray is visible, in any of the preserved fish. Venezuelan Field Notes. The only place I could observe this species under natural conditions was in the Rio Caripe before its waters were muddied by floods. In the dried jungle pools and the small San Pablo the fish were invisible until dug up with a spade or brought to the sur- face in a hand net. In the Caripe they were abundant, living in loose schools in the back water above riffles. As I waded about I could distinguish them from other cichlids by the conspicuous black lateral spot. This seems the only pattern element which is present under all conditions. Several times when fish escaped from the seine and rejoined their free fellows, they became at once in- distinguishable, having undergone no ap- parent pattern changes as a result of fear or excitement. A number of specimens were kept in a small air-supplied aquarium in the labora- tory, past which several of us walked many times a day. These showed almost no change during the hours of daylight, and even when excited by the sudden appearance of food, the lateral spot was the only dominant mark. Pattern Change in a Single Individual. As already mentioned, Figure 1 shows the typical diurnal pattern, as observed in un- disturbed and disturbed fish in their normal haunts and in the aquarium. At night, whether in the dimmest light, or in the il- lumination from a powerful flashlight, the pattern came and went in bewildering com- binations of the four shifting pattern ele- ments, caudal ocellus, longitudinal stripe, vertical bands, sub-ocular spot, in intricate changes around the permanent lateral ocel- lus. Extremes are shown by Figures 2 and 5. In the former, the fish for the most part is white, shading into pale brown on head and back. Figure 6 shows the two ocelli and the eye all connected by a longitudinal stripe. Figure 3 has the sub-ocular spot, and a slight discontinuity of the stripe. In Figure 7 the stripe ends at the lateral ocellus, the posterior half and the caudal ocellus having vanished. Three vertical half bands are present on the postero-inferior quarter of the body. Figure 4 has vertical bands 1, 2, 3, 4, 5 and 7 developed (counting from the caudal), with bands 6, 8 and 9, and the post- erior half of the longitudinal stripe lacking. Figure 5 is a pale phase of what persists in all preserved specimens, the most intense amount of pigmentation, lacking only the narrow longitudinal stripes of the diurnal pattern and the caudal ocellus. In this phase even the iris of the eye changes its pigmen- tation to emphasize and complete the ninth band on the lower part of the head. Pattern of Other Living Individuals. Another individual fish of the same species which I examined six times in one night, showed phases approximating Figures 4, 5 and 7, with the vertical bands always in evi- dence, greatly diminished in only one case. A third fish at 11 P.M. and 2 and 3 A.M. showed no trace either of caudal ocellus or longitudinal stripe. This brief treatment is meant only to stimulate, and to emphasize the value of future observation in the field, both by day and night. The nocturnal changes in pattern and color, as I have often observed in tide- pool fish" are something to be reckoned with in our descriptions both of well known and of newly discovered species, and in ex- planation of ecological adaption. In the pres- ent instance I had no opportunity to obtain detailed data on specific causes, whether psychological, sensory or environmental, of - Book of Bays , page 169. 16 Zoologica : New York Zoological Society these changes, but if our keys and careful descriptions dealing with pattern and color of species and sub-species are to withstand both laboratory, aquarium and field scrutiny, individual changes must be taken into ac- count, and reduced to correct values in diag- noses. As one last example, Eigenmann, in the description of a new species, writes' Aequidens potaroensis is “evidently very closely related to A. tetramerus from which it can easily be distinguished by the dark band from the eye through the angle of the opercle.” On this diagnosis alone it appears in the plate as Figure 5, at least on one night, when it became and quite departed from the said species within a period of ten minutes. My sincere thanks go to Miss Francesca LaMonte for graciously given assistance. EXPLANATION OF THE PLATE. Plate I. The Venezuelan Cichlid fish, Aequidens tetra- merus. Seven pattern and color phases of a sin- gle individual fish, within twenty-four hours time. Figure 1 is the diurnal pattern, all others nocturnal. From a painting by George Swan- son. :: Eigenmann, Fishes of British Guiana, 1912, p. 490. BEEBE. PLATE [. PATTERN AND COLOR IN THE CICHLID FISH, AEQUIDENS TETRAMERUS. Delacour: Revision of the Pycnonotidae 17 4. A Revision of the Genera and Species of the Family Pycnonotidae (Bulbuls). Jean Delacour. (Plates I & II; Text-figures 1-13). The bulbuls constitute one of the most clearly defined groups of perching birds (Passeres). This means that the different genera and subgenera which belongs to the group are obviously related to one another, and rather far removed from any others. A few other genera, however, which may be fairly closely related to them, are difficult to classify in the present state of our knowl- edge. In my opinion, however, bulbuls do not belong to the large Muscicapide family group ( Muscicapidae ) , of which thrushes and robins (Turdinae), warblers (Sylvii- nae), babblers (Timaliinae) and a few others are evidently members. They are too different from any of these in their general structure, proportions, feathering, behavior and habits, and there do not seem to be any definite links with them, no more than with some other very different families. As it so commonly happens with families extending over two or more continents and numerous islands, bulbuls have been studied mostly within the artificial limits of local avifaunas, and their general grouping has been often incomplete and fragmentary, their true affinities being ignored. Major studies and lists of Pycnonotidae can be found in Sclater’s Systema Avium Aethiopi- carum, 1930; Chasen’s Handlist of Malay- sian Birds, 1935; Mathews’s Systema Avium Australasiarum, 1930; Hachisuka’s Birds of the Philippine Islands, 1935; Stuart Baker’s Fauna of British India Birds, 1922- 1930; Bannerman’s Birds of Tropical West Africa, 1936; and my own Les Oiseaux de l’lndochine Fran^aise, 1931. All these works, however, deal only with local avifaunas. In 1934-36, the late Baron Snouckaert van Schauburg published a distributional list of the Asiatic and East Indian Pycnonotidae, detailed but incomplete, and very uncritical. Dr. H. Oberholser in 1900 {Trans. U. S. Nat. Museum, XXI, p. 30, and XXII, p, 15), and in 1903 ( Smithsonian Misc. Coll., XLVIII, 1903, p. 155) studied different groups of African bulbuls and some Indian and Moluc- can forms, describing a number of new genera, only a few of which can be recog- nized. Finally, one must mention a study of the so-called Timaliine birds of Madagascar by Finn Salomonsen {Ann. & Mag. N.H., Ser. 10, Vol. XIV, p. 60, 1934) where a number of forms are included which, in my opinion, belong to the Pycnonotidae. The classification proposed below is more or less at variance with that adopted in these various works. Without attempting a complete review of all the different subspecies and their local distribution, I thought it might be useful to endeavor to draw a better picture of the different genera, subgenera and species in- cluded in the Pycnonotidae as I understand them. The following conclusions are based on the examination of the large and excellent collection deposited at the American Museum of Natural History. I have been greatly helped by the work of Drs. E. Mayr and James Chapin in rearranging them. In the field, I have had a good deal of experience with the Asiatic, Malay and Madagascan species, and also with some of the African ones. The observation of these birds in life and the study of their general behavior has given me a better understand- ing of their characteristics. This has been supplemented by the study of many live forms in captivity in my former collections and in others, public and private. Bulbuls are of moderate size, varying from that of an English sparrow to about that of an American robin. They have rather short wings and a comparatively long tail; their beak is never large, but is slender to moderately thick, notched, with rictal bris- tles usually well developed ; nostrils are long or oval, more or less operculated. Their legs and feet are always rather weak, usually small and often very short. The following characteristics are peculiar to the Pycnonotidae: the body feathers are long, soft and fluffy, particularly on the lower back. The skin is thin and the neck is 18 Zoologica : Neiv York Zoological Society [XXVIII: 4 very short and more or less devoid of fea- thers behind, so that a bulbul with an out- stretched neck shows a somewhat bare patch between the nape and upper back feathers, a feature quite characteristic of the group. There are always some hair-like feathers on the nape, often long and conspicuous, some- times weak and difficult to detect. This is, however, not a diagnostic family charac- teristic as it is also found in different other birds. Most of the bulbuls are olive-green, yellow or brown, with or without white, gray, black or yellow marks. A number have lightly colored under tail-coverts, varying from red to yellow, white or rufous, and some have bright yellow or scarlet spots or streaks on the head, and conspicuous ear- coverts. Many bulbuls are crested, but the presence or the absence of a crest and its relative length and shape is of little impor- tance. Frequently this does not constitute even a specific character, since in several cases some subspecies of the same species are crested while others are not. Several groups have also lengthened, shiny and orna- mental throat feathers, white, yellow or red- dish-copper color. Others have long upper tail-coverts and lengthened feathers on the rump and lower back, ornamented with spots and marks. In all bulbuls both sexes are alike in plumage and often also in size, but in a few cases the male is much larger than the female. Immatures are not spotted and differ but little from the adults. In these points, they widely differ from the flycatchers, thrushes and robins. Many live in forests and their edges, others in sparsely wooded or bushy surroundings and quite a number frequent cultivation and gardens. They are noisy, their notes being always short and loud, either melodious or harsh according to genera and species. Many can be con- sidered good songsters. They generally live in groups, outside of the breeding season, some species in large flocks, and they mix with other birds in wandering hunting par- ties. Most bulbuls are fruit and berry eaters, to which they add insects, while a few are mainly insectivorous. The latter are found among the strong or compressed bill forms. The majority of species are sedentary, but some migrate, the northern breeders in Asia to a great extent; they are the longer- winged forms, particularly Microscelis. All bulbuls build cup-shaped nests and lay heav- ily marked eggs. Representatives of the Pycnonotidae, as defined above, are found all over Africa, in Madagascar and the Mascarene Islands, Asia, the Philippines, the Malay Archipelago and the Moluccas. Of thirteen genera, eight are confined to Africa ( Calyptocichla , Boeopogon, Ixonotus, Thescelocichla, Chlorocichla, PhyUastrephus, Bleda, Nicator), three to Indo-Malaya ( Spizixos , Setornis, Microscelis) , while two ( Pycnonotus and Criniger) are found in both regions. In Madagascar one finds spe- cies of the African genus PhyUastrephus and a form of the species Microscelis mada- gascariensis, which is represented in the Mascarene Islands and from India to China and Indo-China. A few isolated and puzzling genera have a much discussed and still uncertain posi- tion : The curious Hapalopteron familiare from the Bonin Islands, usually considered a bulbul, is so far removed in proportions and plumage pattern that it is better rejected from the group. It may be an aberrant Zos- ter ops. Tylas eduardi from Madagascar, also usu- ally considered a bulbul, does not fit in well, widely differing in the proportions of the wings and tail, the shape of bill, the pattern and color of its plumage. In habits it is also peculiar, living in tree tops of forests and being slow in movements, reminding one somewhat of the Asiatic Cochoa. It is purely insectivorous and may be a true Turdine bird. The African Hypergerus atriceps has fea- thers somewhat recalling those of the Pycnonotidae, but its long, decurved beak, its high legs and its ground habits make it unlikely to be a true bulbul. The proportions and color pattern of another African bird, Neolestes torquatus, make it difficult to accept it in the present group, although it has some characteristics suggesting affinities and is certainly not a shrike. It may be a very aberrant Pycnono- tus. Only detailed studies of the anatomy, habits and behavior of these four peculiar birds can decide whether or not they can be included in the Pycnonotidae. Until this is accomplished it is best not to consider them as true bulbuls. They are not discussed any further in this paper. Several other genera have been assigned by different authors to the Pycnonotidae. ; Such are, on one side, the fairy bluebirds {Irena) and on the other the leaf birds and ioras {Chloropsis, Aethorhynchus and Aegi- thina) . They all differ greatly from the true bulbuls in plumage texture, pattern and col- ors, while males, females and young are conspicuously unlike. The proportion of the tail and wings, their legs and feet, are also very different. They are better considered as forming separate families, possibly not very far from the Oriolidae. But this still remains an open question. On the other hand, I have included with the Pycnonotidae several Madagascan birds classed in the Timaliidae by many authors under the generic names of Bernieria , Cross- leyia, Xanthomixis and Oxylabes. In fact, 1943] Delacour: Revision of the Pycnonotidae 19 they cannot be separated from the African Phyllastrephus in which I place them. “Ber- nieria” madgascariensis and “B.” zosterops are evidently quite close to such species as Phyllastrephus icterinus and P. xavieri; all of them show an extraordinary difference in the size of males and females, a very peculiar feature. The so-called “Oxylabes” tenebrosa, xanthophrys and cinereiceps, although aber- rant, rare and restricted in distribution, are certainly nearer to Phyllastrephus than to Oxylabes madagascariensis, a true Timaliine bird, completely different in proportions, shape and color pattern, with a much thicker bill and shorter wings, as is also Mystacornis crossleyi, both of which live on or near the ground. Close resemblance between Mada- gascan Bernieria and African Phyllastre- phus has been long ago pointed out, particu- larly by Hartlaub (Vog. Madag., 1877, p. 144), by Milne-Edwards and Grandidier (Hist. Phys. Madagascar, Vol. XII, p. 348) and more recently by J. Chapin. That it has escaped others shows once more the danger of the separate study of local avifaunas. The bulbuls, that is to say, the species which show the general characteristics which I have mentioned above, can in my opinion be divided into thirteen genera, a very much smaller number, of course, than is usually accepted1, and some can con- veniently be subdivided into subgenera. Their relationships are more or less close 1 Sclater (Systema Avium Aethiopicarum ) recognizes 24 genera (including Bernieria and allies, and Nicator , not classified by him as Pycnonotidae), while Snouckaert (Orgaan Nederl. Vogelk.) recognizes 21 for Asia and Malaysia (not including Chloropsis, Aethorhynchus and Aegithina, leaving out — work not completed — many species of the Criniger-Microscelis group). Text-fig. 1. Relationship of the bulbuls. 20 Zoologica: New York Zoological Society [XXVIII: 4 and of four principal groups, the three more specialized ones can be linked to a central and more generalized one (see Text-fig. 1). The species are reduced to 109, an average of 8.4 per genus, but five of the latter remain monotvpic, while one includes 47, another 23. In defining these genera, an ensemble of characters of real importance has been con- sidered. Others have been disregarded. As I have said above, the presence or absence of a crest is a negligible feature as it varies within the same species (. Pycnonotus leu- cotis and P. dispar, for instance). Most im- portant are the general pattern and color of the plumage, the shape of the bill and nostrils, the length of the tarsus and toes. A few species show slight serrations near the tip of the upper mandible, close to the terminal notch and this has been considered important enough to separate them generi- cally. That it has no great significance is shown by the fact that such otherwise closely allied species as virens and mazukuensis, the first of which has serrations and the second none, would have to be placed in two allegedly different genera, Andropadus and Arizelo- cichla, which then become ridiculously het- erogeneous ; also one is at a loss to know where to put other forms with a certain roughening of the edge of the upper man- dible, which cannot be called serration. The general shape of the rectrices also is of value, but not their relative length, as more or less graduated or forked tails are found in subspecies of the same species ( Pycnono- tus flavescens and Microscelis madagascari- ensis, for example). Life habits, voice, be- havior, nesting and eggs all have to be taken into consideration and have as great a value as morphological characters. The least specialized group among the Pycnonotidae is the African genus Chloro- cichla, with moderately long, strong bill and legs, rather long wings and tail, colors varying from brown to bright olive-yellow, and a white or yellow throat patch. They are evidently linked to the short-billed Pycnonotus and allies ( Spizixos , Calypto- cichla, Boeopogon, Ixonotus) through the larger Andropadus; to the Criniger group (including Setornis and Microscelis ) and to Thescelocichla, and also to the straight- billed, long-legged Phyllastrephus, Bleda, and Nicator. The Pycnonotus group does not appear to have any near allies on its more specialized end, neither has the Criniger- Microscelis group, if, as I have said before, one does not believe in their close relation- ship with Irena, Chloropsis, Aethorhynchus and Aegithina. On the contrary the Phyllas- trephus-Bleda-Nicator group, less special- ized, with stronger legs and longer beak, has some resemblances to some Timaliine birds on one side, and to certain hooked- billed groups such as Campephagidae, Pri- onopidae, Vangidae and certain African genera so far included in the Laniidae. A. Pycnonotus GROUP. This first group of bulbuls is composed of numerous species, the greatest number of which cannot be generically separated and form the genus Pycnonotus, while a few are different enough to be retained as forming small distinct genera. They all have a comparatively short bill; the culmen is curved and not sharply ridged, the gonys almost straight, nostrils oval or elongated, operculate to different degrees. Rictal bristles rather weak. Tarsus short or moderate. Wings rather short, about equal to the tail, which is comparatively long, square, rounded or slightly graduated. It is in this group that the plumage is most varied or vividly marked, ranging from black to olive-green, yellow, white, gray, brown, sometimes brightened by red and yellow spots and patches on the head, while a num- ber have bright under tail-coverts, white, yellow or crimson. Although completely arboreal, the great- est number of the bulbuls of this group fre- quent bush or park-like country; many have become accustomed to garden life and are found in cultivated areas. Some, however, are forest birds. All these bulbuls have loud short notes, quite characteristic of the group. I. GENUS Spizixos. Blyth, 1845. Type: Spizixos canifrons. Bill short and very thick, whitish-yellow; plumage olive-green and gray. Crested or not. Text-fig. 2. Spizixos canifrons. 1. S. canifrons : Eastern Himalaya, Burma, S. W. China, Northern Indo-China. 2. S. semitorques : Central and Southern China and Northern Indo-China, Formosa. These bulbuls live at high altitudes in the south of their range among brush and near villages and on the edge of forests. They stand close to Pycnonotus, but their very peculiar beak and color entitles them to generic rank. S. canifrons is crested while S. semitorques is not. 1943] Delacour: Revision of the Pycnonotidae 21 II. Genus Pycnonotus. Boie, 1826. Type: Turdus capensis. Bill moderately thick and long for the group; colors and pattern very variable. Feathers on the crown erectile and either elongated or normal in length. Subgenus Euptilosus. Gray, 1853. Type: Brachypus eutilotus. Bill rather long and strong, rictal bristles long and black; feet dark gray or black. Wings rather long and tail graduated. Gen- eral color brown above and whitish below. Long crest of rounded feathers on occiput. Rump feathers ample and very long, marked with white and dark brown. All lateral rectrices with white tips. 1. P. eutilotus : Malay Peninsula, Sumatra, Banka and Borneo. 2. P. urostictus : Philippines and Borneo. P. urostictus has a fleshy bare ring round the eye, but otherwise resembles closely P. eutilotus; it cannot be subgenerically separated. They are birds of open ground and cultivated areas. Subgenus Microtarsus. Eyton, 1839. Type: Microtarsus melanoleucus. Bill moderately long and thick, black like the legs. Wings rather long and tail gradu- ated. General color black, marked with white on the wings. No crest. Rump feathers and upper tail-coverts very long. Lives in forest. 3. P. melanoleucus : Malay Peninsula, Su- matra, Siberut I. and Borneo. Subgenus Brachypodius. Blyth, 1845. Type: Turdus melanocephalus = atriceps. Bill, feet and legs small. General color bright olive-yellow or green, and black or gray. No crest, but feathers of crown erec- tile and rounded. Wings short; tail gradu- ated. Rump feathers and upper tail-coverts very long and marked with black. 4. P. atriceps : Assam, E. Bengal, Burma, Andaman Islands, West and South Indo- China, Malay Peninsula, Sumatra, Java, Borneo and neighboring islands, Pala- wan. Forms with gray underparts and back, in- stead of olive-yellow, are but color phases. 5. P. poiocephalus : S. India. Forest bulbuls, keeping to high trees. Subgenus Rubigula. Blyth, 1845. Type: Turdus dispar. Bill and legs rather short and weak, black or dark brown. Wings short; tail rather long, square or rounded. Upper parts bright olive-yellow, head black or gray ; underparts yellow, gray or mottled black and white. Crested or not. Iris crimson or yellow. 6. P. dispar : Western, Central, N. Eastern India, Burma, Yunnan, Indo-China, Siam, Malay Peninsula, Sumatra, Bor- neo and Java. 7. P. melanicterus : Ceylon. 8. P. squamatus : Malay Peninsula, Su- matra, Borneo, Java. 9. P. cyaniventris : Malay Peninsula, Su- matra, Borneo, These are all forest birds. I consider flavi- ventris (and its local races), mantis and gidaris as subspecies of dcspar; flaviventris is similar to dispar except for its pointed crest which becomes gradually shorter as it ranges south ; dispar has a shiny red throat and flaviventris has usually a black one, but speci- mens with a red throat, as well as intermedi- ates, are not uncommon in the southern and eastern parts of its range; gularis is but a small, duller form of dispar. Subgenus Otocompsa. Cabanis, 1851. Type: Lanius jocosus. Bill moderately strong, black; legs small and black. Wings short, tail long, square or rounded. Upper parts plain brown or green- ish-gray; crown black; ear-coverts forming patches; throat white; underparts whitish- brown, with or without yellow streaks; a black or brown broken collar on the upper breast, more or less definite; under tail- coverts crimson, orange yellow or yellow- ish-white. Crested or not. Iris brown. 10. P. jocosus : India, Burma, S. China, Indo-China, Siam, the Andaman and Nicobar Is., Malay Peninsula. Intro- duced into different other countries. 11. P. xanthorrhous: N. E. Burma, N. Siam, N. Indo-China, S. and C. China. 12. P. sinensis: E. C. and S. E. China, E., N. and C. Indo-China, Formosa, Hainan. All these bulbuls live in open country with scattered trees, clearings around cultivation and human dwellings. They all have similar habits and voices. They are familiar garden birds, particularly jocosus and s. sinensis. Northern forms of sinensis only are migratory. The general pattern of coloration is alike in all species, but jocosus and xanthorrhous have no olive green in their plumage, while sinensis is suffused with it; jocosus only has a long- pointed crest. That these species are closely related is furthermore shown by the following facts: there is a scarlet patch on the cheeks of jocosus, a smaller one at the base of the lower mandible in xanthorrhous, and a still smaller one at the same place in sinensis taivanus; it is absent in other races of sinensis, including hainanus; jocosus and taivanus have white ear- coverts, other races of sinensis brown or white, and xanthorrhous brown; jocosus has wide white tips to the rectrices, except the central pair, broader on the outside ones ; xanthorrhous has very narrow white tips, while sinensis has only indistinct pale edgings. 22 Zoologica: New York Zoological Society [XXVIII: 4 Subgenus Pycnonotus. Boie, 1826. Type: Turdus capensis. Bill and legs rather strong and black. General color brown or brownish-gray, dark- er above, lighter below, plain or mottled, exceptionally tinged with olive ( leuco - genys ) ; head black or brown with more or less elongated feathers on the crown forming a crest in some cases. Wings mod- erate; tail long and square, often tipped with white; under tail-coverts crimson, yel- low or white. Ear-coverts white, brown or black. These bulbuls are larger, longer and coarser than those of the Otocompsa sub- genus. 13. P. capensis : S. Africa. 14. P. nigricans : S. and S. W. Africa. Text-fig. 3. Pycnonotus barbatus tricolor. 15. P. barbatus : Africa, Arabia, Pales- tine, Syria, Cyclades I. 16. P. leucotis : Mesopotamia, Persia, Baluchistan, N. W. & C. India. 17. P. leucogenys : Himalaya and N. W. India from Afghanistan to Assam and north of the Brahmaputra to Dihang River. 18. P. cafer: Ceylon, India, Burma, S. China, Siam, Indo-China, Java. All these typical bulbuls are birds of cul- tivated and open countries, like those of the Otocompsa group, but they are stronger, with longer wings, broader tails and a more uni- form plumage. P. barbatus includes all African (Ethiopia and North Africa), Arabian, Palestinian and Syrian birds, which replace one another geo- graphically and have white under tail-coverts in the north and west, gradually changing to pale and bright yellow to the south and to the east, where they have light tips to the tail, black head and throat, thus approaching leuco- tis. On account of their very different color pattern, I consider leucotis and leucogenys two different species. P. cafer includes all the Asiatic forms with a white rump, almost black with crimson under tail-coverts in the west gradually changing to light gray in the east. Isolated forms in S. Siam, S. Indo-China and Java ( thais , ger- maini and aurigaster) are pale, with yeilow under tail-coverts, but evidently belong to the same species and have nothing whatever to do with xanthorrhous, which they only super- ficially resemble. Some birds from western Siam have orange under tail-coverts, thus pro- viding a link. Subgenus Alcurus. Blyth, 1843. Type: Trichophorus striatus. Bill and legs dark brown, rather weak; general color olive, striped with white, or plain above, and mottled with white or yel- low below; throat white or yellow. Crested or not. 19. P. striatus: Himalaya, Burma, Yun- nan, N. Siam, N. W. Indo-China. 20. P. leuco grammicus: Sumatra. 21. P. tympanistrigus : Sumatra. These three species are certainly related, the first two rather closely, having both a striated plumage above and a full crest. But the sharp differences in the size and in the shape and length of the elongated feathers of the crown are sufficient to make me consider them as two separate species. P. tympanistrigus, a rare species, is more distantly related but all the same certainly not very far from leucogram- micus, which it much resembles in size and in color, with the exception of the crown and upper back. All three frequent forests and scrub at high altitudes. Subgenus Hemitarsus. Bonaparte, 1850. Type: Turdus ochrocephalus=zeylanicus. Largest sized species of the genus; bill and legs rather short, but strong; crown with short, stiff, decomposed yellow fea- thers; upper parts olive brown and lined with white. Text-fig. 4. Pycnonotus zeylanicus. 22. P. zeylanicus: Malay Peninsula, Suma- tra, Nias I., Borneo, Java. An open country, lowland bird. Its particu- larly large size and the peculiar feathers of its crown do not seem, however, to warrant generic differentiation, the bird being a true Pycnonotus in every other way. In spite of the tremendous difference in size, it resembles closely, P. tympanistrigus in shape and color. Subgenus Loidorusa. Cabanis, 1851. Type: Muscicapa goiavier. Short-winged bulbuls, with rounded or slightly graduated tail, weak or moderate bill and feet, black or brown. General color olive gray or brown, many species with a green wash on the upper parts and the edge of wing and tail feathers; bend of wing yellow or whitish. Less differentiated forms almost plain brown, more differen- tiated ones with white or yellow supercilium or other bright marks on the face and yel- 1943] Delacour: Revision of the Pycnonotidae 23 low under tail-coverts. Feathers of the crown sometimes elongated, but no pointed crest. 23. P. bimaculatus : Sumatra, Java, Bali. 24. P. firdaysoni : Burma, Yunnan, Siam, Indo-China, Malay Peninsula. 25. P. xantholoemus : S. India. 26. P. penicillatus : Ceylon. 27. P. flavescens : Assam, Burma, Siam, S. Yunnan, Indo-China, Borneo. 28. P. goiavier: S. Siam, S. Indo-China, Malay Peninsula, Sumatra and neigh- boring islands, Java, Bali, Borneo, Philippine Is. 29. P. luteolus : Peninsular India and Cey- lon. 30. P. plumosus : Malay Peninsula, Sumatra and neighboring islands, Java, Borneo, Palawan, Calamianes, Cagayan, Sulu. 31. P. blanfordi: Burma, S. Indo-China, Siam, N. Malay Peninsula. 32. P. simplex'. Malay Peninsula, Sumatra and neighboring islands, Borneo, Java. 33. P. brunneus: Malay Peninsula, Suma- tra and neighboring islands, Borneo. 34. P. erythrophthalmos : Malay Penin- sula, Sumatra and neighboring is- lands, Borneo. The members of this large group of bulbuls are certainly much more closely related among themselves than any others, and one leads fairly well to the other, as arranged above. They are all birds of forest undergrowth and scrub, some, particularly goiavier, becoming familiar garden birds. Leucops, from Borneo, is evidently a subspecies of flavescens from which it hardly differs; the slightly graduated tail of the latter does not entitle it to generic distinction, as often accepted (Xanthixus) . In fact, it is the northern or mountain repre- sentative of goiavier, as luteolus is its western counterpart. The separation of penicillata in the genus Kelaartia also seems unnecessary, although it is a very well-marked species; but its face ornaments are only an over development of those found in neighboring forms. P. xan- tholoemus is a rare and peculiar bird, but seems to be an unstreaked, grayer and paler repre- sentative in the west of P. finlaysoni. Subgenus Andropadus. Swainson, 1831. Type: Turdus importunus. Bill small or moderate, black or horn brown, with or without small serrations be- fore the terminal notch on the upper man- dible. Legs and feet moderate in size and variable in color like the bill. Wings and tail moderately long; general color uniform olive brown or green, lighter below, some- times gray or yellow with a black or gray head; one with yellow malar patches. Not crested. 35. P. masukuensis : East Africa (high- lands) . 36. P. virens: West Africa to Kenya and Nyasaland. 37. P. gracilis : West Africa to Uganda and Kenya. 38. P. curvirostris : Liberia to Uganda, Congo and Angola; Fernando Po. 39. P. importunus : South Africa. 40. P. insularis : East Africa (coast) . 41. P. latirostris : West Africa to Uganda and Kenya. 42. P. gracilirostris : West Africa to An- gola, Uganda, Kenya. 43. P. nigriceps : Kenya, Tanganyika and Nyasaland. 44. P. montanus: Cameroon. 45. P. tephroloemus: Cameroon, Uganda and Tanganyika. 46. P. milajansis : Kenya, Tanganyika, Nyasaland, E. Rhodesia. 47. P. chlorigula : Tanganyika. Several generic names proposed by Oberhol- ser cannot be recognized on account of the con- fusion or lack of importance of the characters invoked, and the following are synonyms of Andropadus : Eurillas, Stelgedillas, Stelgido- cichla, Charatillas, Arizelocichla. Bulbuls of this large African subgenus are shy forest birds of somber plumage and several bear close resemblance to the Indo-Malayan Loidorusa. P. tephroloemus and P. milajansis, for example, show the same pale shaft stripes on the ear- coverts as P. plumosus. Bill and legs vary much in length and strength throughout the sub- genus, as in the previous ones. These variations are not linked with the presence or absence of serrations or roughening on the upper man- dible, nor to the color or pattern of the plum- age. III. Genus Calyptocichla. Oberholser, 1905. Griniger serinus. The African species C. serina is undoubt- edly closely related to the subgenus Andro- padus, but it differs considerably from all Pycnonotus by its almost straight and slen- der bill, which is narrow, higher than broad and flesh colored; operculum com- pletely feathered. Plumage light olive above, bright yellow below. Tail rather short and wings long. A forest species. Text-fig. 5. Calyptocichla serina. l.C. serina : West Africa (Sierra Leone to the Congo) ; Fernando Po. 24 Zoologica: New York Zoological Society [XXVIII: 4 IV. Genus Boeopogon. Heine, 1860. Type Criniger indicator. Not far from Andropadus, but has the bill short and broad, wider than high at the base, nostrils operculate and half covered by front feathers. Bill and legs black or gray. Rictal bristles weak. Tail short. (—% wing), the two median pairs of rectrices greenish black, the others whitish-yellow. Upper parts olive green; under parts paler; throat gray. Forest bird of the tree tops. 1.5. indicator : West Africa to Uganda and Angola. 1 think Stresemann is right in considering 5. damans as a color phase of B. indicator, and not as a separate species. V. Genus lxonotus. Verreaux, 1851. Type: lxonotus guttatus. Resembles Boeopogon in having the outer pairs of rectrices yellowish-white. Bill long- er and narrower. Very peculiar plumage pattern, olive gray above with large white spots on the wing and lower back, yellow- ish-white below. A bird of the tree-tops, more active than Boeopogon. 1. I. notatus : Liberia to Angola, the Con- go and Uganda. B. Chlorocichla GROUP. VI. Genus Chlorocichla. Sharpe, 1881. Type: Trichophorus flaviventris. Bill black, moderately long and thick, slightly compressed; nostrils oblong with operculum slightly feathered; culmen and gonys gently curved ; culmen notched at the tip. Nuchal hair short; rictal bristles mod- erately developed. Tail equal to wing, round- ed or slightly graduated. Feet strong, legs rather long, gray. General color brown to olive yellow above, paler below, the throat forming a white or yellow patch, more or less clearly defined. These bulbuls live in forest thickets. Text-fig. 6. Chlorocichla falkensteini falken- steini. 1. C. falkensteini : Cameroon and Angola. 2. C. simplex : West Africa, from Portu- guese Guinea to N. Angola, east to Ituri and Semliki. 3. (7. flavicollis: West Africa to Bahr-el- Ghazal, Uganda, and Lake Tanganyika. 4. C. flaviventris: East and South Africa. 5. C. loetissima: Belgian Congo, Uganda and Kenya. The five species of Chlorocichla have been scattered by different authors over many gen- era. Others have associated them with Pyrr- hurus scandens, which I consider a Phyllastre- phus, having a straight compressed bill. These five species are much alike in shape, propor- tions and color pattern, but as a group they are altogether allied to Pycnonotus, Crmiger and Phyllastrephus, between which they con- stitute a link. VII. Genus Thescelocichla. Oberholser, 1905. Type: Phyllastrephus leucopleurus. Bill like Chlorocichla; feet and legs strong. Tail long and graduated, fan-shaped, the four lateral pairs of rectrices with large white tips. Color pattern peculiar, olive brown above, the face and breast gray streaked with white, the remaining under parts white tinged with yellow. Habits also peculiar; a bird of the raphia swamps. Text-fig. 7. Thescelocichla leucopleura. 1. T. leucopleura : West Africa. C. Phyllastrephus GROUP. VIII. Genus Phyllastrephus. Swainson, 1831. Type: P. terrestris. Bill long, slender, compressed and cari- nated ; culmen almost straight, curved only near the tip, with a terminal notch, often strong; gonys slightly convex. Nostrils slit- shaped set in a depression; rictal bristles variable. Legs longer than bill ; feet strong. General color pale brown, olive green or yellow, lighter and sometimes brighter be- low; a few have the head gray, the throat white; wings and tail often reddish-brown. Tail rounded or graduated. Sexes different in size ; in some cases the male is strikingly larger than the female and has a much longer bill. The bulbuls of this genus are forest birds, some living near the ground, in thickets and the lower vegetation; they are mainly insectivorous, and the nearest of the family to Timaliine and Sylviine birds. l.P. scandens : West Africa, east to Bahr-el-Ghazal. 1943] Delacour: Revision of the Pycnonotidae 25 X- : Text-fig. 8. Phyllastrephus terrestris terrestris. 2. P. terrestris2 3: East and South Africa. 3. P. cerviniventris : Kilimanjaro to Nyasaland, Katanga and N. Rhodesia. 4. P. poensis : Fernando Po and Mt. Cam- eroon. 5. P. hypochloris : W. Uganda and E. Congo. 6. P. fulviventris : Angola. 7. P. olivaceo-griseus : Kivu (Congo), Ankole (Uganda). 8. P. flavostriatus : S. E. Africa. 9. P. rabai : 3 Rabai and Shimba Hills near Mombasa, E. Africa. 10. P. lorenzi: Semliki and Ituri districts (B. Congo). Not examined. 11. P. albigularis : West Africa to Uganda. 12. P. fischeri : Kenya and Tanganyika. 13. P. baumanni: Togo and Ivory Coast. 14. P. polio cephalus : Mt. Cameroon. 15. P. orostruthus: Portuguese E. Africa. Not examined. 16. P. cabanisi: 4 Uganda, E. Belgian Con- go to N. Rhodesia. 17. P. icterinus : West Africa to Uganda. 18. P. xavieri: Angola, S. Cameroon, and French Congo. 19. P. madagascariensis : Madagascar gen- erally. 20. P. zosterops: N. and E. Madagascar. 21. P. tenebrosa: C. E. Madagascar. 22. P. xanthophrys: E. and S. E. Mada- gascar. 23. P. cinereiceps : E. and S. E. Madagas- car. I have arranged in a new way the sequence of species, but for specific grouping and names I have used those adopted by J. Chapin in yet unpublished notes. The first species are those with a thicker bill and browner coloration, the last those with a narrower and more com- pressed bill and yellower plumage. 2 P. strepitans I provisionally consider a subspecies of terrestris, although Friedmann & Loveridge report both forms from Bagamoyo and Dar-es-Salam {Bull. C.M.Z., LXXX, p. 231). 3 debilis is probably a synonym. 4 P. sucosus is probably a subspecies. IX. Genus Bleda. Bonaparte, 1857. Type : Dasycephala syndactyla. Bill narrow and high, compressed ; culmen straight and strongly hooked at the tip; gonys convex; nostrils oval; strong rictal bristles. Legs long and strong; middle and outer toes fused as far as the first joint. Olive green above, bright yellow below, wings and tail reddish or green; head gray in one species. Female much smaller than male, which has a larger bill. These birds live in the undergrowth of forests and are insectivorous. Text-fig. 9. Bleda syndactyla syndactyla. 1. B. syndactyla-. West Africa, east to Uganda. 2. B. eximia : West Africa, east to Uganda. 3 . B. canicapilla: N. West Africa. X. Genus Nicator. Hartlaub & Finsch, 1870. Type: Lanius chloris. Bill strong, compressed and hooked; thicker near the tip than in Bleda; gonys convex; legs and feet strong; general color olive yellow and gray, with yellow spots on the wings; tail margined with yellow below. Soft, fluffy abundant plumage of the bulbuls, not the hard, close one of the shrikes. Female much smaller than male as in Bleda and Phyllastrephus. Chattering voice of bulbuls and brilliant song. Birds of bush and forest; insectivorous. 1. N. chloris: Tropical Africa. 2. N. vireo : Cameroon to N. Angola and Belgian Congo. D. Criniger GROUP. XI. Genus Criniger. Temminck, 1920. Type: Criniger barbatus. Bill strong and high, much thicker at the base than near the tip, which is hooked; culmen curved all over; gonys almost straight. Nostrils oval. Rictal bristles well developed. Tarsus short and strong. Wings 26 Zoologica: New York Zoological Society [XXVIII: 4 and tail rounded. Plumage very long and soft; general color olive yellow or brown, with a conspicuous throat patch of long feathers, white or yellow. Long hair-like feathers on the nape. All species resemble one another closely in color and pattern. Many are fully crested, while others only show indication of a crest, and two none at all. All Criniger are forest birds. Their voice is harsh. l.C. barbatus : Upper Guinea. Text-fig. 10. Criniger chloronotus chloronotus. 2. C. chloronotus : Lower Guinea and Congo. 3. C. calurus : Upper and Lower Guinea to Uganda. 4. C. olivaceus : Senegal to Gold Coast. 5. C. flaveolus: Himalaya, Assam, Burma, Java, Bali. 6. C. tephrogenys : Burma, Siam, Yunnan, Indo-China, Hainan, Malay Peninsula, Sumatra, Borneo, Palawan, Balabac, Camianes. 7. C. ochraceus : S. Burma, S. Siam, S. Indo-China, Malay Peninsula, Sumatra, Borneo. 8. C. finschi : Malay Peninsula, Sumatra, Borneo. 9. C. phaeocephalus: Malay Peninsula, Sumatra and neighboring islands, Bor- neo. The grouping of subspecies in the three close- ly allied and puzzling species, C. flaveolus, C. tephrogenys and C. ochraceus, is difficult and we still have to learn a great deal about them before we can form a final opinion. I adopt here a new grouping proposed by E. Mayr, which I think is the best at the present time; it is fol- lowing almost exactly one suggested in part by C. B. Ticehurst ( Journ . Bombay N.H. Soc., XXXVI, 1933, p. 923-925). 1. Criniger flaveolus. flaveolus : Himalaya, Assam, N. Bur- ma. burmanicus: C. Burma. xanthizurus : Java, except East. balicus : Bali and E. Java. 2. Criniger tephrogenys. griseiceps: C. Burma. henrici : N. Indo-China, Yunnan. pallidus : Hainan. annamensis : C. and S. E. Indo-China. robinsoni: Tenasserim (S. Burma). tephrogenys : Malay Peninsula, Su- matra (lowlands). gutturalis: Borneo (lowlands). / rater: Palawan, Balaban, Camianes. 3. Criniger ochraceus. ochraceus-. S. Burma, S. Siam, Cam- bodia, Cochin-China, N. Malay States. cambodianus : S. Cambodia (moun- tains) . sacculatus : Malay Peninsula. sumatranus-. Sumatra (mountains). ruficrissus: Borneo (mountains). XII. Genus Setornis. Lesson, 1839. Type: Setornis criniger. Bill strong and much hooked, the culmen almost straight and gonys convex. Nostrils and rictal bristles as in Criniger. Very long hair-like feathers on nape. General color brown above, yellowish-white below. Rec- trices tipped with white on the inner web except central pair. No crest. Legs short, feet small. In its strongly hooked bill, thick and al- most straight, Setornis recalls Nicator, but it is more depressed and the nostrils, rictal bristles, hair on nape and plumage char- acteristics are completely different and show its really close relationship to Criniger. It is a very peculiar bulbul. Text-fig. 11. Setornis criniger. l.S. criniger: Banka I., Borneo. XIII. Genus Microscelis. Gray, 1840. Type: Hypsipetes amaurotis. Bill slender and long; culmen carinated and slightly depressed at the base, gently curved ; nostrils oval ; gonys nearly straight ; rictal bristles weak or moderate. Nuchal hair short. Legs very shoi’t. Wings mod- erate to long and pointed. Tail long, rounded, square or slightly forked, the rectrices more or less curved outwards in many species. Feathers of the crown erectile, either round- ed or pointed and lengthened. Color very variable. Most of these bulbuls live in for- est, on high trees in general, some in large flocks; others in scrub and small trees. Their voice is loud and harsh. Frugivorous and insectivorous. 1943] Delacour: Revision of the Pycnonotidae 27 Subgenus T richolestes. Salvadori, 1874. Type: Brachypodius criniger. Culmen straight in its basal half, then curved, and hooked, nostrils oval and ex- posed. Long rictal bristles and very long hairs on the upper back, a unique feature in the family. Wings and tail equal and rounded. General color olive green and brown. Live in brush, bushes and small trees. Text-fig. 12. Microscelis criniger criniger. 1. M. criniger: Malay Peninsula, Sumatra, Borneo and neighboring islands. Subgenus Iole. Blyth, 1844. Type : Iole olivacea = clmrlottae. Bill strongly carinated, pale brown or gray, like the feet. Feathers on crown nor- mal in shape and only slightly elongated. General color uniform olive green, brown or yellow, paler and brighter below. 2. M. charlottae :3 From Cachar and Sylhet to Burma, Yunnan and Indo-China, Siam, Malay Peninsula, Sumatra, Bor- neo and neighboring islands, Palawan. 3. M. nicobariensis: Nicobar Islands. 4. M. ictericus : S. W. India and Ceylon. 5. M. affinis: Moluccas, Togian, Sangi, Pel- ling, Banggai, and Sula Islands. M. affinis varies tremendously in size and the smaller forms are strangely similar to M. icter- icus, while the larger ones, with a particolored tail, dark olive and bright yellow, are not far from M. everetti. It is reasonable to assume that they have a Philippinian origin. On ac- count of their distribution, it seems to be diffi- cult to divide these birds into two separate spe- cies as one would be tempted to do according to their size and tail pattern. Subgenus Microscelis. Gray, 1840. Type: Hypsipetes amaurotis. Bill moderately carinated, of various col- ors like the feet. Feathers on crown pointed 5 Replaces Iole olivacea and Iole virescens. See H. G. Deignan, Auk, p. 313. I. striaticeps is a synonym of M. c. palawanensis. Type, in the Rothschild Collection, examined. and more or less elongated. General color very variable, with distinct patterns, never uniform olive green, brown or yellow. 6. M. everetti Philippine Islands (not Luzon) . l.M. gularis :6 7 Philippine Islands. 8. M. siquijorensis : Philippine I.: Siqui- jor, Tobias, Romblon, Cebu. 9. M. amaurotis : Japan and neighboring islands, Corea and E. China (mi- grant), Formosa, Riu-Kiu and N, Philippine Islands. 10. M. virescens:8 Himalaya, Burma, S. China, Indo-China, Siam, Malay Pen- insula, Sumatra, Borneo, Java. 11. M. flavalus :8 Himalaya, Burma, Siam, Indo-China, S. China, Hainan, Malay Peninsula, Sumatra, Borneo. Text-fig. 13. Microscelis madagascariensis. psaroides. 12. M. madagascariensis :9 Madagascar, Mauritius, Aldabra, Reunion, Sey- chelles, Comoro Islands, Ceylon, India, Assam, Burma, Siam, Indo-China, C. and S. China, Hainan, Formosa. 13. M. thompsoni :10 S. Shan States, N. W. Siam, E. C. Burma. 6 Although both forms are found on Mindanao, I con- sider provisionally M. ruflgularis as a subspecies of everetti and its lowland and open forest representative. M. haynaldi, from Sulu and other southern islands, is exactly intermediate. 7 T urdus philippensis Gmelin is. preoccupied by Muller, 1776, and Boddaert, 1783. The next available name is Philemon gularis Pucheran in “Cuvier” Arch. Mus. N. H. (Paris), 1855, 7, 344, pi. 18. 8 I agree with Deignan that the different forms ascribed by most authors to malaccensis, tickelli , or mac - clellandi are conspecific with virescens, and that castano - tus, canipennis and cinereus are subpecies of flavalus. See Auk, 1942, pp. 313, 314. 9 I follow Mayr, Deignan, Danis and others in uniting specifically all the closely allied birds formerly referred to madagascariensis, psaroides and leucocephalus. 10 I do not consider that a bare skin space round the eye and a chestnut patch on the vent are worth a generic separation under the name Cerasophila, as thompsoni is evidently very close to M. m. leucocephalus. Zoologica : New York Zoological Society EXPLANATION OF THE PLATES PLATE I. Fig. 1. Spizixos semitorques semitorques. Fig. 2. Pycnonotus dispar dispar. Fig. 3. Pycnonotus barbatus xanthopygus. PLATE II. Fig. 4. Pycnonotus sinensis sinensis. Fig. 5. Microscelis amaurotis amaurotis. Fig. 6. Microscelis nvadagascariensis psaroides. DELACOUR. PLATE I. FIG. I. FIG. 2. FIG. 3. A REVISION OF THE GENERA AND SPECIES OF THE FAMILY PYCNONOTIDAE (BULBULS). DELACOUR. PLATE II A REVISION OF THE GENERA AND SPECIES OF THE FAMILY PYCNONOTIDAE (BULBULS). Delncour: Two New Subspecies of Pycnonotus cafer 29 5. Two New Subspecies of Pycnonotus cafer. Jean Delacour. (Text-figure 1). The crested and white-rumped bulbuls, common in cultivated areas and open jungles of India, Burma, Siam, Indo-China and Java, are evidently all eonspecific ( Pycnono- tus cafer), but it has long been difficult to trace the transition from the widespread red-vented forms to the yellow-vented races isolated in southern Indo-China and Java: germaini and aurigaster. An interesting light was thrown on the subject when C. B. Kloss (Journ. N. H. Sty. of Siam, 6. No. 3, 1924, p. 291) described a yellow-vented bird from Bangkok as Molpastes aurigaster thais. This bird resembles very closely the Javan aurigaster in general color, even in the rich golden shade of the under tail- coverts. He, however, was wrong in con- sidering a Chantabun bird as similar, as it is a little browner and less distinctly mottled above and has its under tail-coverts of a decidedly different tone of yellow, ap- proaching the lemon yellow of germaini. Text-fig. 1. Distribution of races of Pycnono- tus cafer in southern Siam. 1. Um-Phang ( P.c . klossi ± schauenseei) . 2, Si-Sawat (P.c. scha- uenseei). 3, Ban-Pong, 4, Bangkok (P.c. thais). 5, Chantabun (P.c. deignani) . The nearest relative of these yellow-vent- ed bulbuls certainly is the small, crimson- vented form of central Siam and Tenas- serim, klossi, differing only very slightly from chrysorrhoides, from the Shan States, Indo-China, N. Siam and S. China, in its smaller size. As one goes south, however, these bulbuls show a tendency to vermillion instead of crimson under tail-coverts. A good series collected by W. P. Lowe (Ver- nay Expedition) 53 miles east of Um-Phang, in W. C. Siam, shows either crimson, ver- million or orange vents, indicating a mixed intermediate population. At Si-Sawat, some distance to the south, birds have orange vents, while at Ban-Pong, still further south, they have golden-yellow vents, are similar to the Bangkok birds and must be called thais. The orange-vented birds have no name, nor have the Chantabun birds with black heads like thais, but lemon yellow7 vents almost like germaini. The latter is the dullest race, with a dark brown crown; it is found in all suitable parts of Indo-China, at low and moderate altitudes, south of Tourane in the east, of Nape and Vien- tiane in the west, and in the neighboring parts of eastern central Siam. I therefore propose for them the follow- ing appellations: Pycnonotus cafer schauenseei. Intermediate between P.c. klossi and P.c. thais, in having the under tail-coverts mixed vermillion and yellow, producing an orange color. Type $, No. 1369. Academy of Natural Sciences, Philadelphia. Coll. R. M. de Schau- ensee, July 6, 1939. Si-Sawat, W. C. Siam. Wing: 88; tail: 79; tarsus: 20; culmen : 18 mm. Birds from Um-Phang are intermediate and variable, some examples being similar to the Si-Sawat birds. Twelve specimens ex- amined from Si-Sawat and Um-Phang. Named in honor of Rodolphe Meyer de Schauensee. Pycnonotus cafer deignani. Intermediate between P.c. thais and P.c. germaini, in having the black crown and general color of the former, only slightly browner and less distinctly mottled, and 30 [XXVIII :5:1943] Zoologica: New York Zoological Society 7 lemon-yellow under tail-coverts, only slight- ly more golden than in germaini. Pale tail- tips intermediate in size and color between thais and germaini. Type 8, No. 337.130, U. S. Nat. Museum, Washington. Coll. H. G. Deignan, April 20, 1932. Chantabun, S. E. Siam. Wing: 90; tail: 80; tarsus: 21; culmen: 18 mm. Six specimens examined from Chantabun and Chantaburi. Named in honor of H. G. Deignan. Crane: Uca schmitti Sp. Nov. 31 6. : ;/ Eastern Pacific Expeditions of the New York Zoological Society. XXXI. Uca schmitti, a New Species of Brachyuran Crab from the West Coast of Central America.1 2 Jocelyn Crane Research Zoologist, Department of Tropical Research, New York Zoological Society. (Plate I; Text-figure 1). [This is the thirty-first of a series of papers dealing with the collections of the Eastern Pacific Expeditions of the New York Zoologi- cal Society made under the direction of Dr. William Beebe. The present paper is concerned with specimens taken on the Eastern Pacific Zaca Expedition, and, through the kindness of Dr. Waldo L. Schmitt and Dr. Thomas Bar- bour, with material in the collections of the United States National Museum in Washington and of the Museum of Comparative Zoology at Harvard in Cambridge, Massachusetts.] Except for nine specimens from the east- ern Pacific coast, referred to Uca mordax by Dr. Rathbun (1917) and myself (1941), the known range of the species is restricted to the western Atlantic. A reexamination of each of these nine examples and their com- parison with many Atlantic specimens of mordax have convinced me that they should be referred to a new, homologous species, described below. With the taxonomic sepa- ration of these forms, no species of Uca remains which, in the consideration of modern taxonomists, occurs on both the At- lantic and Pacific coasts of the hemisphere. Uca schmitti sp. nov." Text-fig. 1; PI. I. References: Uca mordax Rathbun, 1917, p. 393 (part.). Crane, 1941, p. 176; text-figs. 2, 3, 4E, 5. Description: This proposed new species, although completely distinct from its Atlan- tic homologue, U. mordax, differs from it noticeably and invariably only in the follow- ing characters: 1. The oblique ridge inside the palm is much lower than in mordax, and the tuber- 1 Contribution No. 652, Department of Tropical Re- search, New York Zoological Society. 2 Published by permission of the Secretary of the Smithsonian Institution. cles with which it is irregularly covered are smaller. Similarly, the tubercles between this ridge and the one at the base of the dactyl are fewer, cover a smaller area and are themselves smaller. These characters alone enable one to separate a mixed group of moderate-sized specimens without the aid of a lens. 2. The palm is relatively deeper than in mordax, and the fingers shorter. 3. In contrast to mordax, on each of the first three ambulatories there is little pile on the upper sides of the merus and carpus, and none on the lower sides of the manus. 4. Hair on tips of minor chelae usually almost lacking in adults, and always far sparser than in mordax. 5. Marginal line of front almost straight, not convex; front appearing slightly broader and shallower than in mordax, although this character is variable and cannot be definite- ly measured. 6. Abdominal appendage thicker, with subterminal arm closely applied to tip of appendage, instead of projecting from it, as it does in mordax. 7. All four males observed alive were definitely spotted (Crane, 1941, p. 177) ; none of the many examples of U. mordax which I observed in Venezuela ever showed a trace of spots (see paper immediately following, Zoologica, Vol. XXVIII, p. 37). Measurements: Male holotype (U.S. Na- tional Museum No. 80451), length 13.4 mm., breadth 20 mm.; base of manus to tip of pollex 35 mm. Three male paratypes (U.S. N.M. No. 22306), lengths 13, 14 and 14.5 mm. Four male paratypes (Dept. Tropical Research, N.Y. Zoological Soc. Nos. 381,116, 381,117, 381,118), lengths 9.7, 13, 14 and 14.5 mm. One male (Museum Comparative Zoology No. 5892), length 15.3 mm. 32 Zoologica: New York Zoological Society [XXVIII :6: 1943] L' F Range: Acapulco, Mexico, to Golfito, Costa Rica Material: Holotype: U.S.N.M. No. 80451, San Bias, Tepic Territory, Mex. ; 3 para- types: U.S.N.M. No. 22306, same locality; 4 paratypes : Dept. Tropical Research, N. Y. Zool. Soc. : No. 381,116, San Juan del Sur, Nicaragua; No. 381,117, Negritos Island, Costa Rica; No. 381,118, Golfito, Costa Rica; 1 male: Museum Comparative Zoology, No. 5892, Acapulco, Mexico. This species is named in honor of Dr. Waldo L. Schmitt, Curator of the Division of Marine Invertebrates at the United States National Museum. Bibliography. Crane, J. 1941. Eastern Pacific Expeditions of the New York Zoological Society. XXVI. Crabs of the Genus Uca from the West Coast of Central America. Zoo- logica, Vol. 26, No. 19, pp. 145-208. Rathbun, M. J. 1917. The Grapsoid Crabs of America. Spec. Bull. U. S. Nat. Mus. No. 97, pp. xxii & xxxx, 1-461. EXPLANATION OF THE PLATE. Plate I. Fig. 1. Uca schmitti. Male holotype (U.S.N.M. No. 80451), dorsal view. Carapace length 13.4 mm. Fig. 2. Uca schmitti. Major chela of paratype (U.S.N.M. No. 22306), inner view. Carapace length 14 mm. X 2.4. Fig. 3. Uca mordax. Major chela of specimen from Caripito, Venezuela (Dept. Trop. Research, N.Y.Z.S. No. 4252a), inner view. Carapace length 14 mm. X 2.4. Fig. 4. Uca mordax. Major chela from Cari- pito, Venezuela (Dept. Trop. Research, N.Y.Z.S. No. 4252b), inner view, X 2.4. Text-fig. 1. Right abdominal appendage of adult male in Uca mordax and U. schmitti. A, mordax, outer lateral view; B, same, tip, outer lateral view; C, same, tip, anterior view; D, schmitti, outer lateral view; E, same, tip, outer lateral view; F, same, tip, anterior view, p: genital opening; a: subterminal arm. (Exact numbers of hairs not shown in full-length views, although apparent proportions and ex- act locations are indicated; hairs omitted from drawings of tips). CRANE. PLATE I FROM THE WEST COAST OF CENTRAL AMERICA. Crane: Uca from Venezuela 33 7. Crabs of the Genus Uca from Venezuela.1 Jocelyn Crane Research Zoologist, Department of Tropical Research, New York Zoological Society. (Plate I; Text-figure 1). [This is a contribution from the Forty- third or Venezuelan Expedition of the De- partment of Tropical Research of the New York Zoological Society made under the direction of Dr. William Beebe. The expedi- tion was sponsored by grants from the Committee for Inter-American Artistic and Intellectual Relations and from four trustees of the Zoological Society, George C. Clark, Childs Frick, Laurance S. Rockefeller and Herbert L. Satterlee, and by invaluable as- sistance from the Standard Oil Companies of New Jersey and Venezuela.] Contents. I. Introduction 33 II. Ecology 33 III. Phylogenetic Notes 35 IV. Species of Uca Taken by the Venezuelan Expedition of the New York Zoological Society 35 U. maracoani (Latreille) 35 U. mordax (Smith) 37 U. murifecenta sp. nov 38 U. pugnax rapax (Smith) 40 U. cumulanta sp. nov 42 V. Bibliography 44 I. Introduction. The present study is based on 1,049 speci- mens of Uca collected on the Venezuelan Expedition of the Department of Tropical Research of the New York Zoological So- ciety, under the direction of Dr. William Beebe. The expedition extended from Febru- ary to September, 1942, with headquarters at Caripito, in northeastern Venezuela. The specimens are distributed among five species, of which two are apparently new to science, and were taken in four general localities: the Caripito region of the San 1 Contribution No. 653, Department of Tropical Re- search, New York Zoological Society. Juan River; the swamp near the mouth of the San Juan; the Pedernales region at the mouth of the Cano Manamo at the southern end of the Gulf of Paria; and the Maracaibo region, in western Venezuela. For the excel- lent and extensive collection made at Mara- caibo, I wish to express my thanks to Mr. Henry Fleming, the expedition’s entomolo- gist, who collected the material for me during a week’s stopover in that locality. During the course of collecting the re- mainder of the crabs, I was able to make a number of ecological observations, which are incorporated in the present report, in- cluding the displays of four of the five species taken. For general working methods, and for definitions of measurements and special terms, refer to Crane, 1941, p. 148. The making of a series of 16 mm. color motion pictures of displaying fiddlers at Pedernales was the only new addition to the working procedure. Detailed drawings of abdominal appendages are presented, since the genital organs are proving, in this as in other groups of animals, to be of definite aid in taxonomy ; more work needs to be done, how- ever, before basic structural plans and their evolutionary significance can be adequately described and evaluated. Miss Rathbun’s synonymy (1917) has been accepted throughout, except in regard to U. mordax (see p. 38). II. Ecology. Habitat: All of the crabs were taken from slightly brackish river banks and delta shores. Although each locality was under tidal influence, the salinity in all cases was very low. The majority were collected dur- ing the rainy season in the summer months, so that the salinity was at its minimum. Adequate series of salinometer readings were not made, but Table I will give a gen- eral idea of conditions. 34 Zoologica: New York Zoological Society [XXVIII: 7 Table I. Salinity (cf. with normal Place ’ Date Season Tide sea water) Caripito Apr. 10 Dry 2 hrs. after high 0.6% Near mouth San Juan R., Apr. 10 Dry High 1.2% 8 miles above bar Pedernales Aug. 25 Rainy High .23% Pedernales Aug. 26 Rainy 1 hr. after low 1.2%2 Maracaibo, Yacht Club Sept. 7 Rainy High 2.0% In spite of these relatively small differ- ences, the five species taken group them- selves definitely into a series, the proportions shifting from river to near river mouth to Maracaibo mud flats in a way that cannot be wholly accounted for by changes in vegeta- tion, soil or amount of light, since in these respects certain stations in each of various localities were very similar. The occurrences of the crabs may be tabulated as in Table II. The second habitat (near the river mouth) was collected only during a two-hour period on a single day at one spot, in deep shade; Table II. River: Minimum salinity; mangroves rare: Caripito, Guanoco. Near river mouth: mangroves dominant: San Juan R., 8 miles above bar. Lower delta swamp: mangroves and marsh grass only: Pedernales, at mouth of Cano Manamo. Tidal flats: mangroves adjacent or distant; marsh grass sometimes present: Maracaibo. back from the banks, among the highest mangroves ; hence the results obtained there would doubtless be considerably modified by more extended collecting in the vicinity. The other regions, however, were exten- sively collected, and although additional species would almost certainly be added by exhaustive collecting over a protracted period, the relative abundance indicated by the table would doubtless in general hold true. The major fact appearing is that up rivers, toward the limit of the Uca range, only U. mordax occurs, but that it is here extremely successful ; all along the banks of the upper San Juan and its tributaries, and many yards back in the adjacent swampy land, mordax was exceedingly abundant, the colonies often lining the banks in practically unbroken series. At the opposite end of the brackish water habitat, on the tidal fiats - The apparently contradictory results obtained at Pedernales, where the salinity near low tide was almost six times more than at a high tide near full moon, is doubtless due to the erratic currents in the neighborhood. (both in the semi-shade of mangroves and on the open flats) around Maracaibo, mor- dax was completely absent while pugnax rapax was far and away the dominant form. Of 526 specimens of Uca collected by Mr. Fleming at Maracaibo from five different stations, giving as varied habitats as pos- sible, all except 15 were pugnax rapax, the remainder being murifecenta, the proposed new species taken also near the mouth of the San Juan and at Pedernales. At Pedernales, the only locality where all five species were taken, the preferences for e e *K> s e "e s § §| s e s e o Cl e s- o s g s* e § a § s o § Dominant — — — — yg. only (April) Rare Dominant — — Abundant Common Rare Abundant Common Dominant Rare particular habitats could be best compared. Here the differences noted in Table III were obvious, and these observations tallied with preferred habitats in the other localities. Size: No statistical work has been done as yet, but it is interesting to note that adult pugnax rapax taken at Maracaibo from the No. 2 type of habitat — i.e., from close to or among mangroves — were on the average definitely larger than those from the barren mud and sandy-mud flats in the same neigh- borhood. Breeding: Apparently in Venezuela the main breeding season is in the spring. U. mordax at Caripito was displaying vigor- ously from February throughout April, and ovigerous females were seen at this time. Between May and September there was pro- gressively less waving seen, and by late summer it appeared to have practically died out, although during the third week in August, at Pedernales, it was continuing in desultory fashion. The season of pugnax 1943] Crane: Uca from Venezuela 35 Table III. 8 Key ft. 5 e X, Preferred ft <£> X, Present e 8 e Si • 3.00) and that form change was taking place during the post-metamorphic growth. By means of sta- tistical methods Keys was able to show that variations in both k and b denote change of form. Therefore, if we are to study ade- quately the implications of weight in rela- tion to form the equation should be deter- mined empirically. The formula W=aLn2 implies the same conditions that have been expressed for the growth of a part relative to another part or the remainder of the body. Weight ( W ) and length (L) are the variables under consid- eration ; the exponent n indicates the rela- tive increases of these variables during the growth of the fish. The coefficient a is in- dicative of the units of measurement used (Keys, 1928) and also denotes the value of the weight when the length is mathemati- cally equal to 1. It thus serves as an index are reduced to a single theoretical unit. It is true that a will fluctuate seasonally with- 2 These symbols are used for the allometry formula of weight in relation to length as distinguishable from the general formula, y = bxk, to denote relationships of dimen- sions, i.e., width and depth. 1943] Shapiro : Weigh t and Body Form in Scombroid Fishes 91 Table 3. Values of k and b for the Partition of Increments of Body Width Relative to the Standard Body Length. The exponent k„x denotes the value for the regression line of width upon length, kxy is the value for length increase upon width, but still expressed in terms of the standard length, and k is the geometric mean value for kyx and kxy. The species are arranged in descending order, within each family, for the value of b. The range of data for width and length, both in millimeters, is given for each species. Range in Size Species kyx kXy *' k b Depth Length Pomatomidae Pomatomus saltatrix 1.170 1.236 1.20 0.040 16.0— 49 145— 363 Carangidae Caranx crysos 1.014 1.107 1.06 0.109 13.0— 57 96— 203 Caranx hippos 1.091 1.145 1.12 0.086 18.0— 49 116— 312 Selene vomer 1.065 1.079 1.07 0.086 4.1— 15 30— 122 Vomer setapinnis 1.107 1.131 1.12 0.057 5.3— 20 55— 180 Trachinotus carolinus 1.034 1.042 1.04 0.012 3.1— 40 22— 304 Scombridae Scomberomorus maculatus 1.075 1.114 1.10 0.057 16.0— 55 167— 552 Trichiuridae Trichiurus lepturus 1.318 1.344 1.33 0.0025 4.2— 19 263— 810 Istiophoridae Makaira nigricans marlina 1.313 1.533 1.42 0.0040 190.0—425 1950—3290 Makaira nigricans ampla 1.230 1.725 1.46 0.0027 165.0—320 1950—3040 * These values are the reciprocals of They express the increase of x on y the regression values for increase of in terms of the standard length. length (x) upon width (?/). I in the limits of variation and consequently some investigators (mostly those who have worked with the cube law) have used this coefficient as a measure of the condition of the fish or the state of sexual development. We prefer to use it not as a measure of fluc- tuating heaviness but of relative bulk, spe- cific for each type of fish. The empirical determination of the rela- tionship between weight and length shows that each species maintains its specific growth ratio. The range of the exponent n is from 2.80 in Caranx hippos to 3.54 in Marlina nigricans ampla (Table 1, Text-figs. 1-4)* 1 * 3 4. Those species with a value below 3.00 are maintaining a negatively allometric in- crease of weight for each unit increase of the length. Thus, among the carangids, Caranx hippos and C. crysos (Text-fig. 1) and Selene vomer 4 and Vomer setapinnis (Text-fig. 2) become relatively lighter with increase in length. Among the species stud- ied of closely allied families, Scombero- morus maculatus shows a similar condition. The weight increases exactly as the cube of the length only when an isometric rela- tion between these variables is maintained. This is a special case of growth similar to j| 3 The growth constants for Caranx crysos have been de- 1 termined only for those fish up to 203 mm. in standard length. The specimens above this size depart significantly I from the curve in a more positive direction. Whether this deviation is due to a change in the growth rate at this point or to an insufficiency of accurate measurements can- not be determined as yet. 4i Selene vomer is currently recognized as being synony- ! mous with Argyreiosus vomer. the previously-explained isometric condition where linear dimensions were considered, but here growth would proceed either iso- metrically along all three geometric axes, or the allometric growth along different axes would be exactly compensating. The only species approximately showing this special isometric length-weight relation is the blue- fish ( Pomatomus saltatrix) . For this species the value of n is 2.98 but the regression of weight on length shows a value of 2.962 and of length on weight 3.001 (Table 1). If we interpret this almost ideal weight — - cube-length relationship on the basis of no change in form (as the cube law would demand), P. saltatrix should maintain con- stancy of body form during growth. The linear dimensions of depth and width, how- ever, show no approach to the condition of unchanging proportions. The constant dif- ferential partition of growth increments in relation to the length is 0.88 for the depth (Table 2) and 1.20 for the width (Table 3). Since these exponential values deviate about equally from the value (1.00) for isometric growth they are approximately compensat- ing, and the weight increases about as the cubic function of the length. The remaining species studied show a positive allometric increase of weight in terms of length. Trachinotus carolinus and T. falcatus show a slight tendency to be- come heavier with increase in length (Text- fig. 3, Table 1) but Trichiurus lepturus and Total body weight (in grams) 92 Zoologica: New York Zoological Society [XXVIII: 12 Standard length of body (in mm.) Text-fig. 1. Increase of total body weight with standard body length; logarithmic plotting for both weight and length. both subspecies of Makaira maintain a high positive allometry (Text-fig. 4, Table 1). In studies of length-weight relations the measurement of maximum length, depth and width have been uncritically accepted as rep- resenting precisely the form of the fish throughout life. Changes in shape, however, may take place in portions of the body which have not been measured, for instance the head and peduncular regions. In an an- alysis of numerous dimensions of Makaira nigricans ampla (Shapiro, 1938), the gra- dients for different parts of the fish were shown to differ considerably, the tendency being for the exponential values to be more intense posteriorly and less so anteriorly. Only if the shape of all sagittal and all frontal sections remained constant, would the maximum depth and width measure- ments accurately represent the form. In that case the sum of the exponential values for the three axes should equal the value of n in the length-weight formula. Selene vomer and Vomer setapinnis are the only species in this study that maintain such a condition (Table 4), although the possibility still ex- ists that various sections of their body too may be growing at different rates, since positive exponential values in some sections of the body may just compensate negative values in other parts. The results obtained with the remaining species indicate that unmeasured proportions of the body change with age, for the sum of the exponents of the linear dimensions often does not closely approximate the value of n (Table 4). A comparison of the initial growth index (a) with the size attained by the various species indicates that the larger fish tend to show a smaller coefficient of bulkiness (Table 1). At the inception of the post- metamorphic straight line relation Trichi- urus lepturus (a=0. 0000000674) and Mak- aira nigricans ampla (o=0. 000000161) and 1943] Shapiro: Weight and Body Form in Scombroid Fishes 93 Table 4. Sum of Exponents for Linear Dimensions Compared with Exponent for Length-Weight Relationship. If the three geometric axes of length, depth and width adequately represent the form of the body, the sum of their exponential values should equal the value of n for the length-weight relation. If the sum of the k values of these axes does not closely approxi- mate the value of h, changes in shape of other portions of the body are taking place and the maximum measurements of length, depth and width are not truly representative of the form of the body. Value of k for Sum, 1+2+3 Value of n (1) Length (2) Depth (3) Width Pomatomus saltatrix 1.00 0.88 1.20 3.08 2.98 Caranx hippos 1.00 Caranx crysos 1.00 Vomer setapinnis 1.00 Selene vomer 1.00 Trachinotus carolinus 1.00 Scomberomorus maculatus 1.00 Trichiurus lepturus 1.00 Makaira nigricans ampla 1.00 Mukaira nigricans marlina 1.00 M. n. marlina (a=0. 000000118) have de- cidedly smaller bulk per unit length than do the deeper-bodied and shorter forms of the carangids, such as Vomer setapinnis (a = 0. 0000648) and Selene vomer (a=0. 0000483) . In its a value of 0.0000288, Scomberomorus maculatus approaches slight- ly the condition of the specialized fusiform Istiophoridae and the elongate Trichiuridae. The value of a for Trachinotus carolinus (0.0000264) and T. falcatus (0.0000350) is low as compared with the other carangids. It is similar to the value for a central scombrid form like S. maculatus. Thus, for unit length, the longer fish, at the point where differentiation has been completed and post-metamorphic growth is beginning to follow the straight line relation, show a low weight coefficient as compared with the |J high value for the deep-bodied forms. If a is accepted as an indicator of the type of body form, the eleven species of fish can be arranged in a series grading from the deep-bodied to the elongate type. The use of the initial growth index for this pur- pose disregards the changes in body form that take place during post-metamorphic growth. The value of a is high for the deep- bodied compressed forms such as Selene vomer and Vomer setapinnis. Their greater relative bulkiness at the beginning of the post-metamorphic growth period seems to be directly correlated with their extreme depth and shortness of body. The moderately elongated fusiform type, represented by the mackeral, bluefish and pompano ( Trachi- notus carolinus), show central values of a (smaller bulkiness per unit length than in the deep-bodied forms). The most special- ized, elongated types, represented by the cutlass fish ( Trichiurus lepturus ), give the : low values at the other extreme of the series. In this species the theoretical bulk 0.81 1.12 2.93 2.80 0.90 1.06 2.96 2.87 0.75 1.12 2.87 2.84 0.92 1.07 2.99 2.97 1.13 1.04 3.17 3.05 0.87 1.10 2.97 2.84 1.18 1.33 3.51 3.36 1.32 1.46 3.78 3.49 1.24 1.42 3.66 3.54 per unit length at the beginning of the nost- metamorphic period is incredibly small. It should be pointed out that the standard length of the marlins was determined with the enormous bill included. The value of a would be somewhat higher if the sword were discounted and would place them closer to the category of the fusiform mackeral type. That there is a relationship between a and n can be shown by plotting these values for the different species on semi-log paper (Text-fig. 5). The general tendency is such that n increases in arithmetical progression as a decreases in geometrical progression, thus conforming to an exponential function wherein log a shows a definite rate of de- crease in terms of increase of n. Hersh (1931. 1934) found that, in the relative growth of genetic mutants of Drosovhila, and in the evolutionary relative growth of the Titanotheres, this same exponential re- lationship holds true. Lumer (1939) also obtained an inverse relationship between the constants for the radius-humerus and leg-arm relations in the Pongidae. Thus, if the value for a suggests a slim fish per unit length at the termination of its larval growth, its weight increase, relative to the lenvth, for the post-metamorphic period is correspondingly greater (compare values for a and n in Table 1). Makaira nigricans ampla, M. n. marlina and Trichiurus len- turus, with the low a values of a slim fish, show a steeper gradient for the distribution of weight as the fish increases in length, whereas the reverse is true of the deeper- bodied fish. The trend of a in relation to n for the various families has been indicated bv the broken lines on Text-fig. 5. It may signify that the exponential relationship is distinctive for each familv and therefore of some value in determining phylogenetic trends. Total body weight (in grams) 94 Zoological New York Zoological Society [XXVIII: 12 Text-fig. 3. Increase of body weight with in- crease of standard body length; logarithmic plotting both ways. The values of n (Table 1) indicate that closely allied members of the various fam- ilies tend to have similar coefficients of growth partition. The species Caranx hippos and C. crysos have values of 2.80 and 2.87 respectively, Trachinotus carolinus 8.05 and T. falcatus 8.09, and Makaira nigricans ampla 3.49 and M. n. marlina 3.54. Although similar within each genus, the species of Caranx and of Trachinotus show no over- lapping of the regression lines for nyx and nxy (Table 1). Therefore the species can be separated by their constant differential growth ratio, as well as by differences in weight at the same length. T. falcatus is relatively heavier than T. carolinus and C. hippos relatively heavier than C. crysos (re- fer to Text-figs. 1 and 3) . There is a distinct tendency for convergence between the spe- cies of Caranx (eventually the index of weight to length becomes the same for both), but for divergence between the species of Trachinotus. The relative growth rates for the post-metamorphic period ap- pear to be broadly similar for closely re- 95 1943] Shapiro: Weight and Body Form in Scombroid Fishes Value of n for weight in relation to length Text-FIG. 5. Values of a for weight in relation to length plotted logarithmically against the growth coefficients (n) for weight relative to length; n plotted on an arithmetic scale. lated species, but specific differences are maintained. Depth and Width as a Measure of Body Form. For the body measurements (depth and width in relation to length) the relation- ships are linear and have been expressed either by the equation y = a + hx or by the geometric form y — bxk. Which equation is more applicable to the data? Meek (1905) and Hecht (1916) used the first equation, for they maintained that, if the lines for the growth of the dimensions of the body con- verge to zero when plotted in relation to the length, the equation becomes y = bx. This isometry equation is therefore applicable when the growth coefficient is close to 1 and will fit the data as well as the allometry formula. But what of the instances when the data does not show the specific case of isometry? The first equation ignores the difference of k from one, whereas the sec- ond equation ignores the difference of a, from zero. If the growth involved is geo- metric the factor a is small and no valid error results from ignoring it. Simpson and Roe (1939, p. 370) show that the allometry formula fits the facts better than does the isometry formula when growth is markedly allometric, as it often is. The values for the constant differential partition of depth in relation to length show negative allometry for some species, and definitely positive allometry for others. Among the Carangidae the values for the 96 Zoologica : New York Zoological Society [XXVIII: 12 Standard length of body (in mm.) Text-fig. 6. Relative increase of depth in terms of the standard length for the Pomatomidae and the Carangidae. Scale both ways in logarithmic increase. partition of the growth increments range from 0.75 for Vomer setapinnis to 1.25 for Trachinotus falcatus (Text-fig. 6). Thus in V . setapinnis the body is becoming slenderer to the extent that each unit increase of depth is 0.75 of each unit increase in length. The reverse is true for the positively allo- metric T. falcatus, which becomes relatively deeper as it grows longer. None of the earangids show an ideal isometric condi- tion. S comb eromorus maculatus and the al- lied Trichiuridae and Istiophoridae show relations similar to those exhibited by the Carangidae (Table 2, Text-fig. 7). S. macu- latus, with a value of 0.87 for k, grows in- creasingly slender, whereas Trichiurus lep- turus (k = 1.18) and Makaira nigricans ampla (k = 1.32) and M.n.marlina (k = 1.24) shows definitely positive allometric increases in depth. The significance of such trends may be correlated with the total size that the fish attain and with the value of the gradient for the deposition of weight. According to Glaser (1938) form is of prime importance and therefore the usefulness of an organ and the adaptiveness of an organism to its environment are to some degree functions of the weight. Changing proportions place upon the organism the necessity of coping with the environment in a constantly dif- fering way. If the environment does not directly affect the direction of growth of the various dimensions, selection in the course of evolutionary change must ad- just these factors to maintain an efficient mechanism for the locomotion of the fish as its body weight increases. With an increase in the relative bulk of the fish (i.e., when m>3.00) there must be an increase in one or more of the transverse body dimensions. Of the species studied, all those which be- come relatively heavier with increased length, namely Trachinotus carolinus, T. falcatus, Trichiurus lepturus, Makaira ni- gricans ampla and M. n. marlina, also be- come relatively deeper (compare values of n in Table 1 with values of k in Table 2). All species studied show positive allometry for the relationship of width to length ( Table 1943] Shapiro: Weight and ''Body Form in Scombroid Fishes 97 Standard length of body (in mm.) Text-fig. 7. Body depth relative to body length for the Scombridae and the allied Trichiuridae and Istiophoridae. Logarithmic plotting both ways. 3). They grow in such a manner that the width increases at a proportionally greater rate than the length (Text-figs. 8-11). Thus fishes {S comb eromor us, Selene, Vomer, or Caranx ) which become relatively lighter (»<3.00) must do so through a relative de- crease in depth. In all the genera, with the exception of Trachinotus5 , the differential growth in width proceeds more rapidly than the growth in depth (compare values of Jc in Tables 2 and 3), even in Trichiurus and Makaira, which show a positive allometric increase in each dimension. The change to a heavier fish proceeds more rapidly along the transverse than along the vertical axis (except in Tra- chinotus). In Trachinotus carolinus both depth and width become proportionally greater than length, but the depth maintains a steeper gradient of growth than does the width. The species studied also show significant 5The value for k for the relation of width to length in T. falcatus could not be determined because of the large error involved in measuring the small fish which alone were obtainable. resemblances and differences in the value of the coefficient b (Tables 2 and 3). In general this value is lower in the more specialized forms ( Makaira nigricans ampla, M. n. mar- lina, Trichiurus lepturus and Trachinotus carolinus ) than in the more deep-bodied, compressed species. For example, in T. lep- turus b = 0.022 for depth and 0.0025 for width, whereas in Vomer setapinnis b — 1.97 for depth and 0.057 for width. In general there is a negative correlation between the values of the coefficients a and 6 on the one hand and the values of the ex- ponents n and k on the other hand (Tables 1-3 ). This correlation holds particularly well when the comparison is made between spe- cies of the Carangidae and between those referred to the Scombridae, Trichiuridae and Istiophoridae. In biological terms, the species which are the deeper and the wider initially become relatively slenderer and thinner with age, and vice versa. In the spe- cies studied, therefore, the body form is the more diverse in the young stages, and what might be called an interspecific regulation Greatest width of body (in mm.) 98 Zoologica: New York Zoological Society [XXVIII: 12 Standard length of body (in mm.) Text-fig. 8. Body width relative to standard body length plotted logarithmically both ways. in growth tends to mold the adults into a more uniform and conventional fish-form. Relative Growth in Relation to Phylogeny and Systematics. The growth intensity of the body as a whole is being distributed according to an orderly system of growth gradients (Thomp- son, 1917; Huxley, 1932). Each species of scombriform fish here analyzed has been shown to possess an individual and specific pattern of growth, and the relative changes in depth and width have indicated certain similarities for closely related species. It is our purpose to evaluate these specific growth gradients and determine whether or not there is an interspecific trend into which these individual growth gradients will fall. Such allometric growth analysis has been used by other workers to indicate trends in evolutionary growth-partition. Hersh (1934) has pointed out that evolution in the Titanotheres proceeds by constant change in the proportion of dimensions as the size of the species increases. In some skull rela- tions, all the species show a uniform rela- tion between b and k and a single straight line curve is obtained, while for other rela- tions a set of curves, each distinct for a genus, results. The differences within any curve are apparently due only to differences in size that the species within a group curve attain. Each curve is based entirely upon measurements of adult individuals and the ontogenetic growth constants are not con- sidered. Similar results have been obtained with the horses. All fossil and modern horses show a distinct group trend with reference to the muzzle length-total skull length rela- tion (Robb, 1935a), while for the relation between the length of the splints and the length of the cannon bone, two distinct bands are produced, one for the four-toed and three-toed horses and the other for the one-toed forms (Robb, 1936). The horse is characterized by a progressive increase in body size from the Eocene on and, therefore, these curves represent stages in geological time. In the horses the ontogenetic curves are of the same value as the evolutionary group curve (Robb, 1935b). Lumer (1940) has shown that various tribes of 'dogs may be separated upon their different allometric trends, but here again, as in Hersh’s study, no indication of the ontogenetic development is given. When orthogenetic trends are not present the group curve will represent merely aver- ages of the ontogenetic curves. In this study of scombriform fishes, species within the same genus show distinctly different growth constants. Evolutionary relative growth (which makes use solely of adult measure- ments) should, in the strict sense, be applied only to groups where the larger species re- capitulate in their development the body proportions of the adult stages of smaller species. The scombriform fishes continue growing throughout life (with body proportions con- stantly changing) and no definite mature stage can be selected for an evolutionary comparison. Furthermore there is evidence that this group has numerous phyletic lines. Therefore, in order to set up a phyletic sys- tem for these fishes (or for any group that does not show an orthogenetic trend) on the basis of relative growth data, the onto- genetic growth constants must be consid- ered. There ought to exist among the mem- bers of a genus, or of a higher group, similar trends in these growth constants. That such trends exist in the scombriform fishes is in- dicated below. If the values of k for the growth partition of width and depth (each relative to length) are plotted against each other (Text-fig. 12), it is possible to observe the interspecific trends of the gradients. Since these trends exist, it becomes important to determine whether the ontogenetic change in propor- tions, characteristic of each species, is a clue to their phylogenetic relationships. Some evidence favorable to this view has been obtained. The lines obtained by the changes in the intensity of the growth con- stants divide the Scombroidei into two major groups — on one hand the Scombridae with the allied Trichiuridae and Istiophori- dae, and on the other the Carangidae (this point is further discussed below). In the scombroids there is a general tendency from the central mackerel type in the direction of 1943] Shapiro: Weight and Body Form in Scombroid Fishes 99 Standard length of body (in mm.) Text-fig. 9. Increase of width relative to length; logarithmic plotting both ways. Standard length of body (in mm.) Text-fig. 10. Greatest width of body relative to the standard body length; logarithmic plotting both ways. 100 Zoologica : New York Zoological Society [XXVIII: 12 Standard length of body (in mm.) Text-fig. 11. Increase of greatest width of body with increase of standard length; both variables plotted on a logarithmic scale. the elongate forms. The gradient (fc) for width increases as does that for depth, but in the sequence to the specialized mackerel- like types, the increase for the depth is slightly greater than that for the width. For the Carangidae this tendency is even more marked and results in a complete re- versal of the intensity of growth in the transverse and vertical dimensions. Per unit increase the width shows greater growth than the depth in Vomer setapinnis and there is a general trend in the direction of Trachinotus carolinus where the depth now maintains greater increase than the width. These are broad tendencies, as shown here, and it should be emphasized that for a more exact phylogenetic scheme, where generic and specific tendencies may be followed, more complete data are necessary. There is a pattern of growth that is indi- vidual for each species but each pattern conforms, nevertheless, to a general scheme indicative of the trends of the larger groups. This seriation of the gradients of growth for the species suggests that phyletic lines can be determined by the allometric method and that change in the growth 'gradients is the mechanism by which the course of evo- lution proceeds. If so, the mathematical analysis of the interspecific changes in the 1943] Shapiro: Weight and Body Form in Scombroid Fishes 101 Value of K for Width Relative to Length Text-fig. 12. The various species of the Scombroidei plotted according to the values of k for depth relative to length and width relative to length. Both ordinate and abscissa are for increasing units of k. The value for the increase of weight relative to the length ( n ) is given for each species. proportions and form of the organism should be a further tool to the phylogenist and systematist. The divergence of the gradients for the Carangidae on the one side and the Scombri- dae and immediate allies on the other (Text- fig. 12) is in harmony with the view that these gradients reflect phylogeny. The sepa- ration of lines for the carangids and scom- broids indicate broadly the phylogenetic division accepted by many workers, either as separate lines within the Scombroidei or as independent modifications of a typical percomorph (serranid) stock. The Carangi- dae are grouped with the Scombridae and related families into a single suborder, the 102 Zoologica: New York Zoological Society [XXVIII: 12 Scombroidei, by many authorities, including Jordan and Evermann (1896, p. 862), Starks (1909, 1910) and Gregory (1933, p. 300). In their view the carangids, although pos- sessing many percoid characters, appear to be even more closely related to the true scombroid families and can therefore be bracketed into a percomorph suborder, the Scombroidei. Tate Regan (1909, 1929, p. 321 ) , however, refers the carangids to the percoid suborder and the true scombriforms to a separate suborder, the Scombroidei. The primary fusiform body, illustrated by the mackerels, is a specialized stage away from the general, although somewhat deep- bodied, percomorph types. Scomberomorus maculatus has the initial growth index char- acteristic for the fusiform body ( a — 0.0000288 ) . This mackerel, however, main- tains the early fusiform condition as it grows {k for depth = 0.87, Table 2). The most advanced stage away from the primary fusiform body is the elongate type repre- sented by Trichiurus lepturus. The fusiform condition has been lost by excessive elonga- tion (a=0. 0000000674) . Makaira may be classified as a specialized fusiform type. The initial growth index is very low (a = 0.000000161 for M. n. ampla and 0.000000118 for M. n. marlina ) and like Trichiurus they are very slender fish early in development. They attain their fusiform condition only through the intense gradients at work dur- ing their growth. There is no evidence of any Scombridae appearing until the Eocene, when they are already fusiform, primarily elongated forms with a vertebral number of 30-50. In this group total length has increased due to the increased number of vertebrae (the primi- tive acanthopterygian number is 10 + 14). The increase in vertebral number would seem to be a normal method of assuming the fusiform condition early in ontogenetic growth and maintaining it throughout life, providing there was no excessive increase in depth. The geological evidence would place the mackerel group as an early (Eo- cene) offshoot of its percomorph ancestor. The mackerels illustrate the first step to- ward the elongated form represented by the Trichiuridae. The independent trichiurid line begins with Ruvettus, which is close to the mackerel type, and follow through by successive stages in modification and elon- gation until the end stage represented by Trichiurus is reached. Trichiurus is secon- darily elongated (its vertebral number is 159). The marlins, with 12+12 or 12+13 ver- tebrae, are already specialized with respect to this deviation from the basic percomorph pattern of 10+14. Early in their post-meta- morphic development they are extremely long-bodied as indicated by their initial growth index. These fish lose their early slenderness and attain a thick-fusiform streamlined condition, during their growth, through the maintenance of steep gradients; that is, through increase in the gross size of their parts. Thus the Istiophoridae pre- sent a distinct line of specialization resem- bling that of the Scombridae, in that both have a fusiform body — attained in the mar- lins during ontogenetic growth but present, however, at the beginning of and maintained during post-metamorphic growth in the mackerels. We are dealing with two widely divergent lines of evolution, each with strik- ing differences (in vertebral number and growth constants) but nevertheless with considerable anatomical evidence to indicate a common heritage. The bluefish ( Pomatomus ) has been con- sidered the connecting link between the Ser- ranidae and the Carangidae (Tate Regan, 1909 ) , because of the intermediate character of its skull, general body form, caudal ped- uncle and caudal fin. In its external body form, however, it seems to have assumed the specialized contours of the fusiform mackerels, with which other authors have classified the genus. Its vertebral number still remains the primitive 10+14 perco- morph number, but its form of body appears to be a case of parallel development of the fusiform type of body correlated with a predatory habit. The change has not been incurred in the mackerel manner, by an in- crease in vertebral number, but rather as in Trachinotus by a modification in relative growth (the initial growth index = 0.0000229 for Pomatomus) . Since deepening of the body does not take place during growth as it does in Trachinotus ( k for depth = 0.88 rather than 1.13, Table 2), the fusiform external mackerel condition established in the young stages of growth is maintained. Pomatomus could have well arisen from a serranid ancestor indepen- dently from the carangids — but perhaps from the same basic stock. The relative growth method does not in- dicate the direction of phyletic lines and thus creates some difficulty in interpreting the line of carangid evolution. Since the in- creasing specialization of the scombroids ap- pears to follow the line of divergence as in- dicated by the allometric growth method (see Text-fig. 12), such a trend may also be characteristic of the carangids. The value for the initial growth index is high for the deep-bodied Selene vomer (a = 0.0000483) and Vomer setapinnis (a = 0.0000648) but these fish become slenderer as they grow in size. Caranx hippos and C. crysos show the same relations for their post-metamorphic development. These fish can be designated primary deep-bodied types, regardless of whether the deep body of the early stages recapitulates the deep body of extinct an- cestors or is a juvenile adaptation. Trachino- 1943] Shapiro : Weight and Body Form in Scombroid Fishes 103 tus carolinus shows a distinctly opposite trend. It is slenderer and more fusiform in the early stages of growth (a = 0.0000264) but becomes deepened during growth ( k for depth is 1.13). It can thus be designated a secondarily deepened type. Apparently two interpretations of this evidence are possible. These differences in the growth relations might well serve as evi- dence either that Trachinotus has been in- dependently derived from the typical caran- gid line, or, as Gregory (1933) maintains, it is a longer-bodied derivative of the more primitive deep-bodied compressed types. The seriation as seen in Text-fig. 12 would seem to favor the interpretation that Trach- inotus is a more specialized carangid in the direct line of ascent from the deep-bodied Selene, Vomer and Caranx. The initial growth index for Trachinotus is similar to that for the specialized Scomber omorus (Table 1). Trachinotus is only fusiform early in its post-metamorphic growth, but reverts to a deep-bodied condition during its ontogenetic growth. The Scombridae by increasing the number of vertebrae, have taken what seems to be a more normal and advanced method of assuming the fusiform condition. Thus their manner of attaining the fusiform condition was a more complete break from their ancestral type than Trach- \ inotus was able to accomplish from the typi- cally deep-bodied carangids or deep-bodied carangid ancestors, if it has come off inde- I' pendently from the carangid ancestors. Summary. Relative growth methods are used to de- lj termine the relation between weight and 1 body form in several fishes generally classed in the suborder Scombroidei. The cube relationship between weight and length is only an approximation and in most of the species analyzed is not even closely i realized. An allometric increase or decrease I of weight in relation to length is corre- 1 lated with the fact that the dimensions of the body are not necessarily a linear function j of length. This non-linear relation may hold . even when an isometric weight — cube-length condition is realized, for the changes in different dimensions may be compensatory. Fish that attain a greater total size show a smaller, bulkiness (initial growth index) early in their post-metamorphic growth, whereas the deep-bodied shortened fish have greater relative bulk at this stage of their growth. However, fish that are slim early in ontogenetic development tend to possess a high positive allometric distribution of weight during growth, and vice versa. This inverse relationship of the initial growth index to the growth constant is exponential and may be of some value in determining phylogenetic trends. The depth in some species shows positive allometry, in others it is negative. All species studied, however, show positive allometry for the relationship of width to length. It is thus indicated that fishes which become rela- tively lighter during growth can do so only through a relative decrease in depth. The differential growth in width proceeds more rapidly than the growth in depth in all the species with the exception of Trachinotus carolinus, where this condition is reversed. Thus, in the majority of scombriform fishes studied, the change to a larger fish proceeds more rapidly along the transverse than along the vertical axis. In general there is a negative correlation between the values of the initial growth index and the values of the growth constants for the dimensional proportions. Species that are deeper or wider initially become relatively slenderer or thinner with age, and vice versa. Closely allied members of the various families have similar coefficients of growth partition, but specific differences are never- theless maintained. The interspecific trends of the ontogenetic growth constants suggests the possibility that phyletic lines can be determined by the allometric method. The pattern of growth that is individual for each species seems to conform to a general scheme indicative of the evolutionary trends of the larger groups. Differences in relative growth trends within the Scombroidei of earlier authors charac- terize separate groups that have been or might be regarded as representing either separate divisions within the Scombroidei or entirely independent parallel phyletic lines. Bibliography. Champy, C. 1929. Le croissance dysharmonique des caracteres sexuels accessoires. Ann. Sc. Nat. Zool, (10), 12: 193-244. Crozier, W. J. and Hecht, S. 1913. Correlations of weight, length, and other body measurements in the weakfish, Cynoscion regalis. Bull. U. S. Bureau of Fisheries, 33: 139-148. De Beer, G. L. 1940. Embryology and Taxonomy. In ‘The New Systematics.’ Edited "by Julian Huxley. Oxford: Clarendon Press: 365-393. Glaser, O. 1938. Growth, time, and form. Biol. Rev., 13: 20-58. Gray, J. 1929. The kinetics of growth. Brit. Journ. Exp. Biol., 6: 248-274. 104 Zoologica: New York Zoological Society Gregory, W. K. 1933. Fish skulls, a study of the evolution of natural mechanisms. Trans. Amer. Philos. Soc., N. S., 23: 75-481. Hecht, S. 1916 Form and growth in fishes. Journ. Morph., 27 (2) : 379-400. Hersh, A. H. 1931. Facet number and genetic growth constants in bar-eyed stocks of Droso- phila. Journ. Exp. Zool., 60: 213-248. 1934. Evolutionary relative growth in the Titanotheres. Amer. Nat., 68: 537-561. Huxley, J. S. 1924. Constant differential growth-ratios and their significance. Nature, 114: 895-896. 1932. Problems of relative growth. London, Methuen & Co. Huxley, J. S. and Tessier, G. 1936. Terminology of relative growth. Na- ture, 137: 780-781. Jordan, D. S. and Evermann, B. W. 1896. The fishes of North and Middle Ameri- ca. A descriptive catalogue of the species of fish-like vertebrates found in the waters of North America north of the Isthmus of Panama. Bull. U. S. Nat. Mus., No. 47, Part 1. Key'S, A. B. 1928. The weight-length relation in fishes. Proc. Nat. Acad. Sci., 14: 922-925. Lumer, H. 1939. Relative growth of the limb bones in the anthropoid apes. Human Biology, 11: 379-392. 1940. Evolutionary allometry in the skeleton of the domesticated dog. Amer. Nat., 74: 439-467. Marshall, S. M., Nicholls, A. G., and Orr, A. P. 1937. On the growth and feeding of the lar- val and post-larval stages of the Clyde herring. Journ. Mar. Biol. Assoc., 22: 245-267. Meek, A. 1905. Growth of flatfish. Sci. Investig. Northumberland Sea Fish. Commis .: 58. Needham, A. E. 1935. Relative growth in the jaws of certain fishes. Proc. Zool. Soc. London, 2 (Part 4) : 773-784. Pezard, 1918. Le conditionnement physiologique des caracteres sexuels secondaires chez les oiseaux. Bull. Biol. Fr. et Belg., 52: 1-76. Regan, C. Tate 1909. On the anatomy and classification of the scombroid fishes. Ann. Mag. Nat. Hist., 3: 66-75. 1929. Fishes. Encycl. Brit. (14), 9: 305-328. Robb, R. C. 1935a. A study of mutations in evolution. I. Evolution in the equine skull. Jour. Genet., 31 : 39-46. 1935b. A study of mutations in evolution. II. Ontogeny in equine skull. Jour. Genet., 31: 47-52. 1936. A study of mutations in evolution. III. The evolution of the equine foot. Jour. Genet., 33: 267-273. Shapiro, S. 1938. A study of proportional changes dur- ing the post-larval growth of the blue marlin (Makaira nigricans ampla, Poey). Amer. Mus. Novitates, No. 995: 1-20. Simpson, G. G. and Roe, A. 1939. Quantitative zoology, numerical con- cepts and methods in the study of recent and fossil animals. McGraw- Hill Book Co., New York. Spencer, H. 1871. The principles of biology. Vol. 1. New York. Starks, E. C. 1909. The scombroid fishes. Science, N. S., 30: 572-574. 1910. The osteology and mutual relation- ships of the fishes belonging to the family Scombridae. Journ. Morph., 21: 77-99. Thompson, D’Arcy 1917. On growth and form. Cambridge University Press. Van Oosten, J. 1928. Life history of the lake herring ( Leucichthys artedi Le Sueur) of Lake Huron as revealed by its scales, with a critique of the scale method. Bull. U. S. Bureau of Fisheries, 44: 265-428. Williams, G. and Needham, A. E. 1938. On relative growth in Pinnotheres pisum. Proc. Zool. Soc. London, 108 (Series A) : 539-556. Cor dier : Breeding of Quetzal in Captivity 105 13. A Record of the Successful Breeding of the Quetzal ( Pharomachrus mocinno costaricensis ) in Captivity. Charles In March, 1942, I had the good fortune to meet in San Jose, Costa Rica, Senora Amparo Zeledon, widow of the well-known ornithologist who died a few years ago. At that time I was starting on a collecting trip having as its objective the capture of the Bare-necked Umbrella Bird ( Cephalop - terus ornatus glabricollis ) and other rare birds for the New York Zoological Society. It was with considerable elation that I noticed at Senora Zeledon’s home a large, well-planted aviary, approximately 30 X 15 j X 10 feet, in which was a pair of Costa Rican Quetzals, Pharomachrus mocinno cos- taricensis, in superb condition. The male i trailed behind in its flight a pair of upper tail coverts nearly a yard long. Subsequently I was able to purchase this pair and the birds are now (February, 1943) on exhibi- tion in the New York Zoological Park, to- gether with several other Quetzals. The most interesting part of my discovery was that Senora Zeledon told me that the Quetzals had reared young successfully on two occasions and that she had made notes of these events at the time. She was kind enough to lend me her notes, which I have translated and reproduce herewith : Senora Zeledon’s Notes on the Breeding of the Quetzal. “We noticed the young bird peeping through the entrance hole on August 8, 1940, and on the 12th it left the nest. “Both male and female were feeding it as long as it was confined to the nest, but once outside, only the mother cared for it, feeding it earthworms (Of all things — Translator) in the morning, and fruit (By which is meant cut-up little plantains known locally as “guinea” — Translator). On the 16th of the same month it fed itself for the first time. I suppose the birds started nesting at the end of June. “The tree trunk in which they nested had been in the aviary for more than two years, exposed to the weather, before they showed any interest in it, the male preferring to enter a bigger, irregularly-shaped hole in the wall forming one side of the aviary. The CORDIER female, however, would only reluctantly go near it, the entrance hole probably being too wide. Fortunately they decided on nesting in the tree trunk after two long years, the wood having become soft enough to be worked on by them. “The male moulted at the end of August and lost its long tail coverts at the end of September. “The plumage of the young bird at the moment it left the nest was uniform coffee- color, except the wing coverts which were green. “Both parents had been kept for three years in the big aviary and previously for two years in a much smaller aviary in which they would get very wild as soon as ap- proached. In the big enclosure they are quite steady and take no notice of visitors. The pair was composed of a young male and an adult female purchased from the natives.^ "February, 1941 : The male, already in good plumage, started to enter the nest in the middle of February and to clean it out, carrying loads of dust in his breast feathers which he would shake out, once outside. Both started uttering their calls and the female entered the nest some days later, mornings and afternoons. Both adults started chasing the young Quetzal from the vicinity of the nest and stayed continuously near it, even in March. “On March 31, 1941, I noticed the female coming out of the nest to let the male get in. Shortly afterward the male came out and the female re-entered. “April 19: Today I had the gardener trim the trees in the aviary and he noticed two young birds in the nest, maybe 8 days old. “Five days later we found a dead young Quetzal on the ground. One side showed signs of having been injured. Possibly the birds got frightened when the gardener worked in the aviary and threw one of the nestlings out. “The young Quetzal, reared the previous nesting season, is still being relentlessly persecuted by its parents. On Friday, May 2, they bit it almost to death. I took it out, did my best to cure it, but found it dead the 106 Zoologicn: Nero York Zoological Society [XXVIII :13:1943] following day. I had made the mistake of putting it back in the aviary. Even when it was lying dead on the ground, the female would swoop furiously down on it. “On May 12, 1941, this season’s young Quetzal could be seen at the nest hole. It left the nest on May 18, early in the morn- ing. Male and female re-entered the empty nest several times that day.” Further Observations in the Field. On discussing the birds’ performance, Senora Zeledon said she believed she knew why the Quetzals waited two years before nesting. She thought the wood in the tree trunk was too hard and they could not work it and shape it to their liking. In 1942 she had the old tree trunk replaced but no at- tempt at breeding was made, the wood prob- ably again being too hard. I measured the old trunk and found it to have a diameter of about 12 to 13 inches. The nesting hollow was approximately 9 inches in diameter. At the bottom was a shallow cup hollowed out, leaving a rim all around of a width of about 2 inches. The lower rim of the entrance hole, almost 5 inches in diameter, was about 6 inches from the rim of the cup. Any future attempt at breeding should take into consideration the condition of the log, which should be so well rotted through- out that bits can be pried off with the finger- nail, and should have an entrance hole and nesting cavity smaller than the measure- ments given above. The birds will accom- modate it to their liking themselves. It would also be a good plan to wet the log thor- oughly from the outside, with a hose, daily, to reproduce natural conditions. Quetzals breed in cloud forests which are dripping wet almost the year around. While in Costa Rica I found two nests at an alti- tude of 4,000 feet. One was located deep in the cloud forest in a tree stump about 15 feet tall, so rotten that it swayed when a finger was pressed against it. The diameter of the stump was at most 9 inches, which goes to show that occasionally the Quetzal takes to cramped quarters. The entrance hole was about 12 feet from the ground. I caught the male, who was brooding inside, by walking noiselessly up to the nest and covering the hole with a butterfly net on a short pole. This was in the month of June. The second nest was situated in an enor- mous tree trunk, charred by fire and stand- ing in a clearing. The entrance hole was a good 20 feet from the ground. The native who climbed up to it risked his neck by doing so. The nest was deep. He brought up two eggs, the size of pigeon’s eggs, uniform- ly blue in color. He replaced them but the birds abandoned the site. The natives said that they invariably do this when the eggs have been touched. At first I thought the Quetzal lives in association with the great woodpecker, whose abandoned nest he would take over and accommodate to his liking, but later I was not so sure of this. Quetzals breed up to 10,000 feet and at the higher levels the big woodpecker is not found, being replaced by a medium-sized woodpecker, so that the Quetzal would have to accommodate a much smaller nest. Despite the Quetzal’s diminutive feet and short bill, he must be a good carpenter. I once observed in Guatemala a Red-bellied Ti’ogon hollowing out, with feet and bill, an occupied termite nest, and these are quite hard. Nigrelli & Smith: Leeches on Tumors of Marine Fishes 107 14. The Occurrence of Leeches, Ozobranchus branchiatus (Menzies), on Fibro-Epithelial Tumors of Marine Turtles, Chelonia my das (Linnaeus). Ross F. Nigrelli New York Zoological Society & G. M. Smith Yale Medical School and New York Zoological Society (Plates I-III). In the course of study on parasites of Chelonia mydas, Nigrelli (1941) reported the presence of large numbers of leeches, Ozobranchus branchiatus (Menzies, 1791), on fibro-epithelial tumors removed from turtles caught off the West Coast of Florida and sent to us by Mr. Stewart Springer. The leeches were previously described from the same host by MacCallum and MacCallum (1918) but these investigators did not men- tion the tumors. The tumor masses were removed from the neck and eyelids of turtles and on a piece one-half inch square as many as fifty leeches in various stages of growth were counted. The largest of these ectoparasites measure about 10 mm. in length and 5 mm. in width (Fig. 1"). They are strongly contracted as the result of preservation, and it is assumed that when alive and fully expanded they must be at least three times as long. There is a powerful sucker at the posterior end of the animal and the body possesses lateral filaments which are described as gills (Fig. 1, G). The mouth is small, round in shane and furnished with smooth thick lips. On the dorsal side of the anterior region may be found two eye spots (Fig. 2). although these structures are not always demonstra- ble. An examination of the intestinal con- tents shows that these annelids feed almost entirely on the blood of the host. The tumors are of the same type described by Smith and Coates (1938). The tissues appear as papillary growths or as round fibromata arising from the skin. The leeches are attached to the tumors by their pos- terior suckers and on the papilloma they are usually found buried deep in the crypts formed by such growths (Fig. 2). In many j instances it is difficult to distinguish be- tween the growth proper and the leeches (Fig. 3). Microscopically, the papilloma shows a definite thickened epithelium with a certain amount of keratin formation ( Figs. 5, 6, 7) . The stroma is highly vascular, especially in regions where the leeches are attached (Fig. 7). The fibromata is characterized by a dense, intermingling fibrous growth, cov- ered at the surface with a normal amount or slightly thickened epithelium. Nests of leeches also have been found attached on these smooth growths (Fig. 4). The etiology of the growths is still un- known. Smith and Coates (1938) discuss the possibility of a viral or parasitic agent. Later (1939) they reported the presence of trematode eggs in many of the tumors ex- amined. These eggs were eventually identi- fied by Smith, Coates and Nigrelli (1941) as those of blood flukes originally described by Leared (1862) as Distomum constrictum. According to Smith and Coates (1939), “It is probable that ova are deposited in pre- existing vascular tumor tissue by the migrating blood flukes, and remain there without affecting the subsequent course of the growth.” They pointed out, further, that the localization of ova in the stroma and in the venous or lymphatic spaces of the turtle tumors is somewhat similar to the condition in the tissues of the urinary bladder in human bilharziasis. “Such an infection of the human bladder, as is well known, may result in papillomatous and malignant changes.” Just what role the ectoparasitic annelids play in the development of these turtle tum- ors is not certain. They may act as vectors for viral or other parasitic forms that may be the causative agent, although histological 108 Zoologica : New York Zoological Society [XXVIII :14 :1943] examination of a number of sectioned and stained leeches showed no evidence that they are involved in the transmission of helmin- thic parasites of the turtles. However, it is altogether possible that the leeches may have some effect on the subsequent growth of the fibro-epithelial tumors. It is known that they feed on the turtle’s blood ; the lat- ter is prevented from clotting by the action of hirudin. Such a continued flow of blood makes an excellent medium, supplying the necessary nutriment for these slow growing tumor cells. That hirudin, per se, may have some effect on the growths is indicated by certain experiments reported in the litera- ture involving the use of heparin. Thus, Lig- neris (1930) found that transplantations of melanoma of angora goat on homologous animals were successful only when heparin- goat-plasma was used as a nutritive and pro- tective medium surrounding the trans- planted tumor particles. Zakrzewski (1932) showed that better proliferation of tissues was obtained when heparin was added to blood serum used as the culture medium. Summary. 1. Leeches, Ozobranchus branchiatus (Menzies, 1791), were found associated with fibro-epithelial tumors of marine turtles. Chelonia mydas (Linnaeus). 2. The role of the buccal secretion of the leeches (hirudin) is discussed in connection with the growth of the tumor tissues. References. Ligneris, M. J. A. 1930. The use of Heparin-Plasma for graft- ing of spontaneous mammalian tumors into Homologous Animals. British J. Exp. Path., 11: 249-251. MacCallum, W. G. and G. A. MacCallum. 1918. On the Anatomy of Ozobranchus branchiatus (Menzies). Bull. Amer. Mus. Nat. Hist., 38 : 395-408. Nigrelli, Ross F. 1941. Parasites of the Green Turtle, Chelonia mydas (L), with Special Reference to the Rediscovery of Trematoda Described by Looss from This Host Species. J. Para., 27 : 15-16. Smith, G. M. and C. W. Coates. 1938. Fibro-Epithelial Growths of the Skin in Large Marine Turtles, Clielonia mydas (Linnaeus). Zoologica, 23: 93-98. 1939. The Occurrence of Trematode Ova, Hapalotrema constrictum (Leared), in Fibro-Epithelial Tumors of the Marine Turtle, Chelonia mydas (Lin- naeus). Zoologica, 24: 279-382. Smith, G. M., C. W. Coates and R. F. Nigrelli 1941. A Papillomatous Disease of the Gall- bladder Associated with Infection by Flukes, Occurring in the Marine Tur- tle, Chelonia mydas (Linnaeus). Zoo- logica, 26:13-16. Zakrzewski, Z. 1932. Die Rolle des Prothrombins und Heparins bei der Proliferation und Differenzierung von Geweben. Unter- suchungen in vitro. Arch. Exp. Zell- forsch. bes. Gewebezucht., 13: 152-175. EXPLANATION OF THE PLATES. Plate I. Fig. 1. Photograph of leeches removed from turtle tumors. G, gill filaments; S, posterior sucker. About 4 X. Fig. 2. Leeches buried in papilloma crypts. Note minute eye spots. About 3 X. Plate II. Fig. 3. Nest of leeches among the papillae. Some of the smaller leeches are diffi- cult to distinguish from growth prop- er. 3 X. Fig. 4. Nest of leeches, in various stages of development, on smooth fibromata of the turtle. Plate III. Figs. 5, 6, 7. Photomicrographs of the turtle tumor. Note relationship of leech to tumor. The stroma of the tumor is highly vascular. About 5 X. NIGRELL1 & SMITH. PLATE I. FIG. 1. FIG. 2. THE OCCURRENCE OF LEECHES. OZOBRANCHUS BRANCHIATUS (MENZIES). ON FIBRO-EPITHELIAL TUMORS OF MARINE TURTLES. CHELONIA MYDAS (LINNAEUS). NIGRELL1 & SMITH. PLATE II. FIG. 3. FIG. 4. THE OCCURRENCE OF LEECHES. OZOBRANCHUS BRANCHIATUS (MENZIES), ON FIB RO- EPITHELIAL TUMORS OF MARINE TURTLES. CHELONIA MYDAS (LINNAEUS). NIGRELLI a SMITH. PLATE FIG. 6. THE OCCURRENCE OF LEECHES, OZOBRANCHUS BRANCHIATUS (MENZIES). ON FI BRO- EPITHELIAL TUMORS OF MARINE TURTLES, CHELONIA MYDAS (LINNAEUS). FIG. 7. FIG. 5. NEW YORK ZOOLOGICAL SOCIETY General Office: 630 Fifth Avenue, New York City OFFICERS Fairfield Osborn, President Alfred Ely, First Vice-president Laurance S. Rockefeller, Second Vice-president Harold J. O’Connell, Secretary Cornelius R. Agnew, Treasurer SCIENTIFIC STAFF General John Tee-Van, Executive Secretary Jean Delacour, Technical Adviser William Bridges, Editor and Curator of Publications Claude W. Leister, Education Donald Marcy, Associate, Education Zoological Park Lee S. Crandall, General Curator & Curator of ' Birds Leonard J. Goss, Veterinarian Claude W. Leister, Associate, Mammals Harry C. Raven, Associate, Anatomy John Tee-Van, Associate, Reptiles Grace Davall, Assistant to General Curator Charles M. Breder, Jr., Director Christopher W. Coates, Aquarist Ross F. Nigrelli, Pathologist Myron Gordon, Research Associate in Genetics G. M. Smith, Research Associate in Pathology Homer W. Smith, Research Associate in Physiology Aquarium Department of Tropical Research William Beebe, Director Jocelyn Crane, Research Zoologist Henry Fleming, Entomologist William K. Gregory, Associate Gloria Hollister, Associate John Tee-Van, Associate Mary VanderPyl, Associate Editorial Committee Fairfield Osborn, Chairman William Beebe Charles M. Breder, Jr., William Bridges Jean Delacour Claude W. Leister John Tee-Van . ZOOLOGICA SCIENTIFIC CONTRIBUTIONS of the NEW YORK ZOOLOGICAL SOCIETY VOLUME XXVIII Part 3 Numbers 15-19 Published by the Society The Zoological Park, New York 60, N. Y. December 6, 1943 CONTENTS 16. On the Relationship of Some Common Fishes as Determined by the Precipitin Reaction. By Douglas G. Gemeroy. (Text-fig- ures 1-7) 109 16. Spiders of the Families Lyssomanidae and Salticidae (Magoninae) from British Guiana and Venezuela. By Jocelyn Crane. (Text-figures 1-4) 125 17. Notes on the Breeding of the Empress of Germany’s Bird of Para- dise in Captivity. By Prince K. S. Dharmakumarsinhji of Bhavnagar. (Plate I) 139 18. Evidence of Healed Hungerosteomalacia (Late Rickets) in a Green Monkey ( Cercopithecus sabaeus). By Joseph Pick. (Plate I) 146 19. Eastern Pacific Expeditions of the New York Zoological Society. XXXII. Mollusks from the West Coast of Mexico and Central America. By Leo George Hertlein & A. M. Strong. (Plate I) 149 i Gemeroy: Relationship of Some Common Fishes 109 15. On the Relationship of Some Common Fishes as Determined by the Precipitin Reaction. Douglas G. Gemeroy From the Zoological Laboratory, Rutgers University (Text-figures 1-7). Introduction. It is generally agreed that the goal of taxonomy is the tracing of phylogenetic re- lationships and since the time of Darwin the emphasis in classification has been on the phylogenetic method in taxonomy. The chief source of data in phylogeny and classifica- tion has been morphological, and there is a need of some independent source of data with which to check the morphological inter- pretation. Such an independent source is available in the serological or biochemical constitution of animal proteins. Nuttall’s pioneer studies with the pre- cipitin reaction applied to the determination of animal relationships have been followed by many others in both plants and animals, Mez and Ziegenspeck (1926 ) ; Landsteiner (1936) and others whose work is covered in the recent review of Boyden (1942). In the thirty-nine years since the publication of Nuttall’s “Blood Immunity and Blood Rela- tionship,” progress has been made both in results and their interpretation, but the im- portance of serological data in taxonomy and phylogeny is still a controversial matter. In order to provide further data from which a conclusion might be drawn as to the relative importance of the serological method of at- tack on phylogeny and taxonomy, a study of the relationships of common fish was under- taken, using the old or common precipitin technique of the ring test and the new pho- tron’er (photronreflectometer) methods in- troduced by Libby (1938) . Such an approach is both quantitative and objective and is in- dependent of morphology which it may com- plement and whose findings it may check. Further it may shed new light on many groups whose relationships are still uncer- tain with the methods available to morpholo- gy. The present paper represents the results of more than 500 tests upon the blood sera of 31 species of fresh water and marine fishes. Materials and Methods. I. Sources of Serum. The sera for the data of the present paper were procured from the species of fish listed below. Both common and scientific names are given. Common Name Scientific Name Lamprey Common Shark Yellow Shark Scyllium Dog-Fish Barn-door Skate Eagle Ray Sturgeon Garpike Bowfin Tarpon Brook Trout Rainbow Trout Brown Trout Carp Buffalo Fish Red Horse or Mullet Petromyzon marinus ( Linnaeus ) Carcharhinus sp. Hypoprion brevirostris Poey Scyllium canicula (Linnaeus) Squalus acanthias Linnaeus Raja laevis Mitchill Aetobatis narinari (Euphrasen) Acipenser rubicundus ( LeSueur ) Lepisosteus osseus (Linnaeus) Amia calva Linnaeus Tarpon atlanticus (Cuvier and Valenciennes) Salvelimts fontinalis (Mitchill) Salmo irideus Gibbons Salmo fario Linnaeus Cyprinus carpio Linnaeus Ictiobus cyprinella (Cuvier and Valenciennes) Moxostoma aureolum . (LeSueur) 110 Zoologica: New York Zoological Society [XXVIII: 15 Catfish Gaff -topsail Catfish Sea Catfish Pike Muskallunge Barracuda Dolphin or Dorado Black Bass Large Mouth Bass Rock Bass Blue Gill Sun Fish Perch Pickerel Kingfish Ameiurus lacustris (Walbaum) Bagre marinus (Mitchill) Galeichthys felis (Linnaeus) Esox lucius Linnaeus Esox masquinongy Mitchill Sphyraena barracuda (Walbaum) Coryphaena hippurus Linnaeus Micropterus dolomieu Lacepede Huro floridana (LeSueur) Ambloplites rupestris (Rafinesque) Helioperca incisor (Cuvier and Valenciennes) Perea f lav escens (Mitchill; Stizostedion vitreum (Mitchill) Menticirrhus americanus ( Linnaeus) II. Collection of Serum. The blood sera of the different fish listed above were obtained from various sources. The fresh water species were procured from Northern Ontario, the State Fish Hatchery at Hackettstown, New Jersey, the New York Aquarium and live fish markets in New York City. The salt water species were pro- cured from a number of the different Bio- logical Laboratories along the Atlantic and Gulf Coasts.1 III. Handling of Serum. Two methods are used regularly to obtain blood from fish. 1. By syringe direct from the heart. 2. By drip method from the caudal ar- tery. With the fresh water species the first method was used throughout, while with the salt water species the second method was used in all cases except the shark, where the pericardial cavity was cut open and the blood allowed to drain into a receiving bowl direct- ly. All bloods collected were allowed to clot for a period ranging from 6 to 12 hours. The serum that had been expressed from the clot by this time was poured off and centrifuged at 2,500 r.p.m. until free from cellular ele- ments. In all cases where the serum was stored without any preservative it was first filtered through a Seitz filter and then bot- tled under sterile conditions. The greater portion of the sera collected however, was preserved by the addition of 0.02 ml. of 10 per cent, formalin to every ml. of serum. A few samples collected in the fall of 1940 were preserved by the addition of 1 ml. of 1:1000 merthiolate solution to each 10 ml. of serum. All the formolized sera were kept at room temperature during the course of the investi- gation, while the native and merthiolated 1 The author wishes to thank Drs. C. M. Breder, Jr. and Alan A. Boyden who contributed all the marine species used and Mr. C. O. Hayford, Superintendent of the State Fish Hatchery at Hackettstown, New Jersey, for a number of the fresh water species. He also wishes to express his appreciation to Drs. T. C. Nelson, C. M. Breder, Jr. and Alan A. Boyden for valuable aid and criticism of this paper. Submitted to the Graduate Faculty of Rutgers University in partial fulfillment of the requirements for the Degree of Doctor of Philosophy. sera were kept in the electric refrigerator, except when small samples were being used for antisera production or in carrying out the required tests. IV. Protein Determinations. Prior to the use of the different sera in antibody production, total nitrogen and non- protein nitrogen determinations were car- ried out by the micro-Kjeldahl digestion method of Koch and McMeekin (1929). From these figures the total protein was obtained by multiplying the T. N. — N.P.N. by factor 6.25. For N.P.N. determinations the protein was precipitated with trichlor- acetic acid. In the ring test where the end point of the reaction is the standard em- ployed in determining relationships, com- parable amounts of protein must be used in the different dilutions of the various anti- gens to give a quantitative basis for such relationships. In the tests reported here only standard and comparable antigen dilutions have been used. The protein concentration of the fish sera in gms./lOO ml. of serum varied from 1.50 gms. to 5.50 gms. The majority, however, were between 2.5 and 3.5 gms. The N.P.N. concentration in mgms./lOO ml. of serum ranged from 20 to 125 mgms. with the ex- ception of the elasmobranchs. Here the highest figures were recorded, being 1,000 mgms. or more in practically all members of this group. V. Injection Methods. Where antisera of maximum specificity are desired a single series of injections is administered. These injections are made on alternate days with the initial injection in all cases being 5 milligrams of protein per kilogram of body weight of the injected rab- bit. Each subsequent injection is double the amount of the previous one until four in- jections in all are given. Where decreased specificity is desired in an antiserum to in- crease the range of cross reactions, multiple series of injections are required, Wolfe (1935). Preliminary tests with fish blood serum showed that the range of cross re- 1943] Gemeroy: Relationship of Some Common Fishes 111 actions in the majority of the antisera made to a single series of injections was rather restricted. To broaden the range of reaction multiple series of injections were employed with three rabbits. In multiple series injec- tions, the initial series is followed after a lapse of one week by a second and even a third series of injections. Eight days after the last injection, a trial bleeding of the animal was taken to deter- mine whether the antiserum was of desired potency. In all but one case the antisera were sufficiently potent. On the following day the rabbits were anesthetized and all the blood obtainable was drawn directly from the heart by means of a large syringe and transferred to a finger bowl. The blood was allowed to clot in the refrigerator for eight |i to twelve hours, after which the expressed serum was decanted off, with minimum dis- turbance to the clot, and centrifuged at 2,500 r.p.m. until clear of cellular elements. The antiserum was then passed through a Seitz filter and bottled under sterile conditions. All the antisera thus treated were stored in the refrigerator at 4°C. ± 1°C. until ) needed. In no case where such a procedure , was canned out did contamination occur in the refrigerated antisera. ' ' VI. Ring and Photron’ er Tests. The ring and photron’er tests were carried out in the usual manner, the technique being that described by Boyden (1942). Experimental Results. The results of the different tests carried out by the ring and photron’er methods are shown in Tables I, II, III, IV. The per- centage relationship found with the various sera used in the heterologous reactions is shown as a per cent, of the homologous tests with both ring and photron’er methods. With the ring test, duplicate readings were taken with all reactions, fresh dilutions of antigen and antisera being made up before the second series of tests was run. The anti- sera used in the ring tests were diluted 1:1 in all reactions, while with the photron’er they were used in full strength, as it was found in preliminary readings that diluting the antisera 1:4 eliminated practically all the heterologous reactions. Most of the anti- sera produced by means of a single series of injections were highly specific, hence no dilu- tion was necessary. Supplementing the tables are representa- tive figures depicting homologous and heter- ologous reactions through their complete range as recorded on the photron’er, also representative graphs of the ring test show- ing the per cent, relationship values as de- termined from the titers of the homologous and heterologous antigens used. Symbols. A number of symbols are used in the tables and figures ; the explanation of which is as follows : (Fo) — Antigen was formolized. Where this symbol appears after an anti- serum it means that the antiserum was produced by the use of a for- molized antigen. In no case was the antiserum formolized. (N ) — Antigen was used without any pre- servative in all reactions. Antisera were produced with the native an- tigen. (S3) — Antiserum produced by means of a triple injection series. Discussion of Experimental Results. I. Technical Considerations. (a) Preservation of antigens with formalin. As it is necessary to use some type of preservative when antigens are collected in the field, formalin was employed wherever such sera were preserved. From the unpub- lished data of M. A. Carriker, working in this laboratory, it was found that 0.2 per cent, formalin could safely be used for the field preservation of animal sera. In the present series of investigations, comparative tests were carried out to determine to what extent such treatment had altered the spe- cies specificity of the different blood sera. Text-figures 2 and 3 give an indication of the parallel results obtained when formolized and native antisera were reacted with the antigens used in their production. The same group of heterologous antigens show parallel relationship with the antinative and anti- formolized antisera. Such results indicate that the use of formalin as a preservative under these conditions has not altered to any appreciable extent the “relationship” values shown by these two series of elasmo- branch blood sera. (b) . Comparison of photron’er and ring test methods. The ring test method of using the precipi- tin reaction in the determination of animal relationships has been the standard for many years and has been used successfully by many workers in serological investiga- tions. With species that are not too closely related, this method is still valuable, as the range of reactions is great enough to include some distantly related species. Although very sensitive at low concentra- tions of precipitated protein, the ring test as commonly performed gives only the end point, with no measure of the degree of re- action at intermediate levels. The pho- tron’er, although less sensitive at the end point, gives a quantitative measure of the degree of reaction at every concentration of 112 Zoologica: New York Zoological Society [XXVIII: 15 the reacting substances. This instrument is far more effective than the ring test, there- fore, where close relatives are involved. The ring test method can be likened to a low power microscope with its broad field of vision, while the photron’er acts like a much higher powered instrument. It pulls apart, as it were, antigens that show no difference in titer with the ring tests. When the two methods are used in the pre- cipitin reaction, one complements the other. Where rather widely separated species are being tested, the photron’er is, as a rule, highly specific and may not show any re- action in such cases. Where the ring test is employed and only the end point of the re- action is determined, the sensitivity of this method is apparent. By its use, cross re- actions that can not be determined with the photron’er are established with the ring tests. II. Bearing of results on the problem of fish relationships. (a). On animal relationships in general. As a fundamental approach in studying the diverse types of living organisms, the problem of the systematic relationship be- tween such groups is of prime importance. The term animal relationship can be used to express any association between animals, but in its common usage it denotes system- atic or genetic relationship. Although the principles on which animal relationships are based are by no means agreed upon by all taxonomists and morphologists, those out- lined by Boyden (1942) seem to be the logi- cal choice as a foundation for any particular attack on the problem of classification. Sys- tematic relationships based upon conserva- tive hereditary traits involve genetic rela- tionship. To distinguish non-essential from essential criteria on which the grouping is based, should be the prime object. Those characters which most faithfully indicate a common heritage, will best serve as guides to the study of animal relationships. Characters are not equal in this respect, however, for although the vast array of structural characters is primarily due to heredity in their mature expression, they may be affected by environment. By such interaction they may or may not help to reflect the common ancestry. Besides the morphological, both chemical and physiologi- cal characters are also determined primarily by inheritance. Some of these may also be modified in their mature expression by en- vironmental influences. On the other hand, certain physiological characters, such as the blood groups, so far as is known, are de- termined so completely by inheritance as not to be affected by environment at all. On such a basis, it seems appropriate to use all kinds of conservative characters, morphological, biochemical and physiological, the choice of such characters depending on the relative constancy with which they indicate a com- mon inheritance. If it be true that the chemical nature of an organism, as well as its morphological expression, is determined primarily by in- heritance, then morphology alone is not a complete basis for determining the degree of animal relationship. The evidence indicates that chemical similarity of proteins of vari- ous animal species denotes the relationship of such species, the closer the similarity of such proteins, the closer the degree of gene- tic relationship between them, Boyden (1942) . The problem of classification is ex- tremely complicated, but with a better understanding from the newer methods em- ployed, it has much greater possibilities of achieving a truly natural classification than was formerly believed. In such a problem, then, we must be assured that the characters picked are conservative and any classifi- cation based on the nature of the organism should use all such characters and not con- fine itself to morphological expressions alone. As a physiological character not apprecia- bly affected by the environment to any ex- tent, the blood sera of various animal species seem to be highly promising. Moreover, the conservativeness of this character is assured by means of the antibody mechanism exist- ing in the organism, which apparently com- pels serum proteins to change but slowly. Since the discovery of the precipitin reac- tion by Kraus (1897) and the statements of Wells (1929) and Landsteiner (1936) that the chemical similarity of proteins can be demonstrated, many studies of the relation- ships of various animal species have been carried out using this method. Besides com- plementing in some cases the earlier findings of morphology, in others it has helped solve problems in classification that could not be worked out from the morphological evidence at hand, as shown in the work of Boyden and Noble (1933). Serological studies then can be an aid in helping to determine the degree of animal relationships. ( b.) On the relationships between Cyclo- stomata, Elasmobranchii and Pisces. One very notable result in the attempt to demonstrate relationship between the above classes of Craniata by means of the sero- logical method, was the marked inability of the antisera employed to show any cross re- actions between these groups, even with the ring test. When antishark serum, produced by the triple injection method, was reacted with the lamprey ( Petromyzon marinus ) and the sturgeon ( Acipenser rubicundus ), no reaction whatsoever resulted (Table I). The inability to get any reaction between Ring' and Photron’er Test Results of Elasmobranchs. 1943] Gemeroy: Relationship of Some Common Fishes 113 M 55 55 < (o^j) uo9S.m;g o © o o (N) © © (OjI) jWh aiSeg © © © © (n) -era © © (oj) apsjis -era 6.25 0.0 © © (N) fisySoa 00 c> © o (OJ) Ms¥Soa 12.5 0.0 OJ 00 (N) umiifAos LO rH © rH (<\l) mnHI^3S 25 12.5 00 rH (N) uoi.idodAjj o to m 56 tests (oj) uoudodAjj (N) 18 50 50 Ring Tes 77.9 90.9 lotron’er I fa-t Oh (oj) 18 q.mqs 50 100 100 <© (N) 68 q-^qs 100 100 o o rH (oa) 68 q-ieqs 100 100 25 O o rH (ojj) Aa.iduieq ° °. o © © O suoiiijm ui ■iapx sno3oiouiojj 1.024 .256 1.024 o fa w o fa CO m fa fa fa OS c3 fa fa m m m fa fa ^ ^ ^ (oq) d.lBQ qqoq © d d d d d (N) lnal£ uM.o.ig (N) ino.tx A\oquiBg (og) qsySui^t (oq) p.iaqoig (og) yo.iaj (og) qsiq 'S'0'a (og) ssBg qaog (oq) sssg -jai q (oq) ssBg qoxqg (og) uiydioa o owoo dodo 03 03 O O O © HHOO dodo cd cd 2 2 10 lo 10 © © © 03 03 03 © © o d cd d zo 10 ID to O 03 to 03 IO * Ol ZD 10 tooiooi LO OOO 0003 10 odd 40 10 cd 03 00 03 OOOOC>OOtHLO dodo ’SScdoi O O O O o rH ( M) ino.ix qoo.ig o d o o o o (°jj) BpnDUJjng OOOOOOi® z H o 5 Z < (oq) ^no.ij, qoo.ig (N) uod.iuj, (og) UOd.IBJ, OOOO OOOO odd© dd2w OOOO dodo in Z w , o O O O O ' H Z < > o : © (Oji) aSunn^snirti (<>£) aqij (oq) ysy^o LDOOOO OOO 03 LO O O n r-^ r-i « o o o o o o o O O O LO ^ 03 05 rH CO O O O o cd 'dodo (N) uyMog (oq) uyAvog (oq) aqia-.i^o (o^) uoa&xruig suoiiUM ui mu ’JOuioh © © oooooo 03 LO © © © iH o 03* o o o cd o © © OOC0000300Tf03 03 © 03 03© TjfQrfHTfOlCOHLOHHLO ooowqqotqwiotcoj 03 Tji 03 03* rH 00 w s c gHSH“ - ^ ' OOO “£?S)tg'S ap-o o aaa jjijioocjcj^^csosoi MWMfqWHP'WWOOO I I l I I I I I I ■ I I *42 ‘42 42 42 42 42 ’-+2 *42 *42 *42 ‘42 (og) djBQ (oq) ino.ij, qooug (oq) uodJBj, (OjI) uyAvog o cq © © © © o* o* © © © © ZD LO OO HOO © d o o cjirniTTiTAT TIT 03 O O O O Oi Tf N (N suoitHH UI HimotoiOH whh '[OUIOJJ lO 03 03 03 03 LO o ^o 30 °r° o O .fz< ' lld § O CO O © © (ojl) o old do (N) as.ioqpoy CO © ho © ©’ oi (N) qsy oimyny °o © © © ©* oi (N) <1.ibq © © © © ooSS (oh) d.iBO . 0.0 0.0 0.0 0.0 100 (N) ino.xx UAvo.iy 60.3 53.2 (N) ino.ix AvoquiBH 00 © © © (Od) lajaqaid © © © © OTtHH (N) ino-CL qoo.iy 100 100 0.0 (o^) ip.iay © © tM © o*tj5nn tH C<8 (oh) ino.ix qoo.ia 0.0 0.0 0.0 78.6 (o^j) ssua qaoy 13.1 30.7 4.0 (N) uod.mx © © © © (°H) qsiy -s'0-a 41.2 56.8 (o^) uod.inx 0.0 0.0 0.0 100 A.NTIGENS («H) ssua -H 'rl 45.1 74.6 11.5 (N) uy^og 100 88.6 0.0 (o^) ssua qoB[ a o © o ■OOOOJ ° ’-H ^ co d (ojl) uyAvoy CO © © 00 © © o o © co 2 °* (oy) BpnDB.i.iBa © © © © © © © © © 2 (OH) aqid .mo co © © © © © © tM © (oy) a.3un[[Bsnj\[ rH © © © HCOHH (oji) uooSrn^g o © © © © © © © © o HHHUU («H) aqiy © © H CO © 22©"©© ^ H (M (oh) Any ai3na 0.0 0.0 (oh) qsyiBQ © © © © © © © © (oh) o^nqg ‘Q-a © © (oj) d.ino © © (oh) (68) >i«qs § © © © © © © O o r\ co rrw H £ £ C m o o o H 0) 0) 0) fi fi H ojO bD bjOc©^ qzj ^ ^ ^ 3 p 2 o o figs*** t-H rv. £ HH MH MH ^ O -J_» -I-J HH ^ O O M ^ ?H ^ H fn CIS C3 +J+J+J+J+J+J4J4J+J-P+J -°fe. < o 05 W w Eh ^ £ 0) P!cI (OjI) yo.iaa o © o © (oji) ssBq^DBia © © © © © © (<\i) uiqd[OQ o © © © (°,l) BpnoB.i.iBa © © © © (o^) aSunpB^snj^ lO 03 © d d (<\a) LO 03 © rH © fa-? h— ' 03 (OjI) qsyiBQ Bag LO O 03 tH LO rH rH (oa) WP30 © l^sclox iPR) 2 03 LO tH (%i) qsypjo O O O O o o o o rH rH T—' rH 03 (oj) d.iBO rH © ® ® -s O 01 co m m H (OjI) ino.ij, .qoo.ig o> 03 Lh tH o ^ © CO O &J0 .£ ^ 3 U (0^) UOd.IBJ, © o r£ © Oh LO (oa) uyAvoa 03 © d © 7.7 0.0 (oa) 9Rtd[ .HR) LO 03 © d © 1 (o^) UOdSjn^g LO 03 © © d © d (oa) Abh 9{Sb3 © J9'RT, snoSo|ouiou © oi fe fa fa < w 00 W HH *w- 1 | CO HI w HH in X X « M © tC r£ r£ W Ifl c£ c£ H z < "eS "cS U O 1 1 "S "5 o o 4-> *-H £ £ < < -H "-H £ £ < < findings in most cases parallei those estab- lished by the methods of morphology. This is best shown with the orders and families within the Teleostei. Cross reactions (pho- tron’er) were found when antibrook trout ( Salvelinus fontinalis ) serum was tested with rainbow trout ( Salmo irideus ) and the brown trout ( Salmo fario). No cross reac- tion could be shown with antibrook trout serum against carp, bowfin, catfish and tarpon. The pike ( Esox lucius) and the muskallunge ( Esox masquinongy') gave good cross reactions with this instrument, but very little reaction was demonstrated even with the ring test with 7 and 8 other species respectively, as shown in Table II. In the group Percoidea of the family Acanthopteri, the antiblack bass (Microp- terus dolomieu) and the antipickerel ( Stizo - stedion vitreum) sera gave excellent cross reactions with closely related species (Table III). With the ring test, as would be ex- pected, slight cross reactions were shown between species somewhat more distantly related (Table II). These results as determined by the sero- logical method, could no doubt, be enlarged when more extensive studies are carried out with other species shown to be closely related by morphological methods. (e). Results which do not parallel the present classification. As just stated, the majority of the tests carried out gave relationships that paral- leled the systematic position as determined by morphological methods. A small number of apparent discrepancies was observed and further studies will have to be carried out before one can be sure that the relationships shown are valid. The cross reaction between the catfish ( Ameiurus lacustris ) and the bowfin indic- ates a 7.7% relationship which is surprising when compaied with a 2.0% for catfish-carp and 1.6% for catfish-trout. Carp and trout are placed much closer to the catfish in their systematic position. On this basis a relation- ship of less than 1.6% was to be expected instead of the 7.7% obtained. Table IV. The antiserum used in these reactions was produced by a multiple series of injections. Such antisera have been shown to have a broader range and a decreased specificity (Wolfe, 1935). Cross reactions may be pro- duced when such antisera are tested with distantly related species that do not give a true indication of the relationships of these species. This is particularly important when the values obtained are small. With a catfish antiserum, produced by means of a series of single injections, no reaction was found with either bowfin, carp or trout. One single series bowfin antiserum, however, in the reciprocal reaction when cross reacted with the catfish, gave a relationship of 2.3%, 1943] Gemeroy: Relationship of Some Common Fishes 117 Text-pig. 1. The comparative values when three different native carp antigens were tested against the same antinative carp serum. Relationships are determined by comparison of total areas of turbidity curves. Although the curves are not identical, the areas of these curves vary less than 2 per cent. Antiserum Anticarp 2 X Antigen Curve Per Cent. No. Area Carp 1 (N) 1 99.5 Carp 2 (N) 3 100.0 Carp 3 (N) 2 98.3 while another showed no reaction with either catfish, carp or trout. The same mul- tiple series catfish antiserum gave a slightly stronger reaction with the pike than it did with two species of marine catfish. Another result showing an apparent dis- crepancy with the present systematic posi- tion, was the tarpon ( Tarpon atlanticus) classed as one of the most primitive of the Isopondyli. If any cross reactions with the ganoid antisera could be obtained, the tarpon, as one of the most primitive species of the Teleostei, should show such reaction. No reaction occurred between antibowfin serum and tarpon antigen, nor did antitarp- on serum react with bowfin antigen. Table II. These few apparent exceptions to the general parallelism, require confirmation but do not seriously detract from the signifi- cance of the results. (f). Possible sources of error in sero- logical studies. Careful consideration should be given to sources of error that may have a bearing on the analysis of any experimental results ob- tained. In serological investigations, the effects of lipoids on cross reactions have been studied ( Landsteiner, 1936). It has been shown that by the addition of lipoids, cross reactions can be greatly increased. Al- though lipoids may not be capable of acting as antigens in antibody production, they can combine with serum proteins and act as haptenes and thus decrease the specificity of the reaction. When the ring test is used in the precip- itin reaction for the determination of animal relationships, comparable amounts of anti- gen and antibody must be used, as only the end point of the reaction is recorded in such 118 Zoologica: New York Zoological Society [XXVIII: 15 1 ANTIGEN DILUTION Text-fig. 2. The relationship between different species of Elasmobranchii. This antishark serum was produced by triple injection series of a formolized antigen and only by such means could a reaction be shown with the dogfish. Formolized lamprey, bam door skate and sturgeon did not react with this antiserum. Curve PerCent. Antiserum Antigen No. Area Antishark (S3) (Fo) X Shark I (Fo) 1 100.0 Shark II (Fo) 2 94.1 Yellow shark (Fo) 3 77.9 Scyllium (Fo) 4 18.9 Dogfish (Fo) 5 8.2 determinations. With the photron’er, where the complete range of reaction is repre- sented by the area under the curve, no such procedure is necessary. Where sufficient amounts of antigen and antibody are pres- ent, the reaction range is always complete. With constant antibody, the curve area re- mains the same, for as the amount of the antigen varies, the position only of the curve on the abscissa is changed. As a rule serum proteins are not' homo- geneous substances and it is of prime im- portance to determine, if possible, just what fractions of the serum proteins are most active in antibody production and titration. The amounts and proportions of these pro- teins may vary in different species and such variation directly affects the antibody pro- duced, which in turn affects the reaction when various antigens are reacted with antiserum. It is especially true that in the ring test, errors may result from compari- son of the sera of different species, equival- ent in total protein, but different in their content of active protein antigen. On the other hand, in the photron’er comparison with the same sera, no similar errors could occur if complete curves are obtained. In this respect, DeFalco (1940) has made an excellent beginning with the blood serum of birds. He found a wide difference in the albumin globulin ratios of certain birds and 1943] Gem.eroy: Relationship of Some Common Fishes 119 ANTIGEN DILUTION Text-fig. 3. The relationships between the Elasmobranchii with an antishark serum made by a single injection series. The antigen used for anti-serum production was not formolized as with the previous one. Text-figures 2 and 3 show the parallel results obtained with the same species, Text-figure 2 involving only formolized reagents, Text-figure 3, involving only native reagents. Antiserum Antishark (N2) X Curve Per Cent. Antigen No. Area Shark I (N) 1 100.0 Shark II (N) 2,4 90.9 Yellow Shark (N) 3 56.0 Scyllium (N) 5 12.9 demonstrated that the globulin fraction of blood serum is by far the most active of the proteins in antibody production and titra- tion. Further investigation may show that I, it is necessary to compare only specific frac- tions of blood serum and not antigenic mixtures, such as the sera of birds and mammals. It does not necessarily follow that the effects found with birds can' be applied to other classes of Craniata, the sera of which may be more nearly constant in their albumin globulin ratios. If reactions are produced with very dis- tant relatives when a multiple series anti- serum with decreased specificity is employed, they should be checked before being ac- cepted. This can be accomplished by dilution of the antiserum. By this means only related species show any cross reactions, thereby eliminating any that may seem doubtful. (g). General conclusions and problems for future study. The data presented in this paper seem to justify the conclusion that the precipitin reaction can be of value as a check on the morphological findings in the study of animal relationships. In most cases these findings parallel the systematic position of the several fishes, especially where closely related species are tested. In other cases they show apparent divergence from the usual systematic arrangement. An examina- tion of some of the more recent classifica- tions of fish, constructed entirely from the morphological approach, reveal the wide range in ideas held by taxonomists in gen- eral as to what principles should be em- ployed as a foundation for determining the phylogenetic relationship of this rather 120 Zoologica: New York Zoological Society [XXVIII: 15 ANTIGEN DILUTION Text-pig. 4. The reaction of antiformolized catfish serum produced by triple series method. The reactions with the formolized antigens indicated here are quite unexpected as enlarged upon in the discussion. Also it may be noted that pike antigen gave a stronger reaction than did either of the marine catfish, and further, the pike cross reaction was stronger than with the carp which is classed in the same order as the catfish. No reactions were produced with formolized barracuda, black bass or sturgeon antigens. Curve Per Cent. Antiserum Antigen No. Area Anticatfish (S3) (Fo) X Catfish (Fo) 1 100.0 Pike (Fo) 2 20.2 Gaff Topsail (Fo) 3 14.7 Sea Catfish (Fo) 4 15.0 Bowfin (Fo) 5 7.7 Carp (Fo) 6 2.0 Brook Trout (Fo) 7 1.6 highly diversified group. Garstang (1931) states that recent classifications can hardly be taken as an expression of their phylo- geny. In any case, a truly natural classifica- tion must take into account all the evidence discernable which bears upon the phylogeny. Only by so doing can a classification with probably correct genetic implications result. Because the serological and the morpho- logical methods do not always agree it does not necessarily mean that the one is entirely right and the other entirely wrong. No phylogenetic method is final, be it morpho- logical or serological. In the absence of cer- tain knowledge of phylogeny we are forced to accept the more probable interpretations of animal relationship and these more prob- able interpretations must be based on more than one kind of evidence. In a real sense There is, therefore, a need for further intensive studies among widely separated groups of fish to determine by serological methods what relationships can be shown between orders, families and species. The possible sources of error in the serological method as enumerated in (f) should be thoroughly investigated and their effects noted, so that correct conclusions may be drawn. With more nearly complete knowl- edge at our disposal, it may be possible to 1943] Gemeroy: Relationship of Some Common Fishes 121 ANTIGEN DILUTION Text-fig. 5. The reaction values between representative species of the group Percoidea. As with the trout species tested the closely related species here are readily differentiated by the photron’er. The representatives of the family Percidae, the perch and the pickerel, seemingly are not closely related to the family Centrarchidae, even though placed in the same group. The black and large mouthed bass, two closely related species, are easily distinguished by the photron’er. Curve Per Cent. Antiserum Antigen No. Area Antiblack Bass II (Fo) X Black Bass (Fo) 1 100.0 L. M. Bass (Fo) 2 74.6 B. G. Sunfish (Fo) 3 56.8 Rock Bass (Fo) 4 30.7 Perch (Fo) 5 4.9 Pickerel (Fo) 6 4.6 understand the real significance of the sero- logical methods, especially the precipitin reaction in the determination of the relation- ships between the different species of fish. It may be that a quantitative basis for the determination of phylogenetic relation- ships can be established by the precipitin reaction. Such a serological “yardstick,” besides its important application in classi- fication, might be of great value in the field of animal breeding and thus mark out new lines of progress in practical genetics. Summary. 1. The precipitin test has been applied to the comparison of the blood sera of 31 species of fresh and salt water” fishes. In all 43 antigens and 26 antisera were used in these tests, which were of two types, the ring test and the photron’er test. 2. Formolized antisera when tested with formolized antigens give parallel values with native antisera tested with native antigens. 122 Zoologica: New York Zoological Society [XXVIII : 15 ANTIGENS ANTIGENS Text-fig. 6. The comparative values by the ring test of the Elasmobranchii tested against antiformolized and antinative shark sera. It will be observed that the barn door skate gave a slight reaction when tested against the antiformolized shark serum produced with the triple series method, while none was shown with an antiserum to a single series. Antishark (S3) (Fo) X A — Shark I ( Fo ) B— Shark II (Fo) C — Yellow Shark (Fo) D — Scyllium (Fo) E — Dogfish (Fo) F — B. D. Skate (Fo) 3. Cyclostomata, Elasmobranchii and Pis- ces, serologically are sufficiently far apart to be considered distinct classes. 4. Within the Elasmobranchii, antishark serum produced no cross reactions with the rays employed. Only one slight re- action was obtained with the skate. 5. In general the results parallel the taxo- nomic position based on morphology, but it is clearly evident that the chemical gulf which separates species and orders among fishes is far wider than that in birds. Bibliography. Boyden, Alan A. 1942. Systematic serology: A critical ap- preciation. Physiol. Zool.,Y ol. 15, No. 2. Antishark (N2) X A— Shark I (N) B — Shark II (N) C — Yellow Shark (N) D — Shark I (Fo) E — Scyllium (N) F — - Dogfish ( N ) Boyden, A. A. and Noble, G. K. 1933. The relationships of some common amphibians as determined by serolog- ical study. Amer. Museum Novitates, No. 606. DeFalco, R. J. 1940. A serological study of some avian relationships. Thesis presented to the Graduate Faculty of Rutgers Univers- ity. Published in part in Biol. Bull., 1942, 83: 205-218. Garstang, W. 1931. The phyletic classification of Teleostei Proc. Leeds Phil, and Lit. Soc. Vol, 2 (Sci. Sect.) Koch, F. C. and McMeekin, T. L. 1924. A new direct nesslerization micro- Kjeldahl method and a modification of the Nessler-Folin reagent for am- monia. J. Amer. Chem. Soc., 46, 2066. 1943] Gemeroy: Relationship of Some Common Fishes 123 Text-fig. 7. The reaction values by the ring test when the antiformolized pickerel and black bass antisera were tested with a number of antigens. It will be seen that the ring test brought in more distant relatives than could be shown with these same sera by the photron’er, but in the more closely related species of the Centrarchidae, the differentiation shown with the photron’er was not paralleled by the ring test. Antiblack Bass II (Fo) X A — Black Bass (Fo) B — L.M. Bass (Fo) C — B.G. Sunfish (Fo) D — Rock Bass (Fo) E — Perch (Fo) F — Pickerel (Fo) G — Dolphin (Fo) H — Brook Trout ( Fo ) I — Barracuda (Fo) J — Kingfish (Fo) Kraus, R. 1897. Uber specifische Reactionen in keim- freien Filtraten aus Cholera, Typhus und Pest bouillon Culturen erzeugt durch homologes serum. Wiener klin. Wochen., No. 32, 736. Landsteiner, K. 1936. The specificity of serological reactions. Charles G. Thomas Co., Baltimore, Maryland. Libby, R. L. 1938. The photronreflectometer, an instru- ment for the measurement of turbidity systems. Immunol., 34, 71. Antipickerel II (Fo) X A — Pickerel (Fo) B — Perch (Fo) C — B.G. Sunfish (Fo) D — L.M. Bass (Fo) E — Black Bass (Fo) F — Rock Bass (Fo) G — Barracuda ( Fo ) H — Dolphin ( Fo) Mez, C. and Ziegenspeck, H. 1926. Zur theorie der serodiagnostik. Bot. Arch., 13, 483. Nuttall, G. H. F. 1904. Blood immunity and blood relation- ships. Cambridge University Press. Wells, H. G. 1926. The chemical aspects of immunity. Chemical Catalogue Company, Balti- more, Maryland. Wolfe, H. R. 1935. The effects of injection methods on the species specificity of serum pre- cipitins. J. Immunol., 29, 1. Crane : Spiders from British Guiana and Venezuela 125 16. Spiders of the Families Lyssomanidae and Salticidae (Magoninae) from British Guiana and Venezuela. Jocelyn Crane Research Zoologist, Department of Tropical Research , New York Zoological Society. (Text-figures 1-4). [This contribution is a result of various ex- peditions of the Department of Tropical Re- search of the New York Zoological Society to British Guiana and to Venezuela, all made under the direction of Dr. William Beebe. The Guiana expeditions were made during the years 1917, 1919, 1920, 1921, 1922 and 1924. The Venezuelan trip, in 1942, was sponsored by grants from the Committee for Inter-Amer- ican Artistic and Intellectual Relations and from four trustees of the Zoological Society, George C. Clark, Childs Frick, Laurance S. Rockefeller and Herbert L. Satterlee, and by invaluable assistance from the Standard Oil Companies of New Jersey and Venezuela.] Contents. I. Introduction 125 II. Family Lyssomanidae 126 Lyssomanes nigropictus Peckham 126 Lyssomanes beebei sp. nov 126 Chinoscopus maculipes sp. nov 128 III. Family Salticidae 129 Subfamily Magoninae 129 Acragas carinatus sp. nov 129 Cobanus scintillans sp. nov 131 Hypaeus porcatus (Taczanowski) 132 Hypaeus flemingi sp. nov 132 Hypaeus duodentatus sp. nov 134 Encolpius fimbriatus sp. nov 135 Mago silvae sp. nov 136 IV. Bibliography 137 I. Introduction. This is the first of a series of papers on the spiders collected at Kartabo, Bartica Dis- trict, British Guiana, and Caripito, State of Monagas, Venezuela, by expeditions of the Department of Tropical Research of the New York Zoological Society under the di- rection of Dr. William Beebe. For maps and ecological data, see Beebe, 1925 and 1943. The present study is based on the lysso- manids and on the salticids of the subfamily Magoninae in the collection. The 84 speci- mens comprise 10 species: 3 species (34 in- dividuals) of Lyssomanidae, of which 25 1 Contribution No. 670, Department of Tropical Re- search, New York Zoological Society. individuals belong to one species, Lysso- manes nigropictus Peckham ; and 7 species (50 individuals) of Magoninae, of which 25 belong to the new species, Mago silvae, there being individually few of the others. Of the 10 species, 8 are regarded as new. Seven were taken in both British Guiana and Venezuela; 2 in British Guiana only and 1 in Venezuela only. In 7 of the species both males and females were taken, and in 4 species enough young, in addition to the adults, to form interesting growth series. Descriptions of colors in life, often so dif- ferent from those in preservative, habitat notes, and dates of capture of various growth stages are included in most of the species discussed. Further summaries, de- tailed development studies, and discussion of breeding seasons must await completion of the study of the remaining salticids. While these data form only a beginning of life history work in these little-known trop- ical salticids, they furnish an example of the many advantages to be gained by studying tropical animals in the field in one or two adjacent localities throughout a number of seasons. Only by such methods can knowl- edge of ecological and psychological char- acters (such as those of display), now recognized as of vital importance even from a purely taxonomic viewpoint, be added to morphological descriptions of species. And only by future comparative studies of all these factors can the evolution and relation- ships be understood of such an important and puzzling group as the salticids. In this paper, total length is measured from anterior margin of orbit to posterior tip of abdomen. For convenience, an arbi- trary division of young specimens into im- mature (“imm.”) and juvenile (“juv.”) stages is made, as follows : A specimen is re- garded as immature when the genital organs are well formed, but some obvious weakness of pigmentation, shortness of chelicerae or palp, etc., indicate a pre-adult stage. A 126 Zoologica: New York Zoological Society [XXVIII: 16 spider is listed as juvenile when there is no external development of the bulb on the male palpal tai’sus or no visible epigynum in the female. Petrunkevitch (1911, 1928, 1933) is fol- lowed in general classification. All types are deposited in the collections of the Department of Tropical Research of the New York Zoological Society, Bronx Park, New York City. My thanks go to Dr. William Beebe for the use of his field notes concerning British Guiana specimens, and to Mr. Henry Flem- ing, entomologist of the Venezuelan Expe- dition, for collecting the majority of the Venezuelan specimens. Where not otherwise specified, the field notes are my own. I wish to express here my appreciation to Dr. Alexander Petrunkevitch and Dr. W. J. Gertsch for their helpfulness with bib- liographic questions. Family Lyssomanidae. Lyssomanes nigropictus Peckham, 1888. Text-figs. 1A, B, C, D, E. References : Lyssomanes nigropictus Peckham, 1888, p. 235, pi. xi, fig. 10; Simon, 1901, p. 392, figs. 394, 395, 410, 411. Color in Life: Male (Venezuelan speci- men) : Integument of cephalothorax pale translucent yellow gi’een with median black stripe on thoracic part and another along each side of same, along margin. Anterior eyes translucent yellow green. A black stripe across clypeus, a black spot at base of each chelicera posteriorly, and black bars on in- ner and outer sides of femur of palpus. All legs paler than cephalothorax, each with black bands on distal end of femur, and on basal and distal ends of tibia; tarsi entirely black; some long black hairs on legs. Ab- domen pale translucent yellow green with four pairs of black, irregular-sized spots, those of the most posterior pair being more or less confluent. Preserved specimens agree with Peckham’s description : that is, all of the markings remain, but are faded to brown or grayish, while the general color is yellowish to white. Female (from a painting of a living Guiana specimen, and descriptions of two Venezuelan examples) : Like male, except that there are no markings whatever save for the usual black spots in which the dorsal eyes are set; the ocular area is ornamented with a few silvery green (in young) to yellowish (in adult) hairs. A few hairs on legs brownish. Abdomen distinctly richer, darker green than in male. Immature male (Venezuela) : Identical coloring with female, except that reddish- brown is mixed with black on the large tu- bercle holding the median lateral eyes. Remarks: There seems to be no question of the identity of the present material from British Guiana and Venezuela with the single “Amazonica” specimen of Peck- ham (from Simon’s collection), both de- scription and figure checking well. The few differences are as follows: the posterior ab- dominal spots are strongly fused in our spec- imens, instead of merely “connected by a brown band;” spots at the base of the cheli- cerae are not mentioned by Peckham, al- though distinct in all our well developed males; finally, Simon’s figure of the cheli- cera (1901, p. 394, fig. 411) shows only the 4 large teeth, not the additional 3 minute basal ones found in our specimens. In some of our examples these basal teeth are so small and close-set that they could be easily overlooked. Because of the relative abun- dance of this species, both males and fe- males, in each of the two localities collected, and the identity of non-sexual structures (chelicerae, eyes, etc.), I refer the females listed below to this species without hesita- tion. Measurements in mm.: Largest male, to- tal length 5.26; cephalothorax 2.46; abdo- men 2.8. Smallest male, total length 3.36. Largest female, total length 6.25; cephalo- thorax 2.25; abdomen 4.0. Smallest female, . total length 3.07. Range: Known from “Amazonica;” Kar- tabo, British Guiana; Caripito, Venezuela. Local Distribution: The Caripito speci- mens were all shaken off bushes and low trees on the edges of sunny jungle roads and along trails through scrub jungle. Material: A total of 25 specimens was taken around Kartabo, Bartica District, British Guiana, and Caripito, State of Mon- agas, Venezuela, distributed as follows: Kartabo: 1917: Sept 2 $, 3 2, inch juv. and imm. (Cat. Nos. 1713, 1716, 1733); 1920: June-Dee., 2 $, 6 2, inch imm. (Nos. 201619, 201620) : 1921: Jan.-April, 1 2 (No. 21309) ; 1922: April 16, 1 2 (No. 22216), Feb.-Oct., 1 $, 4 2, inch imm. (Nos. 221141, 221142). Caripito: 1942: Mar. 11, 1 $ (No. 4236); Mar. 22-30, 1 imm. 2, 1 juv. $ (Nos. 42125, 42126) ; April 1-15, 3 juv. 2 (No. 42438). Lyssomanes beebei sp. nov. Text-figs. IF, G, H, I, J, K. Color in Life: Adult male: Unknown. Adult female : “Translucent green, with ocular quadrangle pink” (Beebe). Young female: Cephalothorax translucent night green (Ridgway’s Color Key) except for a pair of longitudinal patches of yellow hairs, embracing dorsal eyes and confluent imme- diately behind anterior eyes; anterior eyes night green ; chelicerae, palps, legs and ab- domen clear yellow green except for brown fangs and black tarsal tips. Spiderlings, just hatched : “Cephalothorax pale green, abdo- men darker green.” (Beebe). 1943] Crane: Spiders from British Guiana and Venezuela Text-fig. 1. A, Lyssomanes nigropictus, $, left palpus, ventral view (drawn with basal part more elevated than in Peckham’s view) ; B, same, ectal view; C, same, chelicera, ventral view; D, same, abdomen, dorsal view; E, same, 2, epigynum (hairs omitted) ; F, Lyssomanes beebei, $ holotype, left palpus, ventral view,- G, same, ectal view; H, same, chelicera, ventral view; I, same, dorsal view; J, same, 2 paratype, epigynum (hairs omitted); K, same, $ holotype, first metatarsus and tarsus; L, Chinoscopus maculipes, $ holotype, left palpus, ventral view; M, same, ectal view; N, same, cephalothorax and abdomen, dorsal view; O, same, 2 paratype, epigynum (hairs omitted). 128 Zoologica: New York Zoological Society [XXVIII: 16 Color in Alcohol: Both sexes, as usual, have lost all trace of green, the cephalo- thorax being yellowish, the abdomen almost white. Dorsal eyes set in two pairs of black spots. Hairs of ocular quadrangle in male scanty, silvery white, in female bright orange, extending down beyond outer mar- gin of anterior eyes, mingled with silvery white ones around and between anterior eyes; in young females the orange hairs are pale yellow. A band of silvery white hairs across clypeus in both sexes. Fang and spines horn-colored. Abdomen of male with a grayish-black marginal band, of irregu- lar width, surrounding it except at extreme posterior tip. No abdominal markings on female. Structure: Male: Cephalic and thoracic parts of cephalothorax about equally long. Eyes of first row twice as large as those of second. Clypeus about two-fifths diameter of anterior eyes. Chelicerae elongate, diver- gent; 4 teeth on upper margin (2 close to- gether, a third less than half way to tip, and 1 distal) ; 9 teeth on lower margin (a series of 6, almost equally spaced, increas- ing in size distally, plus a cluster of 3 un- equal ones distally) ; a group of large, spiny bristles on outer distal end; no teeth on fang. Legs I, II, IV, III. Metatarsus I 4% times as long as tarsus; fringes on each side feebly developed, increasing distally. Tibias I and II with 4 pairs of inferior spines and 1 pair of strong laterals, very slightly above and behind distal inferior spines. Metatarsi I and II with 3 pairs beneath. All femora with 1 spine above, near base, 3 beyond middle (1 above, 1 on each side) and 3 near tip (of which 1 or 2 are usually broken off). Palpus as figured. Female: Thoracic part of cephalothorax relatively longer than in male. Chelicerae well developed but much shorter than in male, scarcely diverging, without distal tooth on upper margin, and with only 7 teeth on the lower, the distal cluster being absent; the outer distal spiny bristles are only feebly represented. Metatarsus I only 3% times as long as tarsus; fringes scarce- ly weaker than in male. Spines exactly as in male. Epigynum as figured. Measurements in mm . : Male holotype, to- tal length 5.47 ; cephalothorax 2.29 ; abdo- men 3.18; chelicera (excl. fang) 2.46. Fe- male paratype, total length 6.0; cephalo- thorax 2.36; abdomen 3.64; chelicera (excl. fang) 1.0. Male paratype, imm., total length 5.03. Female, imm., total length 5.81. Breeding and Development: “April 19, 1924: Found female beneath leaf with 20 round, green eggs, deposited singly, close together on a small circular area. Egg .85 mm. in diameter. April 21 : Eggs all hatched, or at least ruptured, and gathered in a tiny, white, shrivelled mass at the meeting place of mouth, leg tips and abdomens of the young spiders. These are in a curious post- ova condition, far from able to move or function. Each large, rounded, green ab- domen is bent down, and the legs and palpi all ranged side by side, all centered beneath the cephalothorax. The eyes are very in- distinct, merely sketched in, and the whole cephalothorax is swollen, rounded and pale green. The spiderlings are 1.3 mm. long. April 22: The young spiders have unbent, their legs are free and spread out and the mother has devoured all but five of their egg skins, and has spun a little open work tent over the young, so fine that only by holding it against the light is it visible. When disturbed, the young ones scramble around weakly, and the mother creeps over them on guard, and will not leave even when I put the leaf under the microscope.” (Beebe) . In the young female, there are only 6 teeth on the lower margin of the chelicera and the epigynum is not developed; the leg spines and general proportions, however, appear quite sufficient for identification. The chelicerae of the young male are con- siderably shorter actually and relatively than those of the holotype, but differ in armature only in having the teeth more closely spaced (as in the female) and in having 5 and 7, not 9, large distal bristles on the upper side, on right and left sides respectively. Pigment and palp well devel- oped. The holotype is probably not quite mature, since it appears about to molt. Affinities: The proposed new species seems to be rather closely related to L. man- dibulatus Cambridge, 1900, from Central America, and to L. consimilis Banks, 1929, from Panama. It differs from both, however, in the lack of markings on the thoracic region, in the dentition of the chelicerae, in the spinulation of the legs, and in details of the palp. Range: Known from Kartabo, Bartic-a District, British Guiana, and Caripito, State of Monagas, Venezuela. Material: A total of 4 specimens (not counting just-hatched spiderlings) was taken as follows: Kartabo: 1920: Jan. -April, $ holotype (Cat. No. 201621), $ paratype, imm. (No. 201622) ; 1924: April 19, $ para- type with spiderlings (No. 24,422). Cari- pito: 1942: Mar. 18, 1 imm. $ (No. 4275), shaken from low tree in high jungle. This species is named in honor of Dr. William Beebe, director of the expeditions. Chinoscopus maculipes sp. nov. Text-figs. 1L, M, N, O. Color in Life : Adult male holotype : In- tegument of cephalothorax and abdomen en- tirely translucent lettuce green, except for 1943] Crane: Spiders from British Guiana and Venezuela 129 a narrow brown stripe extending completely around margins of cephalothorax, including clypeus, and a similar one around abdomen. Dorsal eyes each set in a black base, and these in a single pair of patches of shining white hairs. Palpi and labium reddish- brown. Anterior eyes translucent reddish- brown. Legs translucent and colorless except for purplish-brown bands as follows: on distal end of femur, basal and distal ends of tibia, basal and distal of metatarsus and most of tarsus. In addition, the entire fe- mur and tibia of the front legs especially and second legs somewhat appear bright purple in certain lights ; it seems the thread of pigment they contain must be magnified by the curve of the leg, like mercury in a thermometer. Spinnerets reddish-brown. Juvenile male: Differs from adult in be- ing a much yellower green, in lacking the ocular patches of white hairs, and the brown marginal stripes, and in having pigment on the legs only near base and tip of each tibia and base of metatarsus; palpi and spinnerets colorless. Color in Alcohol : In both adult and young described above, as well as in our two re- maining male specimens, the integument of cephalothorax and abdomen is yellowish or pale brownish ( not dark brown or black as in other males of the genus except brasilien- sis ), and whitish in the young. The white cephalic hairs are largely missing, and have turned yellowish, while the brown stripe is variable in strength, not depending wholly on development. Eyes and spinnerets faded, colorless. The banding of the legs remains, somewhat stronger even than in life, the pig- ment of femora and tibiae I and II being much stronger in the adult males, darkening the entire segments of the first and much of the second, instead of being confined to thin median threads of pigment. Adult females : Integument of cephalo- thorax yellowish-white, that of abdomen white. Black bases of dorsal eyes moderate- ly well clothed in yellow hairs. No marginal stripe and no pigment on palpi and spin- nerets. Legs white with spots of brown pigment at basal and distal ends of all tibiae, metatai'si and distal parts of tarsus. Fang and epigynum light brown. Labium unpigmented. Structure: Cephalic part of cephalotho- rax only half as long as thoracic part; ocular quadrangle twice as wide as long; about 5 or 6 very minute teeth on lower margin of chalicera. Palp and epigynum as figured. Measurements in mm. : Male holotype, to- tal length 5.91; carapace 2.04; abdomen 3.83; leg I: femur 4.62, patella .75, tibia 4.35, metatarsus 4.41, tarsus .86. Female paratype, total length 6.02; carapace 1.88; abdomen 4.14; leg I: femur 3.91, patella .65, tibia 3.48, metatarsus 3.59, tarsus .81. Juv. female, total length 3.38; two juv. males, total lengths 5.5, 6.72. Affinities: In generic characters the pro- posed new species is a typical Chinoscopus. It differs radically from all three forms, flavus (Peckham, 1888), gracilis (Taczan- owski, 1872), ernsti (Simon, 1900) and brasiliensis Mello-Leitao, 1917, described, however, as follows: The male is light in color in both living and preserved speci- mens and has a marginal body stripe; black spots or bands are present near the joints of at least the tibiae and metatarsi in all legs in both sexes; details of the palp are distinct; epigynum of the female is quite different from that of flavens, the other known female in the genus. Range: Known from Kartabo, Bartica District, British Guiana, and Caripito, State of Monagas, Venezuela. Local Distribution: two males, both adult and young, were shaken at different times from shrubs beside a sunny jungle road. Material: A total of 5 specimens was taken, as follows : Kartabo : 1924 : April, 1 $ paratype (Cat. No. 241010). Caripito: 1942: April 3, 1 8 holotype (No. 42161) ; Mar. 22-30, 1 imm. 8 (No. 42124) ; May 1-15, 1 imm. $ (No. 42439) ; May 15-30, 1 imm. 8 (No. 42440). The name maculipes is given this species in reference to the marking on its legs. Family Salticidae. Subfamily Magoninae. Acragas carinatus sp. nov. Text-figs. 2A, B, C, D, E, F. Color in Alcohol: Male: Cephalothorax light rufous, paler around margins. Dor- sal eyes set in two pairs of black spots; a crest of reddish-orange hairs mixed with black ones behind anterior eyes ; a pair of subquadrate patches of white hair between middle and posterior dorsal eyes ; a smaller, similar, median patch, often almost rubbed off, behind posterior eyes. All eyes rimmed with reddish-orange hairs; clypeus reddish- brown under eyes, blackish above margin, ornamented medially with a large oval spot of short white hairs. Chelae, mouthparts, tarsus of palp, distal part of all femora and all patellae, tibiae and basal parts of meta- tarsi chestnut brown. Rest of palp and re- maining parts of legs yellowish-white to pale horn. Abdomen above covered with hairs forming a central whitish stripe flanked by two olive stripes, the latter meet- ing anteriorly, where they are darkened and have a reticulated appearance. Anterior third of abdomen with two well separated pairs, in the olive region, of small, faint, dark, reticulated spots. Posterior third with 2 pairs of strong, black, subquadrate spots, 130 Zoologica: New York Zoological Society [XXVIII: 16 Text-fig. 2. A, Acragas carinatus, $ paratype, left palpus, ventral view; B, same, ectal view; C, same, chelicera, ventral view; D, same, dorsal view,- E, same, endite, ventral view (hairs omitted); F, same, 2 paratype, epigynum (hairs omitted); G, Cobanus scintillans, $ holotype, left palpus, ventral view; H, same, ectal view; I, same, 2 para- type, epigynum (hairs omitted) ; J, same, chelicera, ventral view; K, same, juvenile $; L, same, immature $ paratype,- M, same, adult $ holotype; N, Hypaeus porcatus, $, left palpus, ventral view; O, same, ectal view; P. «ame, chelicera, ventral view; Q, same, endite (hairs omitted). 1943] Crane: Spiders from British Guiana and Venezuela 131 and behind them a much fainter third pair. Abdomen below white marginally with en- tire central portion olive, sometimes faint- ly striped. Female: Differs from male as follows: The scanty crest is composed entirely of light yellowish hairs; white interocular, postocular and clypeal spots lacking; inter- ocular ones replaced by reddish hairs; ab- domen white, unmarked except for a pair of very small, faint spots anteriorly (rep- resenting the second faint pair of the male), and posteriorly by the two pairs of strong, black spots; no trace of the fainter third posterior pair of the male. Each of the strong spots is preceded by a small spot of white hairs, better developed before the more posterior pairs of spots. Structure : With the characteristics of the genus. Male chelicera externally with a strong carina ending in a pronounced tooth; upper margin with a large black tooth, com- pressed. with an oblique inner margin but pointed tip; lower margin with up to 6 minute teeth (5 or less in the immature, and sometimes 6 on one side, 5 on the other) ; the basal one considerably the larg- est, decreasing in size distally. External distal part of lamina dilated and subangu- late. Palp as figured. Female differs from male as follows : chel- icera smaller, without keel or keel tooth, and with 2 small teeth on upper margin replacing enlarged one of the male. Outer angle of lamina less pronounced. Epigy- num as figured. Measurements in mm.: Male holotype, to- tal length 5.88 ; cephalothorax 2.63 ; abdo- men 3.25. Female paratype, total length 4.72 plus (probably immature; has been partially dried at one time) ; cephalothorax 2.29; abdomen 2.43. Male paratype, total length 2.63. Two young males, total lengths 4.61 and 5.18. Two juvenile males 4.41 and 3.94. Development: The two juvenile males, in addition to the undifferentiated palp, have the crest rudimentary and entirely red, have female-type chelicera, and lack prac- tically all abdominal pigment except the two pairs of strong black posterior spots, which are as dense as in adults. In one slightly immature male (4.6 mm.), found with the slightly immature female, the dark abdomi- nal stripes are much darker than in any other specimens, almost obliterating the 4 posterior spots. In the other immature male (5.18 mm.) the olive stripes are scarcely developed. Affinities: The proposed new species is closely related to A. leucaspis Simon, 1900, and to A. longimanus Simon, 1900. Accord- ing to the descriptions, these three alone have both a cephalic crest and a white cly- peal spot. The present species differs from the others in the fact that the crest is chiefly red, not entirely black, and in the markings of the abdomen (although all three species share at least two pairs of strong black spots). In addition, it differs from leucaspis in the strongly carinated chelicera ending in a distinct tooth, and from longimanus in having a maximum of 6, not 5, teeth on the lower margin, and in lacking white spots on the legs. ' Range : Known from Kartabo. Bartica District, British Guiana, and Caripito, State of Monagas, Venezuela. Material: A total of 7 specimens was taken as follows: Kartabo: 1922: Feb. -Oct., 1 $ holotype (Cat. No. 221143) ; 1917: Sept., 1 2 paratype (No. 1710), 1 imm. $ (No. 1747); 1920; July-Dee., 1 imm. $, 2 juv. $ (No. 201623). Caripito: 1942: June 1-15, 1 $ paratype (No. 42441). The name carinatus is given this species in reference to the well developed keel on the male chelicera. Cobanus scintillans sp. nov. Text-figs. 2G, H, I, J, K, L, M. Color in Life: Adult male (from holo- type) : Ocular quadrangle and entire upper portion of abdomen covered with scales of brilliant iridescent metallic green shifting to purple and rosy. Lateral and thoracic parts of cephalothorax as well as clypeus with bright iridescent plum scales. Integ- ument of cephalothorax wood brown, of ab- domen paler; under parts not iridescent, sternum brown, abdomen grayish-black. Each of the four anterior eyes surrounded with narrow margin of bright yellow-orange hairs. Legs all unbanded, uniformly pale translucent horn except for anterior pair and femur of second pair, which are cov- ered with scales of bright iridescent plum. Immature male (paratype) : Differs from adult in having iridescence on front legs barely developed. Juvenile male: Ceph- alothorax white except for ocular quad- rangle which is brown overlaid with green iridescence shifting to rosy; a median lon- gitudinal white stripe dividing it, which is bounded by an orange russet border run- ning through and slightly below lateral eyes. Abdomen above brown with two lon- gitudinal rows, one one each side of mid- line, of small white spots, and outside these on each side is a dark stripe. Lateral to these is a pair of white stripes. Ventral side of abdomen brown. Its entire dorsal surface is overlaid lightly with the same green-rosy iridescent scales as are on the ocular quadrangle. Chelicerae, palpi and legs translucent greenish-white; faint iri- descence on joints of legs. Large eyes light brown shifting to black, rimmed with white 132 Zoologica: New York Zoological Society [XXVIII: 16 hairs, except dorsally where they are or- angish. Adult female: Not seen in life. Imma- ture female: exactly like immature male just described. Juvenile female: Cephalo- thorax huffy yellow; legs greenish; anterior eyes chestnut, rimmed narrowly with white hairs. Abdomen above deeper buff than eephalothorax, sprinkled irregularly with hairs of whitish iridescence and brownish- buff. Color in Alcohol : Males, females and young: Iridescence almost vanished, remain- ing chiefly around dorsal eyes and, very faintly, on abdomen. Color otherwise as in life, except as follows : Cephalothorax in adults of both sexes reddish-brown, paler in female. Females of all ages with a few rusty hairs around and below dorsal eyes; large eyes framed in white. First legs of male as dark as cephalothorax, with con- spicuous black hairs on lower margin of tibia, metatarsus and tarsus; second legs slightly paler; rest pale horn as in life. Legs of female all buffy white. Abdomen of male brownish, unmarked, both above and below; of female buffy white with markings as follows: A pair of longitudinal brown bands of varying width extending through- out most of length, enclosing a median series of 5 or 6 white chevrons; in addi- tion there is a pair of lateral brown stripes; underside with a large central brownish area surrounded by white. In the two young- est males, in preservative, faint subdermal traces of one or two pairs of very faint, dark, blotchy abdominal markings, corre- sponding to the dark dorsal stripes of the female, are distinguishable. In the youngest females, the cephalothorax is pale brownish, abdomen white, without markings, except subdermal traces as in the males. Structure ami Affinities : The proposed new species appears to be closely related to Cobanus unicolor Cambridge, 1900, from which it differs in the somewhat less ex- treme development of the chelicera, lacking a recurved tip and having the 2 pairs of teeth somewhat closer together, even in the largest specimens; also, there is a tooth on the tibial spine of the palp below the hooked tip, ; similar but more strongly developed pectination is found in C. erythrocas Cham- berlin & Ivie, 1936. No iridescence is men- tioned in the description of unicolor, but the scales are so easily dislodged, and their brilliance so weakened by preserving liquid that they could easily have passed unno- ticed in unicolor. In fact, each of these three characters may prove to be of no separative value, but since the present specimens are the first of the genus taken in South Amer- ica, unicolor being known only from the Costa Rican holotype, it seems wise to refer them to a new species. The Guiana and Venezuela series are without question iden- tical. Measurement in mm.: Male (holotype), total length 6.05; cephalothorax 2.69; ab- domen 3.36; largest female (paratype), total length 5.69; cephalothorax 2.59; abdomen 3.1; immature male (paratype), total length 6.43; immature female (paratype), 5.47; smallest male, 5.38; smallest female, 4.0. Development: The difference in color be- tween adults and young has already been noted. Text-figs. 2J-M inclusive, shows a growth series indicating the development of the chelicerae, which lag behind the palpi and iridescence in development. Note how similar are the chelicerae of the adult fe- male and juvenile male (Text-figs. 2J, 2K), and the fissidentate character of the teeth. Range: Known from Kartabo, Bartica District. British Guiana, and Caripito, State of Monagas, Venezuela. Local Distribution: Three Caripito speci- mens were shaken off shrubs alongside open jungle roads and from trees in low jungle. Material : A total of 12 specimens was taken as follows: Kartabo: 1921: Jan.-April, 1 $ (Cat. No. 21311), 1 imm. $ (No. 21310); 1922: Feb. -Nov., 1 juv. $ (No. 221144). Caripito: 1942: Mar. 10, 1 $ (holo- type) (No. 4232) ; Mar. 24, 1 imm. $ (para- type) (No. 42442) ; April 1-15, 2 2 (para- types) (No. 42443); Mar. 17, 1 $, 1 2, both imm. (No. 4267) ; April 15-30, 1 imm. $ (No. 42444) ; Aug. 1-15, 1 juv. $ (No. 42445) ; Aug. 15, 1 juv. 9 (No. 42382). Hypaeus porcatus (Taczanowski). Text-figs. 2N, O, P, Q. References: Attus porcatus Taczanowski, 1871, p. 53, pi. iv, fig. 5. Hypceus porcatus, Simon, 1900, p. 44 [ 1901, p. 419, figs. 465, 466. Remarks : A single, slightly immature male was taken at Kartabo, Bartica District, British Guiana, between February and Oc- tober, 1922, total length 5.95 mm. (Cat. No. 221145). Its youth is shown in the length of the palps, slightly shorter than in Taczan- owski’s figure, and in the pigmentation of the abdomen, the spots being less distinct than indicated in Simon’s description. Oth- erwise, the specimen agrees perfectly with the latter, except that there are 5, not 4 teeth on the right chelicera, but the typical 4 are found on the left side. The present specimen extends the known range from French Guiana to British Gui- ana. Hypaeus fiemingi sp. nov. Text-figs. 3A, B, C, D, E. Color in Life: Male: Integument of ceph- alothorax light brown ; ocular quadrangle dark brown, with a pair of patches of white 1943] Crane: Spiders from British Guiana and Venezuela 133 Text-fig. 3. A, Hypaeus flemingi, $ holotype, chelicera, ventral view; B, same, dorsal view; C, same, endite, ventral view (hairs omitted) ; D, same, left palpus, ventral view; E, same, ectal view; F, Hypaeus duodentatus, $ holotype, endite, ventral view (hairs omitted) ; G, same, chelicera, ventral view,- H, same, left palpus, ventral view; I, same, ectal view; J, Encolpius fimbriatus, $ holotype, right chelicera, ventral view; K, same, dorsal view; L, same, 2 paratype, epigynum (hairs omitted) ; M, same, $ holotype, cephalothorax, lateral view; N, same, left palpus, ventral view. 134 Zoologicci: New York Zoological Society [XXVIII: 16 hairs between anterior and posterior dorsal eyes, which are as usual set in black patches. A crest of stiff, short russet and black hairs across ocular quadrangle immediately be- hind anterior eyes. A prominent patch ot white hairs occupying most of clypeus be- tween eyes and c-helicerae. Chelicerae and mouthparts chestnut brown. Integument of palp light horny except tibia and tarsus which are dark brown; femur covered with white hairs; two spots of white hairs in front of patella and tibia respectively. First legs entirely dark brown with white spots in front on tip of femur, base of patella and near base of tibia ; similar spots present on patella and tibia of second leg; metatarsus and tarsus lighter brown than rest, with a yellowish, translucent area in middle of each. Other legs similar but lighter, and with coxa, trochanter and basal three- fourths of femur translucent light horny; tarsi brownish-black. Abdomen above light olive green with a basal transverse band of dark green, and a broad central greenish- black stripe starting one-quarter of way to tip; this stripe is crossed by two transverse bars of same color, one at its origin, one less than half-way to its end; it broadens irregularly toward the tip, suggesting a third bar. Four pairs of small white spots on abdomen as follows : one spot in front and to outer side of each intersection of a cross bar with the median stripe, and the fourth pair behind. Underside of abdomen pale chartreuse with broad median dark green stripe joining at its base laterally with the broad basal transverse band of dorsal side a pale yellow band just before base of spinnerets. Color in Alcohol : So great has been the change in abdominal markings during the year since the specimen’s capture that there is no hint that it is the same spider: the green has completely vanished, as well as all shading, except a trace of the median longitudinal band and three pairs of small, faint spots, the first joined by a very faint cross bar, on the otherwise white abdomen. Underside white, except for two fine, me- dian, subcuticular lines. The cephalothorax and appendages are unchanged, except that in preservative the white spots on palpi and legs are not nearly as conspicuous as in life. Structure : Chelicerae long and slender, without a trace or even a marginal swelling on the outer margin, which is straight. Up- per margin of sulcus with 5 small teeth, lower with 3. Endite of pedipalp with outer corner produced, obtuse. Tibia of first leg with 3 pairs of ventral spines and 3 un- paired, anterior, lateral spines, of which the most distal unpaired spine is lower, longer and stronger than the other two. Palp as figured. Measurements in mm.: Total length 5.97 cephalothorax 2.9, abdomen 3.07. Affinities : H. flemingi appears to be most closely related to concinnus, nigrocomosus and cucullatus, all known only from Simon’s type descriptions (1900). In all of these a white clypeal spot is present, a tooth on the outer margin of the chelicera is lacking or reduced to a projection or convexity, and there are only 3 teeth on the lower margin of the sulcus. The present form differs from these in such details as the markings on the abdomen (in both living and preserved specimens), in the whiteness of the inter- ocular spots and, most important, in the de- scribed details of the palp. Range: Known only from Caripito, State of Monagas, Venezuela. Local Distribution: Collected by beating shrubs along a sunny jungle road. Material: Caripito: 1942: April 3, 1 $ (holotype) (Cat. No. 42155). This species is named in honor of Mr. Henry Fleming, entomologist of the Vene- zuelan Expedition, who collected the major- ity of the spiders. Hypaeus duodentatus sp. nov. Text-figs. 3F, G, H, I. Color in Life : Male : “Cephalothorax terra cotta, with black encircling band. Ocular quadrangle slightly deeper in color, the dor- sal eyes set in black, with patches of warm buff hairs between the anterior and poste- rior dorsal eyes. Chelicerae and mouthparts pompeian red. Palpi warm buff with tarsi brown. Legs warm buff and brown. Cephalo- thorax below warm buff. Abdomen above warm buff with black markings below, black with buff lateral markings.” (Beebe). Color in Alcohol: Markings of legs and upper side of abdomen practically identical with description of these parts in H. flem- ingi in life (p. 000), except that there are only 2, not 4 pairs of white abdominal spots visible (1 pair in front of second cross-bar, 1 pair in front of posterior expansion of median stripe) ; also, the second cross-bar and the posterior expansion each holds a pair of dark spots ; finally, the coloration of these abdominal markings is dull olive brown and buffy white, rather than dark and light green. The “black encircling band” of the cephalothorax mentioned in the field notes of the present species is now only present posteriorly as a narrow marginal line, and anteriorly on the face, on each side of the clypeal patch of white hairs. The rather poorly developed crest is composed chiefly of red hairs with a few black ones in both specimens taken; one of these ap- pears completely mature. Structure and affinities: This species is 1943] Crane: Spiders from British Guiana and Venezuela 135 very closely related to the preceding, H. flemingi, but differs as follows: the cheli- eerae are shorter and broader, with a slight- ly sinuous outer margin and with only 2, not 3, teeth on the lower margin of the sul- cus; also, the lower spine of the palpal tibia is produced and distally slender, not short and blunt. Measurements in mm.: Male holotype, to- tal length 5.59, cephalothorax 2.69, abdomen 2.9. Male paratype, total length 5.38 (imma- ture) . Range: Known only from Kartabo, Bar- tica District, British Guiana. Material: Kartabo: 1922: Feb.-Oct., 1 $ (holotype) (Cat. No. 221146). 1924: Mar. 23, 1 $ (paratype) (No. 241011). The species is named in reference to the two teeth on the lower margin of the cheli- cera. Encolpius fimbriatus sp. nov. Text-figs. 3J, K, L, M, N. Color in Alcohol: Male: Integument of cephalothorax reddish-brown (brightest in ocular quadrangle, except for a single pair of elongate black patches in which the dorsal eyes are inserted. A few russet orange hairs of varying length encircling all eyes; a very sparse, irregular sprinkling of minute yel- lowish white hairs scattered over cephalo- thorax in general, most numerous immedi- ately behind eyes and on sides of thoracic region; a narrow obsolescent band of white hairs immediately above the dark margins of thoracic region; clypeus with a slightly broader, very dense (but easily destroyed) band of white hairs across its entire mar- gin; clypeus otherwise naked and blackish. Mouthparts and sternum horn-colored. Chel- icerae almost black above, brown beneath. Palpi dark brown basally, otherwise pale horn except for brown bulb and spines; all the numerous, long hairs on dorsal and lat- eral sides pure white. Femur of anterior legs almost black, especially anteriorly ; pat- ella and tibia brown, each with a median spot of white hairs in front; metatarsus brown basally and distally, pale horn in the middle; tarsus horn-colored, tips slightly darker. All other legs with femur pale yel- lowish in basal half, brownish-black distally; the remaining segments very like the corre- sponding ones of first legs, except that patches of white hairs are rudimentary on the two posterior pairs. Integument of ab- domen above light brown with broken trans- verse dark markings, forming 3 irregular, speckled bands in anterior half, the most posterior being broadest, and 4 similar bands in front of spinnerets, but wavy and much narrower. Entire abdomen with a plentiful covering of short light hairs. Ven- tral side pale, with patches of dark pigment laterally, and a large fan-shaped patch around spinnerets. Female exactly similar to male, even in presence of long white hairs on palpi, ex- cept as follows: hairs around eyes are yel- lowish to whitish, not orange; anterior fe- mora scarcely darkened; other leg segments lighter than in male; abdomen with less pigment, the anterior bands breaking down altogether medially; underside of abdomen with a faint, dark, broad median stripe, and with fan-shaped posterior marking weaker than in male. Structure and affinities: The proposed new species agrees so well with Simon’s description of the genus and type species from Brazil (E. albobarbatus Simon, 1900, p. 59, and 1901, pp. 421, 427, and fig. 474) that it is possible that adequate series will show them to be identical. The only appar- ent differences in the males are as follows : In the present specimens there are 6, not 3 or 5, teeth on upper margin of chelicera, and 6, not 4 or 5, in a close-net series on lower margin, on the right side; on the left side, however, in both series the teeth tend to be weak and confluent, some in the up- per series being obsolescent; hence, in the genus the teeth number may be quite vari- able within the species. The remaining dif- ferences discernable from the descriptions and single figure consist in the longer tarsus in the palp of the present species, and per- haps a difference in abdominal markings, which Simon has not described in detail. A small, hooked spine on outer, lateral side of tip of palpal tibia in addition to the more anterior elongate process shown in the an- terior view figured ; bulb rather compressed, little projecting. In the female, the median dorsal eyes, al- though definitely in advance of the middle, are slightly less advanced than in the male. Chelicerae more slender, with 5 teeth on upper margin, 6 on lower, on both sides; no tooth near apex of outer upper margin, but this area is swollen with a distinct summit in the same region which in the male gives rise to the spine. There seems to be no rea- son to doubt that the two specimens belong to the same species, in spite of the fact that the male is from Kartabo, the smaller, prob- ably immature female, from Caripito. Measurements in mm. : Male, total length 3.97; carapace 2.09; abdomen 1.88. Female, total length 3.49, carapace 1.68, abdomen 1.81. Range: Known from Kartabo, Bartica District, British Guiana, and from Caripito, State of Monagas, Venezuela. Material: Kartabo: 1924: April, 1 $ (hol- otype; Cat. No. 241012). Caripito: 1942: April 16-30, 1 9 (paratype; No. 42446). The name fimbriatus is given this species in reference to the fringed palpi. [XXVIII: 16 136 Zoologica: New York Zoological Society Text-fig. 4. Mago silvae. A, left palpus of $, ventral view, juvenile (total length 5.57 mm.) ; B, same, immature (total length 5.1 mm.) ; C, same, slightly immature?, holotype (total length 5.89 mm.; D, same, adult, paratype (total length 6.43 mm.) ; E, same, ectal view; F, $ paratype, epigynum; G, ^ paratype, chelicera, ventral view; H, same, dorsal view. (All except G and H drawn to same scale; hairs omitted from A, B, C, F). Mago silvae sp. nov. Text-fig. 4. Color in Life : Adult male : Integument of cephalothorax reddish-brown to brownish- black, the ocular quadrangle and area im- mediately behind it brightest. Dorsal eyes set in black. Four patches of pale yellow hairs as follows: one on each side between median and posterior dorsal eyes: a small median spot between posterior eyes; a large median spot behind posterior eyes. In addi- tion, there are a few reddish hairs around all dorsal eyes. Large eyes chestnut brown rimmed with yellowish-brown hairs. Cheli- cerae black; mouthparts tipped with black. Palpi blackish-brown with moderately long, white hairs. First pair of legs considerably darker than others, with two patches of white hairs on anterior central parts of patella and tibia respectively; all legs band- ed dark and light, the dark ranging from vinaceous brown to dark chestnut, the pale from buffy pink to light horn, the hairs light or dark corresponding to color of underlying integument. Sternum light brown. Abdomen above tawny olive to blackish-brown, with paler markings, varying from pinkish-buff to white, and poorly to sharply defined. A characteristic general pattern is always dis- cernible, however, including two basal pale cross-bars separated by a dark bar and, be- hind these, six more or less coalescent and very irregular pale blotches; these are all often overlaid with uneven superficial pig- ment. Underside of abdomen almost entirely occupied by a dark triangle, vinaceous brown to blackish, traversed medially by a pair of narrow, longitudinal, pale lines, and sur- rounded laterally and posteriorly in front of spinnerets by gray. (From Beebe notes on Guiana holotype and Crane notes on Vene- zuelan paratype) . Adult female: Color in life unknown. Juvenile female : Cephalothorax primuline to ochraceous yellow, naked except for buffy 1943] Crane: Spiders from British Guiana and Venezuela 137 yellow hairs fringing all eyes. Large eyes ochraceous brown to reddish. Anterior two pairs of legs and their hairs ochraceous orange ; remaining legs somewhat paler than cephalothorax ; tarsal claws black. Abdomen above and below same shade of yellow as cephalothorax with three pairs of very dis- tinct, dark, dorsal spots and a longitudinal, dark, ventral band, narrower than in adult male. (From Beebe and Crane notes on Guiana and Venezuelan specimens, respec- tively) . Color in Alcohol : Little change, except for the usual fading. Adult female uniformly white except for three pairs of dorsal ab- dominal spots — larger, paler, and more ir- regular than in young female — and a dark, ventral, abdominal marking as in adult male. Dorsal abdominal markings in both sexes very variable in both intensity and shape, largely, but not altogether depending on developmental stage. Pattern of male practically identical, under outer epidermis, with that of female. In the male, the median pale hairy spot of ocular quadrangle is stronger in the young than in the adult, while that behind the posterior eyes devel- ops late and is strongest in the adult. Both are easily rubbed off in preserved specimens and may be almost lacking. The paired spots between the dorsal eyes are always much stronger. Structure ancl Affinities: Judging from Cambridge’s (1882) and Simon’s (1900, 1901) descriptions, the present species is a typical Macjo. The character of the cheli- ceral teeth alone, however, as well as other details, distinguish it from all of Simon’s species, while the tibial spines of the palp and the abdominal markings separate it at a glance from Cambridge’s intentus, in the description of which the dental formula is not given. In the present form there are 7 to 8 (usually 8) unequal teeth on the up- per margin, most of them large, except for the fourth, which is minute; 4 to 6 (usually 5) on lower margin; the full complement is not developed until late, well grown fe- males, apparently mature save for incom- plete epigyna and more compact abdominal spots, having as few as 2 teeth on lower margin. Unlike the male, the female has no spine on anterior side of first patella, and only one lateral spine (the more distal), on anterior side of first tibia. The development of the palp is shown in the figures. In only one specimen, the largest (Guiana paratype. No. 1718, shown in the shaded drawing) is the longitudinal tubule on the bulb apparent, or the small loop to the right distinct; also, in all other speci- mens, the external, smaller, tibial spine is more pointed than in this old specimen In cheliceral armature, body markings and all other details, however, this male is typical of the species. Measurements in mm. : Male holotype, to- tal length 5.89, cephalothorax 3.17, abdomen 2.72; largest male (paratype, No. 1718), to- tal length 6.43, cephalothorax 3.36, abdo- men 3.07; male paratype (No. 42207), total length 5.38 ; youngest male, total length 5.57. Largest female (paratype, No. 42447), total length 6.53, cephalothorax 3.36, abdomen 3.17; youngest female (No. 24,111), total length 5.76. Range : Known from Kartabo, Bartica District, British Guiana, and Caripito, State of Monagas, Venezuela. Local Distribution: 9 of the total of 25 specimens taken are known to have been beaten from bushes and low trees on the edge of and within jungle of moderate height. Material: Kartabo: 1917 (Sept.): 1 8 paratype (Cat No. 1718) ; 3 8 (Nos. 172, 1711, 1743) ; 1 2 paratype (No. 1748) ; 2 2, inch 1 imm. (No. 1749). 1920 (Nov.): 1 imm. 8 (No. 201625) ; 2 imm. 2 (No. 201- 624). 1921 ( Jan.- April ) : 1 8 (No. 21312). 1922 (Feb.-Oct.) : 2 imm. 8 (No. 221147). 1924 (Feb.-May) : 1 8 holotype (No. 24182) ; 3 8, 2 2, inch imm. & juv. (Nos. 24111, 241013, 241014). Caripito: 1942: April 1-15, 1 2 paratype (No. 42447) ; April 18, 1 8 paratype (No. 42207) ; April 16-30, 2 8, inch juv. (No. 42448) ; Mav 23, 1 juv. 2 (No. 42273) ; Aug. 15-31, 1 imm. 8 (No. 42449). This species is named for its jungle hab- itat. IV. Bibliography. Banks, N. 1929. Spiders from Panama. Bull. Mus. Comp. Zool. Harvard. Vol. LXIX, pp. 53-96. Beebe, W. 1925. Studies of a tropical jungle; one- quarter of a square mile of jungle at Kartabo, British Guiana. Zoologica, Vol. VI, pp. 1-193. 1943. Physical factors in the ecology of Caripito, Venezuela. Zoologica, Vol. XXXVIII, pp. 53-60. Cambridge, F. O. P. 1900. Arachnida. Araneidea and Opiliones, Vol. II, in Biologia Centrali- Ameri- cana. Cambridge, O. P. 1882. On new genera and species of Aranei- dea. Proc. Zool. Soc. London, 1882, pp. 423-442. Chamberlin, R. V. & W. Ivie. 1936. New Spiders from Mexico and Panama. Bull. Univ. Utah Biol. Ser. 3(5), pp. 1-103. Mello-Leitao, C. 1917. Notas arachnologicas. Broteria Lisboa Ser. Zool. 15, pp. 74-102. 138 Zoologica: New York Zoological Society Peckham, G. W., E. G. Peckham, & W. H. Wheeler. 1888. Spiders of the sub-family Lyssomanae. Trams. Wise. Acad. Sci. Arts & Letters, Vol. VII, pp. 222-2 56.2 Petrunkevttch, A. 1911. Spiders of North America, South America and the West Indies. Bull. Am. Mus. Nat. Hist., Vol. XXIX, pp. 1-809. 1928. Systema Aranearum. Trans. Conn. Acad. Arts & Sci., Vol. 29, pp. 1-270. 1933. An inquiry into the natural classifica- tion of spiders, based on a study of 2 The plates (XI, XII) are missing in most of the extant copies of this paper, including the stock now on hand at Madison, Wise. Dr. Petrunkevitch, however, very kindly sent me photographs of tracings he made from a complete copy some years ago. their internal anatomy. Trans. Conn. Acad. Arts & Sci. Vol. 31, pp. 299- 389. Simon, E. 1900. Etudes arachnologiques. xlvii. Descrip- tions d’especes nouvelles de la famille des Attidae. Ann. Soc. Ent. France. Vol. 69, pp. 1-61. 1901. Histoire naturelle des Araignees. Vol. II. pp. 381-429. Taczanowski, L. 1871. Les Araneides de la Guyane frangaise. Horae Soc. Entom. Ross., Vol. VIII, pp. 32-132. 1872. Les Araneides de la Guyane frangaise. Horae Soc. Entom. Ross, Vol. IX, pp. 64-159. Dharmakumarsinhji: Breeding of Birds of Paradise 139 17. Notes on the Breeding of the Empress of Germany’s Bird of Paradise in Captivity. Prince K. S. Dharmakumarsinhji of Bhavnagar. (Plate I). Foreword by Jean Delacour. Birds of Paradise have exerted a strange fascination ever since men have known them. This is not surprising, as their beauty and showiness cannot be surpassed. They have been eagerly sought by aviculturists, as well for public zoological gardens as for private aviaries. Nearly all the many different species have been obtained from their native haunts in and around New Guinea and kept in Eu- rope, in America, in India and elsewhere. So far, however, none had ever been bred in confinement. Eggs had been laid, and as re- cently as the last two years the pair of Long-tailed Birds of Paradise in the New York Zoological Park have nested repeated- ly, but no further results followed. Prince Dharmakumarsinhji is the first aviculturist to have met with complete suc- cess. A young Empress of Germany’s Bird of Paradise has been reared in his aviary. He must be heartily congratulated for such an achievement, more so for having noted so carefully all the observations he has made on this outstanding event. * * * On receiving a telegram from Mr. Shaw Mayer, in May, 1939, to the effect that he was passing through Bombay on his way to England with a collection of tropical birds, including many species of Birds of Paradise, I wired to my friend, the Inspector of Avi- aries of His Highness the Maharaja Sahib of Bhavnagar, to go to Bombay. I was then at Mussoorie recovering from ill-health; had I been well I could not have resisted seeing this marvellous collection of oriental birds as it passed through India. Within a few days I was informed that the Inspector had bought a pair of Paradisea apod, a augus- taevictoriae and a male of the Lesser Superb Bird of Paradise (Lophorina superba minor ) and that they were on their way home to Bhavnagar. On arrival the pair was set free in His Highness’ Palace aviary in a suitable cage facing west, the cage having an interior room higher up if they required more se- clusion. The roofing was of tiles with a good 45-degree slope. Below the tiles was wood. Surrounding the aviaries were large trees so that the cages kept cool during the heat of the day. Moreover, the evening sun could only penetrate at a late hour. This environ- ment evidently suited them very well. In the season during which the birds arrived, they began moulting rapidly, and there was nothing specially interesting ex- cept that they slowly became accustomed to the special diet that is given to our In- dian birds, which I shall later mention. The following spring, at the beginning of February, 1940, the male bird showed signs of display and started calling vigor- ously. The call is a harsh tvauk wauk wauk wauk wauk continued, as described by dif- ferent ornithologists concerning the Greater Bird of Paradise. This call is repeated often during the day and is the general call; there are many others that the bird emits. As the hot weather approached, the male was seen displaying. My assistant, the Avi- ary Inspector in whose charge I had en- trusted the birds, mentions that he saw the male displaying as follows: “He low- ered his head, drooped his wings and erect- ed his plumes above his back. Sometimes the plumes were not completely erected, how- ever. All this could be seen through the first cage. I had a glimpse of him while he was moving from side to side, hopping on his favorite perch — a horizontal piece of straight wood one inch thick, and three feet long.” I was unfortunately not able to witness the full display as described by my assistant. Suddenly during the next month, April, 1940, the male commenced shedding his beautiful plumes and to my astonishment they were all discarded within a fortnight. It was then that we noticed that he showed signs of real courting, chasing the female and calling in tones which varied from gur- gles to grunts. It was his custom to lower 140 Zoologica: New York Zoological Society [XXVIII: 17 his head at the same time, emitting short grunts as he approached the female. These gestures appeared to be his true courting. The particular grunt that he emitted then resembled the sound of a motor-bicycle horn. The courting took place in the early hours of the morning and after the new food was put in the cage, which was at 10 A.M., and also during the quiet hours of the after- noon. During the middle of April the female was seen picking up nesting material. Hence an old crow’s nest was put on a branch erected inside the cage. Twelve days later the female had constructed a cup-shaped nest of coconut fibres and new leaf stems inside the crow’s nest and we found that she had started incubation of a single egg. The hen bird was very regular, sitting tighter as the days advanced, leaving the nest two to three times a day and only coming down to feed, refresh herself and preen her feathers. After the 25th day of incubation the hen bird became very irregular and on the 28th day I could not resist seeing the egg again and examining it. Most unfortunately, the egg was found to be infertile. I now have it in my collection ; its measurements are : length, 3.65 cm. ; middle 2.65 cm. During the month of November, 1941, I moved into my own house, Dil-Bahar, where I had special aviaries constructed to suit my birds and where I took over the pair of Bii’ds of Paradise. Here I devoted most of my time to making notes at the commence- ment of the breeding season. Out of the four cages facing east, the second from the south I selected for the pair. All of the cages were of equal size but the interior was colored differently and there was no extra room as in the palace aviaries. The two middle cages are closed in on three sides with net windows and have a Venetian type of window at the back of the wall for ventilation. The front is wire netting with a door for each cage. The cage in which the Birds of Paradise were kept (cage No. 2) is furnished with plants (cro- tons), dry perches of -teak wood laid hori- zontally, and an old stump of a tree with branches put here and there. A cemented water course passes through all the avi- aries to supply fresh running water for drinking and bathing. The food of the birds has been virtually unchanged since they came into my avi- aries. It consists of papaya, bananas and chogo, which latter is a mixture of flour, eggs, meat and ghee made out of butter. Live food is also given, including grass- hoppei’s, mole-crickets and locusts. I find that the birds relish this mixed diet and thrive on it. The breeding of my pair would indicate an adequate diet. Throughout the winter months the male was calling, although not as vigorously as in February and March. The diversity of calls that he emitted was extraordinary, varying in strength in different degrees and rather difficult to describe. Neverthe- less I was able to notice that the male had definitely a peculiar call during the real courtship or rather during the height of love-making. This call was a short sylla- bled, horn-like sound as mentioned before. There was also another peculiar call that seemed similar to that of our Indian tree- pie; this also was heard only during the breeding season. The former love call is generally emitted at the approach of the female or when close to her. Moreover, it was the male’s daily habit to bow his head from side to side, then suddenly jerking it up like a mallard drake and shaking it in pride. During the entire month of March, the male bird was calling incessantly throughout the day but no special display was to be seen except at the end of the month when he was heard emitting his peculiar court- ship grunts. At this time both birds were very shy, especially the female which had a very acute sense of danger and vigilant eye- sight. The latter was indicated when I was in the habit of looking from the adjoining aviary (No. 3) to see if I could catch the male displaying or luckily happen to see them mating. In order to do so I had drilled two small holes through the cement sheet- ing on the sides of the aviary. The two holes were three feet apart and the size of a 22 calibre rifle bullet hole. From here I could observe from time to time the male coming down to feed, taking his daily bath in the early morning and dancing about on the branches. After taking his dip he would dry himself on his perch, preening his beautiful plumes. The hen bird would also bathe but not so regularly as the male. This may have been because she could per- ceive me through that minute hole and catch any slight movement that I made. She was very wary and became more sus- picious toward the direction of the holes. However, it does not seem unusual among these birds to be suspicious, for in the dense jungles of New Guinea there must be many of their enemies lurking close by. Moreover, they are birds that normally pre- fer absolute seclusion among dense foliage. While the male had his own courtship and display, I noticed also that the female seemed very active. She would exercise in a peculiar manner which coincided with the display of the male. She would leave her perch, fly toward the open side of the cage and make one or two short circles and re- turn. At first I imagined this to be merely the usual exercise. But making further ob- servations I noticed that it was a part of the pro-breeding display, if one may call 1943] Dharmakumarsinliji: Breeding of Birds of Paradise 141 it so. She would sometime fly around the male in this manner, too. This latter be- havior overcame my doubt. The female Bird of Paradise had a habit of moving from one side to the other on a branch, often turning completely round and flipping her wings as do crows. This be- havior of wheeling, as I might call it, is also to be seen among babblers. They are very cautious birds and do not alight on the ground if they can help it. This charac- teristic is so pronounced that they will climb down a branch, lower the head, take a drink, go up again and fly to their perch. When coming down to feed they will fly to a slender branch of a bush, alighting invari- ably in a horizontal fashion. Then they wheel about as they climb down with one foot above the other, see that there is no danger, and then come down to feed. I have seen the female scale an old trunk of a tree not unlike a woodpecker looking into the crannies for insects. During April, although the climate be- came a little warmer as the southeast winds began to blow, there was a cool breeze com- ing from over the sea. This seemed to stim- ulate the birds immensely. The male started his displays. Hopping on his branch, with both feet simultaneously, he would jump up one or two inches, as it seemed; in spring- ing himself up he would also move his wings rapidly. Then he would fly to another branch, do the same and return. He would also hop from side to side on his favorite perch, lowering his head from one side to the other and making his courtship grunts. This kind of display lasted fifteen days from April 5 to April 20. During the second week of April, after the 17th, the hen bird was for the first time seen picking up a stem of lucerne in her beak and carrying it towards nest No. 2, an old crow’s nest com- posed of bits of wire, which had been placed in the northeast corner of the aviary about a month before. This was the first sign of nest building. The southeast winds continued. On April 21 the male was not calling as much as usual. The next day I arranged to put creepers, lucerne and fibers for nesting material into the aviary. New material was given each day but the female preferred the lucerne, flying to nest No. 2 and depositing the succulent stems untidily. Many of them dropped out of the nest, while some were not even properly arranged. The building procedure continued every day but only in the morning hours between 9 and 10:30 o’clock. Some creepers fell down each day while new ones were roughly put on the nest. As the days advanced only a few re- mained on the nest. This type of haphazard nest building went on until April 27 when everything ceased and there were no signs of picking up of material. By this time only a few stems of lucerne, that had fixed them- selves in the wire nest, remained. Whether the hen bird had become suspicious or wheth- er she was bluffing was a question to be an- swered later. However, I kept a close watch and found that the male bird was again calling vigorously and was seen chasing the female. On May 6 I was spraying water with my hose pipe into the cages because of the tremendous heat. It was my habit to spray the casuarina branches that screened part of the south and west walls at the back of the aviary. I did not notice the hen bird move as she regularly did from behind the screen. However, a few minutes later when I had stopped spraying, she flew out of the casuarinas and she perched on a branch in the front part of the cage. She immediately started drying herself, and having done so at once flew back to her usual resting place. In spite of this new observation, which I then took to be quite natural, I passed the cage to water the others. Two days later, on May 8, my aviary boy reported the nest and a single egg of the Paradise Birds among the casuarina branches. His suspicion had been aroused because the hen bird was not to be seen in the cage and on investigating in the casuarina screen, he eventually flushed the bird off the nest. I confirmed his statement by seeing the egg by means of a small mirror fastened to a cane stick. The nest was a reg- ular cup shape and was placed between the stems of the thickest clump of branches. It was quite invisible from outside. A glimpse of the hen bird could be seen from the adjoining aviary but only if the place were pointed out to one. Such was the cunning of the female which had ingeniously avoided our attentions and had surreptitiously built her nest. From May 8 I took particular care and put down my notes. The hen bird would come down from the nest to feed and clean herself three or four times a day and would remain out from five to ten minutes on each occasion. The male kept fairly silent and never interfered, only making his grunts and gurgles when the hen left her nest. I could generally tell when the female had left the nest by the male making his love calls. On May 17 and 18 the hen was very hard set. On May 20 one of my female dayal birds got into the cage and caused a little trouble before I could remove her. Actually I opened the first cage window so that she could return but I found by doing so a pair of spreo starlings entered the cage, too. However, I managed to entice them out except for a spreo which was so obstinate as not to leave the cage. The next day, May 21, at 2 P.M., I found near the front of the cage an empty egg shell which I at once recognized as that of 142 Zoologica: New York Zoological Society [XXVIII: 17 the Bird of Paradise. The first thought that gripped me was that the spreo starling had done the mischief, for often they have swal- lowed and destroyed the eggs of other birds. The next instant I was in the cage with my mirror to examine the nest and to my de- light I saw a little chick, an absolutely pink little thing lying on its side. At once I had the spreo removed. Counting the days from May 8, the incubation period evidently was 13 days, but I am inclined to think that the egg was laid earlier — probably on May 6, so that the incubation period would be 15 days. However, this is a point to be con- firmed when better chances afford. The hen brooded the chick most of the time and started feeding it, as I noticed, at midday. The feeding procedure was ex- tremely interesting. Grasshoppers were put down. She would select one, 1 or 2 inches long, take it to a near branch, then fix it in her claws and start removing its legs, then its wings, and finally would swallow it whole. Seeing that there was no danger, she would fly to the nest and regurgitate the food into the mouth of the young. After four days the hen bird would take two grasshoppers at a time and regurgitate them both into the mouth of the nestling, one by one. During the next week large grasshoppers were given as well as locusts. These the hen apparently preferred to the smaller ones. In this case also she would take a locust or large grasshopper and pro- ceed with it in the usual manner, removing the legs and wings and also the intestines. The latter were cleverly removed. Then piece by piece she would swallow the soft abdomen and lastly the head, which evi- dently was an edible portion of the body. However, there was one special peculiarity that I marked about the parent bird during the feeding; that was that she fed the young only twice at each feeding time. This seemed to me very strange. However, when the nestling was 10 days old she would take two or three small grasshoppers and swal- low them entire at a time, and would re- gurgitate them out one by one. She would only bother herself to remove the legs and wings in the case of larger insects. After each feeding the hen picked up the excreta of the chick, which she swallowed in the manner of many other birds during their parental care. During the first week, I started with only three feedings, one at 9 A.M., the second at noon and the third at 4 P.M. During the second week I increased the feedings to five, at 7 :30 and 11 A.M., 2, 4 and 6 P.M. and continued the same until the young left the nest. The number of insects varied at each feeding. When large insects were given the number was usually small — thus, five or six large insects or as many as twenty small ones. The voice of the chick could be heard at a distance of 10 feet when it was one week old and at 15 days the chattering was audi- ble at 30 feet. The temperature in the shade of June 3 registered 110 F. This was at the hottest part of the day. The heat was intense, especially on the roof and inside, despite two layers of cement sheeting with a one- inch air space. On Friday June 5, there was a distressing scene, the nestling falling to the ground. Fortunately it escaped injury. Why it should have fallen seemed a mystery, but I later concluded that it was a result of activity induced by the extreme heat and the lack of ventilation near the nest. At the end of the first fortnight the tender wing feathers were clearly visible, the pec- torals were merely hairs, and the tail feath- ers were starting to grow. The nestling was handled with care and was returned to the nest. It croaked once or twice while on the hand. The legs were still colorless and white. The iris of the eye was lead gray. During this period of intense heat the nestling fell out of the nest again, but since there was straw on the ground, it did not sustain in- jury and was returned to the nest. On June 8, 10 and 12 I took photographs of the parent bird feeding the nestling. After June 10 the temperature dropped and it varied from 105 to 107 degrees F. The male bird took no part in parental care, but on the contrary became quite a nuisance at times. In fact, he seemed rather henpecked. The female would fly at him and claw him if he ventured too near the nest. The male became sluggish, and would sit placidly in his usual perch in the corner opposite the nest. He was removed as the young got older, as a safety measure. To give a more detailed description of the nest, it was cup-shaped and measured 4x/4 inches wide and 3Y2 inches deep. It was com- posed mostly of causarina leaves and creep- ers of Jacaramontsia, which has a beautiful blue flower. There were, however, a few coconut fibres. The height from the ground was 7 ft. 4 in. and the nest was situated in the southwest corner of the cage. On June 21 the nestling was able to fly out of the nest. It was fully fledged and looked very much like its mother except that the nape was not so lightly colored as in the parent bird. The eyes were different and the legs lighter in color. It would fly behind its mother for food, and slowly started to feed on its own. During the entire period of parental care the hen bird emitted a call that I had not previously noted. It was an alarm call, sounding like Kurr Kurr Kurr, resembling the call of some of the larger woodpeckers. This call was only heard during the period of parental care and was quite different from the call the female ordinarily emitted. 1943] Dharmakumarsinhji: Breeding of Birds of Paradise 143 Her call does not vary as much as that of the male, but is shriller. [On April 28, 1943, Prince Dharmakumar- sinhji wrote Dr. A. Wetmore, of the Smith- sonian Institution, as follows] : “My adult male Paradisea apoda augus- taevictoriae suddenly died on February 17, although a week before he was seen courting and I had every hope that the pair would breed successfully again. “I had been surprised when my cage boy reported that the adult hen was showing signs of nest building. This was on March 7. The situation of the nest this time was in the adjacent corner or N. W. direction of the aviary among the casuarina branches. I promptly inspected the place and could not ascertain it as a nest, although I had seen the hen bird sitting there quite often and breaking leaves. Ten days later nest build- ing commenced in the usual slow manner on a dried bush of a Duranta situated in the front part of the aviary. Creepers with their blue flowers hanging on their stems were placed on the bush each day until active nest building took place on March 20 and 21 and a complete cup-shaped nest was built. Moreover, on March 22 the hen started incubating a single egg. The nest was com- posed of creeper and pieces of coconut fibre with a bit of string and was lined inside with casuarina leaves. The height from the ground was 3 feet 7 inches. “Incubation was very regular. She would not stir even when the cage boy went in regularly to clean the cage and put down the food. On April 14 I took the egg, which I believe was addled, while it was still being incubated most regularly. For the next few days the adult hen bird could be seen perched on the nest with a mournful air. After that she began destroying the nest completely. Let me hope that the young bird of her previous nesting will turn out to be a male.” 144 Zoologica: New York Zoological Society EXPLANATION OF THE PLATE. Plate I. Fig. 1. Female Empress of Germany’s Bird of Paradise on her nest in the aviary. Fig. 2. Female feeding the nestling. DHARMAKUMARS1NHJI. PLATE I. FIG. 1. FIG. 2. NOTES ON THE BREEDING OF THE EMPRESS OF GERMANY'S BIRD OF PARADISE IN CAPTIVITY. Pick: Healed Hunger osteomalacia in a Monkey 145 18. Evidence of Healed Hungerosteomalacia (Late Rickets) in a Green Monkey ( Cercopithecus sabaeus ). Joseph Pick, M.D. Department of Anatomy, New York University, College of Medicine (Plate I). Deformities of the skeleton in domestic and captive animals, particularly in mon- keys, have been repeatedly reported in the literature. The value of many of the earlier reports, however, is limited to the mere recording of the deformation. The patho- genesis of the skeletal changes is often omitted or dealt with rather incompletely, as in the accounts of Bland Sutton (1888) and Draseke (1906) who diagnosed all de- formities found in monkeys, lions, bears, birds and dogs as rickets, although some of their cases had undoubtedly nothing to do with vitamin D deficiency. The confusion in the pathology of bone in animals was largely cleared up by Chris- teller (1923) who definitely distinguished rickets and ostitis fibrosa v. Recklinghausen in a considerable number of his monkey cases. However, Christeller has not suffici- ently separated ostitis fibrosa from osteitis deformans (Paget’s disease), probably be- cause these diseases were at that time not recognized as two entirely different condi- tions. After the comprehensive study of Christeller who had based his diagnosis on careful macroscopical, histological and radio- logical examination, only a few well studied cases of bone softening in animals have been recorded. Osteomalacia in cattle was clinical- ly observed by Meador (1927) and McIntosh (1928), in a racing pony by Hewlett (1927) and in swine by Kinsley (1928) and Bouchet (1931). Rickets in carnivores was observed by Maignon (1931) and late rickets in a fossil bear by Jarisch (1937). Since in the present case a fairly complete clinical, post mortem and laboratory ex- amination could be carried out, a brief record of these findings may be a useful contribution to the knowledge of bone path- ology in captive animals, and a help to the veterinarian in evaluating his cases. Case Report. History. A female green monkey ( Cercopithecus sabaeus ), about three years old, which had been fed on a diet consisting only of milk and bread, was given to the Zoological Park, Bronx Park, New York City, because it had bitten its owner. A deformity of the right thigh was detected, but otherwise the mon- key seemed normal. The animal was trans- ferred to the Department of Anatomy, New York University College of Medicine, for further study. After radiological studies and chemical examination of the blood had been made, the monkey was killed with chloroform and the organs fixed in Zenker’s fluid. For comparison the macroscopical, radiological and histological examination of a normal monkey of the same species was carried out. Gross Anatomical Findings. The complete post mortem examination revealed no pathological deviations in com- parison with the normal subject except for certain abnormalities of the bones of the extremities and of the parathyroid bodies. In the upper limb the humerus on both sides was bent between the upper and middle thirds to form an angle of approximately 120°. The length of each humerus measured 9 cm., the diameter at the point of bending 1.4 cm. The bone appeared to be shortened about 3 cm. but thickened about 0.5 cm. Radius and ulna on both arms measured 11 cm. in length, the same as in the normal. They were slightly bowed within the lower third. The joints of the upper limb were freely movable. In the lower limb the femur on both sides were markedly bowed and twisted in the lower third. The bones meas- ured 9.5 cm. on the right and 10 cm. on the 146 Zoologica: New York Zoological Society [XXVIII: 18 left side in length and appeared to be shortened. Fibula and tibia measured 12.5 cm. in length, the same as in the normal. The fibula was of normal shape but there was bending of the shaft of the tibia. The joints of the lower limb were freely move- able except for the right knee joint which showed marked impairment of mobility due to the deformation at the lower end of the femur. All the bones were of hard consistency. There was nowhere any thickening in the epiphyseal regions or costocartilaginous junction, nor any sign of softening. The periosteum could be easily removed from the bone and did not show any trace of inflammation or haemorrhage. Three para- thyroid bodies were found. A fourth may have been encased by the thyroid, but was not detected as the thyroid was not cut into serial sections. The left upper parathyroid body was twice the length of the two others. Microscopical Findings. Sections were taken from the following organs : clavicle, scapula, sternoclavicular junction, humerus, femur, radius, ulna, fib- ula, tibia, pubic bone, calvarium, basis of skull, fifth lumbar vertebra, sternum, sup- rarenal, heart, thymus, thyroid, ovary, kidney, parathyroid and liver. From the histological examination of the bones, the trabeculae, particularly within the areas of bending of the femur and humerus, showed signs of active resorption and deposition of new bone substance. There was a mod- erate number of fractured trabeculae en- veloped by numerous cells, particularly of the osteoclastic type and numerous thin- walled congested blood vessels. Other tra- beculae were lined by osteoblasts on one surface and on the opposite by osteoclasts. These findings were interpreted as a struc- tural reorganisation of the ground sub- stance. There was, however, nowhere any sign of extensive osteoporosis or patholog- ical bone formation, such as the mosaic structure commonly found in osteitis de- formans Paget or the fibrosis of v. Reck- linghausen’s disease. The costocartilaginous junction and all the epiphyseal regions ex- amined showed no pathological deviation. The bone marrow was mixed, partly cellular, partly fatty in type. All the soft organs were normal except for the parathyroids in which the number of oxyphile cells was considerable increased (Plate I, Fig. 1). Radiological Findings. The X-ray examination of the bony skele- ton revealed marked deformity of the long bones as a result of extensive bowing and twisting. These changes were most marked in the upper third of the humeri (Plate I, Figs. 2, 3) and lower third of the femora. In these regions the cortex had become very thin and was associated with an increase of the medullary trabeculation. The cortex and medullary cavities of the remainder of these bones as well as the bones of the forearm and leg appeared normal. Blood Chemistry Findings. Calcium 10.53 mg. in serum. Inorganic phosphorus 5.856 mg. in serum. Bodansky phosphatase units 9.95. Calcium/inorganic ratio 1.8. Discussion and Conclusions. Interpretation of the findings encountered some difficulties because of the lack of any pathognomonical sign. The only pathological features of the case were the deformities as revealed by gross and radiological examina- tion, the microfractures associated with re- construction of bone substance, previously described in the human by Looser (1920), Fromme (1921), and Salinger (1929), and finally the unusual number of oxyphile cells within the parathyroid bodies. The evalua- tion of the chemistry findings in the blood was difficult, because the normal figures for the green monkey are apparently not de- termined. In the present case they corres- pond approximately to that of a normal human infant and the comparatively high amount of inorganic phosphorus as well as phosphatase might be significant for a heal- ing process of the skeleton. Doubtless, there was no acute process acting, for the animal was in good health and softness of the bones could not be de- tected anywhere. It seemed rather that the monkey had previously suffered from a temporary disease associated with softening of the skeleton which had primarily affected the bones of the extremities, no doubt be- cause the latter, as supports of the body, were particularly exposed to pressure from the body weight. After the disappearance of the noxious cause the bone substance apparently regained its normal histological features, although the deformities were never corrected. In adaption to the now al- tered lines of force a reorientation of the trabeculae within the deformed bones took place, with resorption of fractured or use- less trabeculae and formation of new bone. In accordance with Erdheim (1914) the large number of the oxyphiles in the para- thyroids were interpreted as hypertrophy necessitated by the increased demand for new bone substance. In regard to diagnosis osteitis deformans of Paget, ostitis fibrosa v. Recklinghausen and scurvy can safely be ruled out. The condition was considered to be most prob- ably a case of healed hungerosteomalacia — a variety of late rickets — caused by a tern- 1943] Pick: Healed Hunger osteomalacia in a Monkey 147 porary deficiency in vitamin D and by in- sufficient exposure to sunlight. Ordinary rickets could be excluded, since the epiphy- seal regions were not affected at all. Hungerosteomalacia was observed in the adult human in Vienna, Munich and other cities of the defeated countries after the World War I by Schlesinger (1919, 1921), Wenckebach (1919), Edelmann (1919), Porges and Wagner (1919) Cramer and Schiff (1920), Heyer (1920), Steiner (1927), Eyermann (1932) and Muir Craw- ford (1934). The radiological appearances of the condition were described by Eisler (1919). It is associated with pain and de- formities in these bones which are par- ticularly exposed to pressure and tension such as the spine, long bones and the thor- acic cage. Radiologically there is no change in mild cases, but osteoporosis, microfac- tures, loss of minerals have been observed in the advanced stages of the condition. Successful treatment with cod-liver oil and vitamin D was reported by Dalyell and Chick (1921) and Hume and Nirenstein (1921). Such an interpretation of the present case seems to be justified because the monkey was doubtlessly kept on a deficient diet and had apparently been rarely exposed to sun- shine. In accordance with the treatment em- ployed in the human and that suggested by Hume, Lucas and Henderson (1929) in animals, administration of vitamin D and sufficient exposure to sunlight is recom- mended as preventative measure against the establishment of osteomalacia in captive animals. Summary. Deformities of the long bones of all four extremities in a green monkey ( Ceropithe - cus sabaeus ) was found on gross and radio- logical examination. There was active re- organization of the trabeculae of the bones and hypertrophy of the oxyphile cells within the parathyroids. The condition was con- sidered to be a case of healed hungerosteo- malacia due to insufficient intake of vitamin D and exposure to sunshine. The material for this study was provided by the New York Zoological Park, Bronx Park, New York City, and I want to ac- knowledge the kind cooperation of Dr. Leonard J. Goss in this work. References. Bland-Sutton, J. 1888. J. comp. med. and surg. 10, 1-29. Bouchet, A. 1931. Bull. V Acad. vet. de France. 4, 417-419. Christeller, E. 1923. Erg. all. Path. u. path. anat. d. Mensch. Tiere. 20, 1-184. Cramer, A. & P. Schiff 1920. Rev. med. de la Suisse Romande. 40, 746-763. Dalyell, E. J. & H. Chick 1921. Lancet. 2, 842-849. Draseke 1906. Z. Ethn. 38. 751-752. Edelmann, A. 1919. Wien. klin. Woch. 32, 82. Eisler, F. 1919. Wien. klin. Woch. 32, 605-606. Erdiieim, J. 1914. Wien. Akad. Wiss. 6, 116. Eyermann, Ch. H. 1932. J. Miss. State Med. Assn. 29. 551-554. Fromme, A. 1921. Arch. klin. Chir. 116, 664-680. Hewlett, K. 1927. The Vet. J. 83. 510-511. Heyer. 1920. Munch, med. Woch. 67, 98. Hume, E. M., N. S. Lucas & H. Henderson. 1929. The Vet. J. 85, 490-492. Hume, E. M. & E. Nirenstein. 1921. Lancet. 2. 849-853. Jarisch, A. 1937. Wien. klin. Woch. 50, 729-730. Kinsley, A. T. 1928. Vet. Med. 23. 461-462. Looser, E. 1920. Dtsch. Z. Chir. 152, 210-357. McIntosh, R. A. 1928. Vet. Med. 23. 304-306. Meador, D. J. 1927. The Cornell Vet. 27, 61-63. Maignon, M. 1931. Bull, de I’acad. vet. de France. 4, 411-414. Muir Crawford, A. & D. P. Cuthbertson. 1934. Quart. J. Med. 27. 87-104. Porges, O. & R. Wagner. 1919. Wien. klin. Woch. 32, 385-387. Salinger, H. 1929. Fortschr. Geb. Roentgstr. 39, 1049- 1059. Schlesinger, H. 1919. Wien. klin. Woch. 32, 245-247. 1921. Wien. klin. Woch. 34, 213-214. Steiner, P. 1927. Wien. med. Woch. 77, 972. Wenckebach, K. F. 1919. Wien. klin. Woch. 32. 295-296. 148 Zoologica: New York Zoological Society EXPLANATION OF THE PLATE. Plate I. Fig. 1. Oxypyile cells of the parathyroid, in- cluding a few follicles of the thyroid. Fig. 2. X-ray of the right humerus showing the deformation and increased medul- lary trabeculation. Fig. 3. X-ray of the right humerus of a normal green monkey. PICK. PLATE I. FIG. 1. FIG. 2. FIG. 3. EVIDENCE OF HEALED HUNGEROSTEOM ALACIA (LATE RICKETS) IN A GREEN MONKEY (CERCOPITHECUS SABAEUS). Hertlein & Strong: Mollusks from the Pacific 149 19. Eastern Pacific Expeditions of the New York Zoological Society. XXXII. Mollusks from the West Coast of Mexico and Central America. Part. II.* Leo George Hertlein & A. M. Strong. California Academy of Sciences. Plate I. [This is the thirty-second of a series of papers dealing with the collections of the Eastern Pacific Expeditions of the New York Zoological Society made under the direction of William Beebe. The present paper is concerned with specimens taken on the Templeton Crocker Expedition (1936) and the Eastern Pacific Zaca Expedition (1937-1938). For data on localities, dates, dredges, etc., refer to Zoo- logica, Vol. XXII, No. 2, pp. 33-46, and Voi. XXIII, No. 14, pp. 287-298.] Contents r . PAGE Introduction 149 Genus Glycymeris Da Costa 150 Submenus Glycymeris s.s 150 Glycymeris ( Glycymeris ) gtgantea Reeve 150 Glycymeris ( Glycymeris ) maculata Boderip.. 150 Subgenus Tuceta Bolten 151 Glycymeris ( Tuceta ) multicosta ta Sowerby. . . 151 Glycymeris ( Tuceta ) tessellata Sowerby 151 Glycymeris ( Tuceta ) tessellata canoa Pilsbry & Olsson 152 Glycymeris ( Tuceta ) tessellata strigilata Sowerby 152 Subgenus Axinactis Morch 153 Glycymeris ( Axinactis ) delessertii Reeve 153 Glycymeris (Axinactis) inaequalis Sowerby. . . 153 Family Arcidae . . . 153 Genus Area Linnaeus 154 Subgenus Area s.s 154 Area (Area) fernandezensis Hertlein & Strong, nom. nov 154 Area (Area) mutabilis Sowerby 154 Area (Area) pacifica Sowerby 155 Subgenus Acar Gray 155 Area (Acar) gradata Broderip & Sowerby. . 155 Subgenus Anadara Gray 155 Area (Anadara) biangulata Sowerby 155 Area (Anadara) formosa Sowerby 156 Area (Anadara) mazatlanica Hertlein & Strong, sp. nov 156 Area (Anadara) reinharti Lowe 157 Area (Anadara) similis C. B. Adams 157 Area (Anadara) tuberculosa Sowerby 157 Subgenus Areopsis von Koenen 158 Area (Areopsis) solida Sowerby 158 Subgenus Lunarca Gray 158 Area (Lunarca) vespertina Morch 158 Subgenus Barbatia Gray 158 Area (Barbatia) reeveana d’Orbigny 158 Subgenus Calloarca Gray 159 Area (Calloarca) alternata Sowerby 159 Subgenus Cara Gray 159 Area (Cara) emarginata Sowerby 159 * Contribution No. 671, Department of Tropical Research, New York Zoological Society. 1 Hertlein, L. G., and Strong, A. M. Eastern Pacific Expeditions of the New York Zoological Society. XXII. Mollusks from the West Coast of Mexico and Central America. Part I. Zoologica, New York Zool. Soc., Vol. 25, Pt. 4, December. 31, 1940, pp. 369-430. Subgenus Cun ear ca Dali 159 Area (Cunearca) aequatorialis d’Orbigny ... . 160 Area (Cunearca) bifrons Carpenter 160 Area (Cunearca) esmeralda Pilsbry & Olsson. 160 Area (Cunearca) nux Sowerby 161 Area (Cunearca) perlabiata Grant & Gale. ... 161 Subgenus Larkinia Reinhart 161 Area (Larkinia) grandis Broderip & Sowerby 161 Area (Larkinia) multicostata Sowerby 162 Subgenus Scapharca Gray 162 Area (Scapharca) cepoides Reeve 162 Area (Scapharca) concinna Sowerby 162 Area (Scapharca) obesa Sowerby 163 Genus Noetia Gray 163 Subgenus Noetia s.s 163 Noetia (Noetia) reversa Gray in Sowerby.. 163 Subgenus Sheldonella Maury 163 Noetia (Sheldonella) delgada Lowe 163 Subgenus Eontia MacNeil 163 Noetia (Eontia) olssoni Sheldon & Maury.. 163 Genus Pteria Scopoli 164 Pteria sterna Gould 164 Genus Pinctada Bolten 164 Pinctada mazatlanica Hanley 164 Genus Pinna Linnaeus 165 Pinna rugosa Sowerby 165 Genus Atrina Gray 165 Atrina maura Sowerby 165 Atrina oldroydii Dali 166 Atrina texta Hertlein, Hanna & Strong, sp. nov 166 Genus Pedalion Solander in Huddesford 166 Pedalion chemnitzianum d* Orbigny 166 Introduction. In Part I of this series of papers* 1 the authors outlined a plan for a complete cata- logue of the tropical west American mol- lusks. Since the publication of that plan, conditions resulting from unsettled inter- national political relations have made it necessary to change and drastically reduce the entire plan. Under the existing condi- tions it is thought better to record now the mollusks collected on the Zaca Expeditions, rather than to wait indefinitely for the pub- lication of monographic reports on the fauna as a whole. In the revised plan of publication only the species secured by the Zaca Expeditions of 1936 and 1937-1938 will receive formal headings. These will be followed by a refer- ence to the original description and where desirable an additional reference to an il- lustration or to an important discussion of the species. The type locality, range, col- lecting stations, a brief description or de- 150 Zoologica: New York Zoological Society [XXVIII: 19 scriptive notes, and the distribution will be given. Some of the species heretofore unillustrated will be figured as will the new species. Keys will include only the species collected on the Zaca Expeditions. It is hoped that papers of monographic scope dealing with families of tropical west American mollusks may be published from time to time in this or in other scientific periodicals. In that way it may be possible to continue, in a modified form, the original plan of a complete catalogue of the mollus- can fauna of this interesting region. The results of studies of several of the families are now in manuscript awaiting publication. Acknowledgement is especially due Dr. G. D. Hanna, Curator, department of Paleon- tology of the California Academy of Sci- ences, for assistance and suggestions. Ac- knowledgement is also due Dr. U. S. Grant, IV, of the University of California at Los Angeles, Dr. H. G. Sc-henck and Dr. A. M. Keen of Stanford University, Mr. A. G. Smith, Berkeley, California, Dr. H. R. Hill and Mr. G. Willett of the Los Angeles County Museum of History, Science and Art, and Mr. C. G. Abbott and Miss Viola Bristol of the San Diego Society of Natural History. The preparation of the photo- graphs by Mr. Frank L. Rogers used in this paper is here acknowledged: his work was accomplished during the course of Fed- eral Works Progress Administration Proj- ect Number 8569. CLASS PELECYPODA. Order Prionodesmacea. Superfamily Arcacea. Family Glycymeridae. Genus Glycymeris Da COSTA Key to the Species of Glycymeris. A. Shell large, smooth or with fine radial striations a. Color pattern of reddish-brown zig- zag areas gigantea aa. Color pattern of brown dots on um- bos maculata B. Shell with well developed radial ribs a. Ribs few, heavy, not over 12, strong- ly striated, with wide interspaces b. Ribs rounded, not over 7 inaequalis bb. Ribs square, 9 to 11 delessertii aa. Ribs 24 to 40, not striated c. Ribs 35-40 multicostata cc. Ribs fewer, not over 30 d. Altitude usually not over 30 mm. e. Hinge evenly rounded zessellata ee. Hinge angulated strigilata dd. Altitude over 30 mm., ribs broader canoa Subgenus Glycymeris s.s. Glycymeris I G/ycy men’s I gigantea Reeve. Pectunculus giganteus Reeve, Conch. Icon., Vol. 1, Pectunculus, February, 1843, species 3, pi. 1, figs. 3a, 3b. “Hab. Guay- mas. Gulf of California. (Found in sandy mud at seven fathoms depth.)” Type Locality : Guaymas, Mexico, in 7 fathoms, sandy mud. Range: Magdalena Bay, Lower Califor- nia, and from Punta Penasco, to Cape San Lucas, Lower California; Acapulco, Mex- ico. Collecting Stations: Mexico: Santa Inez Bay, Gulf of California (145-D-1-3), 4 fathoms, sand. Also on beach and at Monu- ment station ; Ceralbo Island ; Arena Point area, Gulf of California. Description : Shell large, orbicular, thick, finely radially striate, ornamented by waved reddish-brown spots which touch each other and often form a zigzag pattern toward the umbos. There are about 29 or 30 hinge teeth, the central ones smaller than the distal ones. Specimens attaining a height of 100 mm. have been collected at Magdalena Bay, Low- er California. The zigzag reddish-brown color pattern serves to separate this species from G. maculata Broderip whose coloration consists of brown spots. The color pattern also serves to easily separate it from the Mediterranean species G. bimaculata Poli. Distribution: The distribution of Glycy- meris gigantea seems to be very limited. It occurs at Magdalena Bay, Lower Cali- fornia, and is known to occur from Pliocene to Recent in the Gulf of California region. It has been recorded from Acapulco, but it does not appear to be common south of the Gulf of California. Glycymeris I Glycymeris I maculata Broderip. Pectunculus maculatus Broderip, Proc. Zool. Soc. London , August 14, 1832, p. 126. “Hab. in Portu Portrero”. “Found in fine gravel in eleven fathoms of water.” — Reeve, Conch. Icon., Vol. 1, Pectunculus, 1843, species 4, pi. 1, fig. 4. “Hab. Puerto Potrero, Central America. Cuming.” Type Locality : Puerto Potrero, Costa Rica, in 11 fathoms, fine gravel. Range: Magdalena Bay, Lower California, and Punta Penasco, Sonora, Mexico, to Zor- ritos, Peru. Collecting Stations: Mexico: Arena Bank, Gulf of California (136-D-30), 35 fathoms, sand, weed; Ceralbo Channel (137-D-3), 46 1943] Hertlein & Strong: Mollustcs from . the Pacific 151 fathoms, rock; Port Guatulco (195-D-9, 17), 6-7 fathoms, gravel, sand, crushed shell; Tangola-Tangola Bay (196-D-6, 7), 6-7 fathoms, sand, crushed shell; Nicaragua: Gulf of Fonseca, Potosi and Monypenny Point; Corinto (200-D-19), 12-13 fathoms, mangrove leaves; Costa Rica: Piedra Blan- ca Bay (208-D-l, 10), 2-6 fathoms, rocks, sand, algae. Description: The shell of Glycymeris maculata is large, orbicular, and ornament- ed by small chestnut brown spots which occur on the earlier part of the shell. This color pattern serves to separate it from C. gigantea Reeve. Radial striae are present on the exterior of the shell. Distribution: This species occurs in shal- low water from the northern part of the Gulf of California to Peru. Subgenus Tuceta Bolten. Tuceta Bolten, Mus. Boltenianum, 1798, p. 172. Type (here designated) : “T. pectuncu- lus.” [In the synonymy of which Bolten in- cluded Area pectunculus Gmelin, illustrated by Chemnitz, Conchyl.-Cab., Bd. 7, 1784, Tab. 58, figs. 568 and 569. Red Sea]. Tuceta Bolten has by some authors been placed in the synonymy of Glycymeris s.s. So far as we know no type has been desig- nated heretofore for Tuceta and we there- fore designate Tuceta pectunculus, in the synonymy of which Bolten placed Area pec- tunculus Gmelin \_—Arca pectunculus Lin- naeus] illustrated by Chemnitz. This is a strongly ribbed form with an arcuate hinge. Tuceta thus becomes available for strongly ribbed species of Glycymeris similar to the type species. Glycymeris ITucetal multicostata Sowerby. Pectunculus multicostatus Sowerby, Proc. Zool. Soc. London, 1832 (issued March 13, 1833), p. 195. “Hab. in America Meridion- ali.” “Found in coarse sand and gravel, in twelve fathoms water, off the Island of Muerte, in the Bay of Guayaquil.” — Reeve, Conch. Icon., Vol. 1, Pectunculus, 1843, species 26, pi. 5, fig. 26. “Hab. Bay of Guayaquil (found in coarse sand and gravel at the depth of twelve fathoms) ; Cuming.” Type Locality: Off the Island of Muerte, Bay of Guayaquil, Ecuador, in 12 fathoms, coarse sand and gravel. Range: Punta Penasco, Sonora, Mexico, to Guayaquil, Ecuador. Collecting Stations : Mexico : Santa Inez Bay, Gulf of California (143-D-l; 144-D-2, also on beach), 21/2-29 fathoms, mud, crushed shell, rock, weed. Ceralbo Island, Gulf of California; Arena Bank, Gulf of California (136-D-18, 27), 40-50 fathoms, mud, sand, calcareous algae, rock; Port Guatulco; Panama: Bahia Honda; Colom- bia: Gorgona Island. Description : The shell of Glycymeris multicostata is elongately rounded in out- line and is usually ornamented with about 35 to 40 well developed flat topped radiating ribs. The ribs vary in number and in width and in some cases are ornamented by long- itudinal incised lines toward the anterior ventral margin. The ribs increase in num- ber both by bifurcation and intercalation. On the beaks the ribbing often occurs in fascicules. The general color of the shell is gray variegated with chestnut. There is usually some brownish color in the interior of the shell, especially the posterior part. The species described as Pectunculus septentrionalis by Middendorff appears to be a synonym of Glycymeris multicostata. Although Glycymeris septentrionalis has been cited in the literature as occurring in Alaska, we have seen no specimens from that region which could be referred to the species as characterized by Middendorff’s original description and illustrations. Mid- dendorff compared his species to Glycymeris multicostata and to the figure referred to G. inaequalis by Sowerby in Beechy’s voy- age which really represents G. bicolor, a form very close to G. multicostata. The shape, hinge, character of ribbing and color pattern in Middendorff’s figures all conform with those of G. multicostata. The ribbing occuring in fascicules on the beaks is very characteristic of G. multicostata. The type specimen of Middendorff’s species was said to have been collected by Wosnessensky at Ukamok Island (also known as Chirikof Island), near Kodiak, Alaska. Wosnessensky also collected in the Gulf of California, a region where G. multicostata occurs com- monly. It appears then, unfortunately, that Pectunculus septentrionalis must be rele- gated to the synonymy of Glycymeris mul- ticostata. Distribution: This species occurs at many localities from the Gulf of California to Peru and is known to occur from Pliocene to Recent. Glycymeris ITucetal tessellata Sowerby. Pectunculus tessellatus Sowerby, Proc. Zool. Soc. London, 1832 (issued March 13, 1833), p. 196. “Hab. ad littora Columbiae Occidentalism’ “From sandy mud and gravel, in from eight to ten fathoms, at Monte Christe and in the Bay of Xipixapi.” — - Reeve, Conch. Icon., Vol. 1, Pectunculus, 1843, species 29, pi. 6, fig. 29. Original lo- cality record cited. Type Locality: Monte Cristi, Ecuador, in 8 to 10 fathoms, sandy mud and gravel, here designated as type locality. Bay of Xipixapi, Ecuador, also cited originally. Range : Arena Bank, northeast of Cape San Lucas, Lower California, to Xipixapi, Ecuador. Collecting Stations : Mexico : Arena Bank, 152 Zoologica: New York Zoological Society [XXVIII: 19 Gulf of California (136-D-6, 13, 15, 18, 22, 24), 40-50 fathoms, mud, Area conglomer- ate, sand, weed, crushed shell, muddy sand; Gorda Banks, Gulf of California (150-D-8), 40-45 fathoms, muddy sand; Manzanillo (184-D-2), 30 fathoms, gravelly sand; Port Guatulco (195-D-9), 7 fathoms, gray sand, crushed shell; Santa Cruz Bay; Costa Rica: Port Parker (203-D-l, 2, 3, also on beach), 12-15 fathoms, sandy mud, crushed shell, shelly mud; Port Culebra (206-D-l, 2, 3), 14 fathoms, sandy mud; 14 miles S. E. of Judas Point (214-D-1-4), 42-61 fathoms, mud, shell, rocks; Panama: Bahia Honda (222) ; Hannibal Bank. Description: The shell of Glycymeris tessellata is triangularly orbicular, some- what attenuated toward the umbos. The ribs, about 25 in number, are rounded and rather wide. The shell is colored by tessel- lated rich purple spots on a grayish-white background. The largest specimens in this collection assigned to Glycymeris tessellata measure about 33.3 mm. in height. The width of the ribs appears to be somewhat variable. From the descriptions of Sowerby and Reeve there appears to be but little difference between Glycymeris tessellata and strigilata except that the latter has a more angular hinge and usually smaller ligamentary area. It is on those characters that we have separated the two forms but it is not at all certain that these are constant. Some of the forms here referred to tessellata have about 25 ribs, a rounded hinge with about 24 teeth which are strong on the sides but which become very weak in the center of the hinge. In a series of specimens these distinctions appear very much less pronounced and there is a suggestion of gradation from this form to those with more numerous teeth with a rounded angular hinge. Glycymeris pectinata Gmelin from the Caribbean region is a similar species. Distribution: This species occurs fairly commonly from the Gulf of California to Ecuador. It is also known to occur in the Pleistocene. Glycymeris (Tucetal tessellata canoa Pilsbry & Olsson. Glycymeris canoa Pilsbry & Olsson, Proc. Acad. Nat. Sci. Philadelphia, Vol. 93, Sep- tember 9, 1941, p. 54, pi. 13, figs. 2, 2a. “Canoa formation, Punta Blanca”, Ecuador, Pliocene. Type Locality. Canoa formation, Punta Blanca, Ecuador, Pliocene. Range: Known living only from Arena Bank, Gulf of California. Collecting Station: Mexico: Arena Bank, Gulf of California (136-D-5, 22), 33-45 fathoms, sand, weed, mud. Description: This form differs from typi- cal tessellata in the much greater size, ex- tremely broad ribs and in that the color markings are more in the nature of concen- tric zigzag lines rather than in large spots. There are about 24 teeth on the hinge but on some large specimens there are only about 20, and the center of the hinge is nearly smooth. A typical specimen measures : height, 48.5 mm., length, 46 mm., convexity (both valves), 28.5 mm. This subspecies appears to be a giant form of Glycymeris tessellata Sowerby but differs in the much greater size and broad- er and flatter ribs. Judging from the origin- al description and illustrations of Glycy- meris canoa, the specimens from the Gulf of California appear to be identical with those from the Pliocene of Ecuador. Distribution: This subspecies is known living at the present time only at Arena Bank, in the southern end of the Gulf of California. It occurs in the Pliocene of Ecuador and it seems likely that it may be found living at other localities between the Gulf of California and Ecuador. Glycymeris ITucetal tessellata strigilata Sowerby. Pectunculus strigilatus Sowerby, Proc. Zool. Soc. London, 1832 (issued March 13, 1833), p. 196. “Hab. ad Sanctam Elenam.” “Dredged from a depth of six to eight fath- oms in sandy mud.” — Reeve, Conch. Icon., Vol. 1, Pectunculus, 1843, species 31, plate 6, fig. 31. Original locality record cited. Type Locality: Santa Elena, Ecuador, in 6 to 8 fathoms, sandy mud. Range : Manzanillo, Mexico, to Santa Elena, Ecuador. Collecting Stations: Mexico: Manzanillo (184-D-1-2), 25-30 fathoms, sand and grav- elly sand; Costa Rica: Port Parker (203-D- 1, 3), 12-15 fathoms, sandy mud, crushed shell, shelly mud. Description: There appears to be little to separate the form described by Sowerby under the name “Pectunculus strigilatus ” from that described as P . tessellatus. The only character cited for the form Glycy- meris t. strigilata which may separate it from G. tessellata is that given by Reeve who stated “A very solid broadly ribbed shell, so peculiarly contracted towards the umbones that the hinge almost describes an angle.” In the present collection, a series of speci- mens dredged off Manzanillo, Mexico, seems to answer that description. The largest specimen measures 28.5 mm. from beak to base. There are about 24 to 25 ribs which in some shells are rounded but in others flattened. There are about 30 rather fine teeth which near the center of the hinge become shorter and weaker. It is uncertain whether this will prove to be a constant 1943] Hertlein & Strong: Mollusks from the Pacific 153 character because there appears to be a gradation to forms with a rounded hinge with stronger and fewer hinge teeth. The triangular ligamentary area is generally smaller than that on typical Glycymeris tessellata. The same sort of angularity of the hinge of the form G. t. strigilata is sometimes no- ticed on specimens of G. pectinata Gmelin. Distribution: This subspecies occurs from off western Mexico to Ecuador along with Glycymeris tessellata. Subgenus Axinactis Morch. Axinactis Morch, Malakozool. Blatter, Bd. 7, 1861, p. 203. Species in original list: “Axinaea ( Axinactis ) inaequalis Sow.” and “Axinaea ( Axinactis ) assimilis Sow.?” Type (here designated) : Axinaea ( Axin- actis) inaequalis Sowerby. Shell roundly triangular, thick, beaks opisthogyrate ; ribs wide and separated by narrower interspaces; ribs, interspaces and margins covered by fine radial ribs which are separated by fine incised lines; cardinal area forming an elongate asymmetric tri- angle and ornamented by six or seven in- cised lines which run diagonally to the base from an impressed line bordering the pos- terior part of the area ; only a small portion of the ligament occurs in front of the beaks; hinge with two series of chevron-shaped teeth, the two series separated by an im- pressed line which is almost directly in line with the beaks ; the posterior portion of the anterior series small and nearly vertical; on perfectly preserved specimens a fine row of denticles is present along the rather straight portion of the posterior margin ; inner margin of valves fluted. Glycymeris I Axinactisl delessertii Reeve. Pectunculus delessertii Reeve, Conch. Icon., Vol 1, Pectunculus, December, 1843, species 52, pi. 9, fig. 52. “Hab. — ?” Glycymeris delesserti Reeve, Pilsbry & Lowe, Nautilus, Vol. 47, No. 3, 1934, p. 85. Maria Madre Island, Tres Marias Islands, Mexico. Type Locality: No locality originally cited. Maria Madre Island, Tres Marias Islands, Mexico, here designated as type locality. Range: Mazatlan, Mexico, to Panama. Collecting Station: Mexico: Port Guatul- co, on beach. Description: A worn left valve of Glycy- meris delessertii is present in the collection from Port Guatulco, Mexico. The shell of this species is ornamented by about 9 to 11 squarish ribs which are separated by inter- spaces a little narrower. The ribs are orna- mented by riblets separated by longitudin- ally incised lines varying in depth and number. These incised lines continue over the interspaces on the anterior and posterior portions of the shell but in the median por- tion the interspaces are only finely striated. The more numerous, narrower, square ribs easily separate this species from Glycy- meris inaequalis Sowerby. Distribution: This species has been col- lected at Mazatlan, the Tres Marias Islands, Port Guatulco, Mexico, and at Panama. It is also known to occur in the Pleistocene of Oaxaca, Mexico. Glycymeris I Axinactisl inaequalis Sowerby. Pectunculus inaequalis Sowerby, Proc. Zool. Soc. London, 1832 (issued March 13, 1833), p. 196. ‘‘Hab. ad Panamam et Real Llejos.” “Found in sandy mud in ten fath- oms.”— Reeve, Conch. Icon., Vol. 1, Pectun- culus, 1843, species 16, pi. 4, fig. 16. Orig- inal locality record cited. Type Locality : Panama City, Panama, in 10 fathoms, sandy mud, here designated as type locality. Real Llejos [near Corinto]^ Nicaragua, also cited originally. Range : San Marcos Island, Gulf of Cali- fornia, to Bayover, Peru. Collecting Stations: Nicaragua: Corinto (200-D-19), 12-13 fathoms, mangrove leaves ; Costa Rica : Piedra Blanca Bay (208-D-l, 10), 2-6 fathoms, rocks, sand, algae. Description: This species is readily rec- ognized by the presence of about 6 rather broad, rounded ribs which on the anterior and posterior portions of the shell give way to one or two small ribs. The ribs are wider than the interspaces and both are covered by small somewhat irregular rib- lets which are due to the presence of in- cised lines. The color of the exterior of the shell is a whitish background with irregu- lar concentric bands of dark brown color. Glycymeris inaequalis can be separated from G. delessertii by the presence of the fewer, broader, rounded ribs. The species described as Pectunculus assimilis by Sower- by can be relegated to the synonymy of Gly- cymeris inaequalis. Distribution: This species has been re- corded from various localities from the Gulf of California to Peru. It is also known to occur in the Quaternary of Ecuador. Family Arcidae.* Key to the Genera and Subgenera of the Arcidae. A. Beaks curved toward the anterior Area a. Inner margin of valves smooth ♦For a classification of the Arcidae, see Reinhart, P. W., Classification of the Pelecypod Family Arcidae,” Bui Mus. Roy. Hist. Nat. Belgique, Tome 11, No. 13 1935 68 pp., pis. 1-5. nnutner important paper published recently by Reinhart is entitled “Mesozoic and Cenozotc Arcidae from the Pacific Slope of North America.” Geol. Soc. America. Special Papers No. 47. June 16, 1943, pp. XI, 1-117, pis. 1-15 3 fiES in text, 3 tables. 154 Zoologica: New York Zoological Society [XXVIII: 19 b. Hinge with a continuous series of teeth c. Ligamental area extremely wide and almost flat; posterior end more or less expanded Area s.s. cc. Ligamental area narrow, V- shaped d. Ligamental area extending equally on each side of the umbos Barbatia dd. Ligamental area extending mainly posterior to the umbos e. Muscle scars prominent, elevated ; anterior and posterior ribs not larger than the others Acar ee. Muscle scars not promin- ent; large anterior and posterior ribs Calloarca bb. Hinge with an edentulous pit or gape in the middle Arcopsis aa. Inner margin of valves crenulated f. Right and left valves equal in size and sculpture g. Ligamental area extending equally on each side of the umbos h. Central teeth perpendicular to the hinge line Anadara hh. Central teeth diverging Larkinia gg. Ligamental area extending mainly posterior to the beaks Lunarca ft. Left valve the larger, overlapping the right i. Posterior dorsal margin with a flattened auriculation .Cara ii. Posterior dorsal margin with- out a flattened auriculation j. Sculpture of the two valves similar or only slightly dif- ferent Scapharca jj. Sculpture of the two valves strongly discrepant Cunearca B. Beaks curved toward the posterior Noetia a. Shell trigonal Noetia s.s. aa. Shell elongate b. Posterior end strongly expanded Sheldonella bb. Posterior end not expanded Eontia Genus Area Linnaeus. Subgenus Area s.s. Key to the Species of Area s.s. A. Shell expanded posteriorly; posterior margin notched pacifica B. Shell not expanded or notched posteriorly but obliquely truncated or rounded mutabilis Area l Area) fernandezensis Hertlein & Strong, nom. nov. Area angulata King & Broderip, Zool. Jour., Vol. 5, July, 1832, p. 336. “Habitat ad Juan Fernardez.” “This shell was dredged up from 80 fathoms water in the offing of Cumberland Bay, at Juan Fernandez; it was attached to a branch of coral.” — Stem- pell, Zool. Jahrb., Suppl. Bd. 5, December 20, 1899, Fauna Chilensis, Bd. 2, p. 219, pi. 12, figs. 1-9. Juan Fernandez Island, on rocks along the coast and in 20 to 40 fath- oms. Not Area angulata Meuschen, Mus. Gev- er„ 1787, p. 426. A study of the west American species of Arcidae revealed that the specific name angulata of King and Broderip had been used earlier for a species of Area by Meu- schen. The new name Area fernandezensis is here proposed for the species from Juan Fernandez Island. Area (Area) mutabilis Sowerby. Byssoarca mutabilis Sowerby, Proc. Zool. Soc. London, May 17, 1833, p. 17. “Hab. in Colombia occidentali.” “Found under stones at the Isle of Plata”. Area mutabilis Sowerby, Reeve, Conch. Icon., Vol. 2, Area, 1844, species 85, pi. 13, fig. 85. Original locality record cited. Type Locality: Island of La Plata, Ecua- dor, under stones. Range : Magdalena Bay, Lower California, and the Gulf of California, to Guayaquil, Ecuador. Collecting Stations: Mexico: Arena Bank, Gulf of California (136-D-21), 45 fathoms, mud; Port Guatulco; Sihuatanejo; Costa Rica: Port Parker; Port Culebra; Piedra Blanca; Isla Cedro. Description: The four to six ribs on the posterior slope of the shell of Area mutabilis are somewhat coarser than the others and in fresh shells these are characteristically dark in color. The ligamentary area is con- cave. Area mutabilis is strikingly like A. im- bricata Bruguiere of the western Pacific. Lamy considered the two forms to differ only subspecifically. Area santamariensis Reinhart from the Pliocene of California is similar to A. mutabilis but is said to differ in ornamentation and in the shape of the ligamental area. Distribution: This species is often found under rocks at low tide from Magdalena Bay, Lower California, and the Gulf of California, to Ecuador. 1943] Hertlein & Stro7ig: Mollusks from the Pacific 155 Area I Area I pacifica Sowerby. Byssoarca pacifica Sowerby, Proc. Zool. Soc. London, May 17, 1833, p. 17. “Hab. ad Sanctam Elenam.” “Found on rocky ground, in from six to eighteen fathoms, adhering to each other in large bunches.” Area pacifica Sowerby, Reeve, Conch. Icon., Vol. 2, Area, 1844, species 75, pi. 11, fig. 75. Original locality record cited. Type Locality. Santa Elena, Ecuador, in 6-18 fathoms, on rocky ground. Range: Scammon Lagoon, Lower Califor- nia, to Paita, Peru, and the Galapagos Islands. Collecting Stations: Mexico: Santa Inez Bay, Gulf of California (145-D-l, 3), 4-13 fathoms, sand; Santa Inez Point, Lower California; Arena Bank, Gulf of California (136-D-l, 17), 45 fathoms, mud and Area conglomerate; Arena Point area, Lower California; Gorda Banks, Gulf of California (150-D-7-8, 16), 20-75 fathoms, sand, mud- dy sand, rock, calcareous algae; Cape San Lucas, Lower California; Passavera Island, Chamela Bay; Nicaragua: Corinto (200-D- 19), 12-13 fathoms, mangrove leaves on bottom; Costa Rica: Port Parker; Piedra Blanca (208-D-1-10), 2-6 fathoms, rocks algae, sand; Panama: Hannibal Bank; Bahia Honda; Colombia: Gorgona Island. Description: The shell of Area pacifica is expanded and notched posteriorly. It is ornamented by radiating ribs and is colored by narrow, rounded V-shaped brown bands. It is more expanded posteriorly than A. occidentalis Philippi, a species living in the Caribbean region. Distribution: This species often occurs from Scammon Lagoon, Lower California, to Peru, under rocks at very low tide and in shallow water. It also occurs in the Plio- cene and Pleistocene of Lower California. Subgenus Acar Gray. Area I Acar) gradata Broderip & Sowerby. Area gradata Broderip & Sowerby, Zool. Jour., Vol. 4, January, 1829, p. 365. “Hab. ad littora Oceani Pacifici.” Also “From Mazatlan.” Barbatia (Acar) gradata Sowerby, Rein- hart, Trans. San Diego Soc. Nat. Hist., Vol. 9, No. 10, 1939, pp. 39-43, pi. 3, figs, la, lb, 5a, 5b, 6a, 6b, 6c, 6d, 6e. Mazatlan, Mexico, and Taboga Island, Panama. Type Locality: Pacific Ocean, littoral. Also Mazatlan. [Mazatlan, Mexico, stated to be the type locality by Reinhart and accepted as such by the present authors]. Range: Point Abreojos, Lower California, to Punta Penasco, Sonora, Mexico, and south to Negritos, Peru, and the Galapagos Islands. Collecting Stations : Mexico : Cape San Lucas, Lower California; Sulphur Bay, Clarion Island; Banderas Bay; Port Guatu- lco (195-D-9, 15), 1.5 to 7 fathoms, gray sand, crushed shell, coral; Tangola-Tangola Bay (196-D-14-15, and beach), 5 fathoms, crushed shell; Nicaragua: Corinto (200-D- 19), 12-13 fathoms, mangrove leaves on bot- tom; Costa Rica: Port Parker; Ballena Bay; Uvita Bay. Description: Shell moderately thick for its size, ornamented exteriorly by reticulate sculpture. On some specimens granules are developed at the point where the concentric crosses the radial sculpture. At the time of of the original description of the species Broderip & Sowerby stated, “This elabor- ately ornamented shell looks at first sight like a piece of Chinese carving.” Large specimens attain a length of about 30 mm. The first published figure of Area grad- ata is in the Zoology of Beechy’s Voyage, 1839, p. 152, pi. 43, fig. 1. The part of that work here involved was prepared by Sower- by so the specimen may be considered to be authentic in the absence of information to the contrary. This figure shows a coarsely sculptured shell. This interpretation of the species is identical with that of Reinhart who has discussed and illustrated the holo- type. Area gradata closely resembles A. reticu- lata Gmelin, a Caribbean species, and there are somewhat similar species in the Western Pacific. Distribution: This species occurs from Lower California to Peru and the Galapagos Islands. It is often found under rocks be- tween tides. Subgenus Anadara Gray. Key to the Species of Anadara. A. Anterior dorsal margin acutely pointed ; about 30 ribs biangulata B. Anterior dorsal margin forming nearly a right angle, or rounded a. Anterior ribs deeply grooved, finely nodulous formosa aa. Anterior ribs not grooved or only faintly so b. Ribs 35-37 c. Ribs tuberculated, shell high tuberculosa cc. Ribs smooth, shell elongate mazatlanica bb. Ribs 40-44 similis Area I Anadara I biangulata Sowerby. Plate I, Figure 3. Area biangulata Sowerby, Proc. Zool. Soc. London, May 17, 1833, p. 21. “Hab. ad lit- tora Columbiae Occidentalis. (Atacamas),” “A single specimen was dredged at a depth of seven fathoms.” Area gordita Lowe, Trans. San Diego 156 Zoologica: New York Zoological Society [XXVIII: 19 Soc. Nat. Hist., Vol. 8, No. 6, March 21, 1935, p. 16, pi. 1, fig. 1. “Acapulco, 20 fath- oms” (type). Also “Guaymas, 20 fathoms” and “off West Mexico.” — R. H. Palmer & Hertlein, Bull. South. Calif. Acad. Sci., Vol. 35, pt. 2, May-August [issued September 10], 1936, p. 70, pi. 19, figs. 1 and 4. Pleistocene of Oaxaca, Mexico. Type Locality. Atacames, Western Ecua- dor, dredged in 7 fathoms. Of Area gordita, Acapulco, Mexico, in 20 fathoms. Range: Guaymas, Sonora, Mexico, to Paita, Peru, and the Galapagos Islands. Collecting Stations : Mexico : Santa Inez Bay, Gulf of California (143-D-l, 5), 18-29 fathoms, sand, mud, crushed shells, weeds; Gorda Banks, Gulf of California (150-D-8), 40-45 fathoms, muddy sand ; Manzanillo Bay (184-D-2), 30 fathoms, gravelly sand; Costa Rica: Port Parker (203-D-l, 3), 12-15 fath- oms, sandy mud, shelly mud, and crushed shells; 14 miles southeast of Judas Point (214-D-l to 4), 42-61 fathoms, mud, shell, rocks; Panama: Gulf of Chiriqui (221-D-l to 5), 35-40 fathoms, sandy mud. Description: Area biangulata was origin- ally described by Sowerby as oblong, ventri- cose, white, radiately ribbed, covered by a brown periostracum, dorsal margin anteri- orly acute, posteriorly obtusely angulated; anterior end shorter and higher; posteriorly subacuminate; ligamentary area elongate, flat, wider anteriorly. Length, 2.0; width, 1.2; height, 1.3 poll. That description exactly applies to the shell described by Lowe as Area gordita. The shell is sharply pointed anteriorly and is ornamented by about 30 ribs. Palmer and Hertlein have pointed out the resemblance of this form to Miocene species of the Carib- bean region. Distribution: This species is now known to occur from the Gulf of California to Peru. It is also known to occur in the Pleistocene of Oaxaca, Mexico. Area I Anadara ) formosa Sowerby. Area formosa Sowerby, Proc. Zool. Soc. London, May 17, 1833, p. 20. “Hab. in America Centrali. (Gulf of Tehuantepec).” • — Reeve, Conch. Icon., Vol. 2, Area, 1843, species 10, pi. 2, fig. 10. “Hab. Gulf of Tehuantepec, Mexico (found in sandy mud at the depth of from ten to twelve fath- oms) ; Cuming.” Type Locality: Gulf of Tehuantepec, Mexico. Range : Cedros Island, off Lower Cali- fornia, to Paita, Peru. Collecting Stations: Mexico: East of Ce- dros Island (126-D-6), 45 fathoms, mud; Tangola-Tangola Bay (196-D-6-7), 6-7 fath- oms, sand and crushed shell; El Salvador: La Union; Costa Rica: Culebra Bay; Piedra Blanca. Description: Shell elongate, posterior dor- sal area rather broad and somewhat concave, ornamented by about 35 to 38 flat-topped ribs. The posterior ribs are wider, the an- terior ones finely nodulous and divided by a groove. Two specimens referable perhaps to this species were dredged about 3 miles east of Cedros Island. A specimen from La Union, Gulf of Fonseca, El Salvador, measures 112.3 mm. in length. Distribution: This species is known to occur from Cedros Island, Lower California, to Peru, but is not very abundant in any of the collections which we have studied. Area I Anadara I mazatlanica Hertlein & Strong, sp. nov. Plate I, Figures 1 and 4. Shell elongately-oval, moderately convex, equivalve, white under a brown epidermis which is thin and rubbed near the beaks, thicker toward the margin and fringed in the interspaces between the ribs; hinge line straight with beaks rising above it at about the anterior third ; anterior end forming almost a right angle at the end of the hinge line, then slightly rounded; posterior end forming an obtuse angle at the end of the hinge line, roundly pointed at the junc- tion with the basal margin which is slightly convex and without a gape; sculpture of 35 broad, smooth, flat-topped ribs with narrow interspaces; ligamental area fairly broad, with 4 incised lines angulated under the beaks in such a manner as to leave a smooth, flattened shelf along the anterior dorsal margin ; hinge with about 60 narrow teeth extending about the full length of the hinge line, small, vertical and closely spaced in the middle, larger, wider spaced and slightly oblique towards the ends. The type meas- ures: longitudinal diameter, 62 mm., vertical diameter, 36.3 mm.; convexity of the two valves, 31 mm. Holotype, and paratypes (Calif. Acad. Sci. Paleo. Type Coll.), dredged at Station No. 153-D-2, in approximately Lat. 23°06'N., Long. 106°47'W., 19 miles west of Mazat- lan, Mexico. Twenty additional specimens were dredged in the same locality. In the same general locality six specimens were dredged at Station 153-D-3. Three odd valves were dredged at Station 155-D-l, in 56 fathoms thirteen miles west of Mazatlan, Gulf of California, Lat. 23°12'N., Long. 106°40'W., in 56 fathoms in mud. One speci- men was dredged at 143-D-5, Lat. 26°54'N., Long. 111°53'W., Santa Inez Bay, Lower California, in 18 fathoms in sand. Range: Santa Inez Bay, east coast of Lower California, to off Mazatlan, Mexico. In many ways this species resembles Area formosa Sowerby but the anterior end is shorter and more sharply angulated at the 1943] Hertlein & Strong: Mollusks from the Pacific 157 hinge line. Also the ribs show no indication of a median groove. The ventral margin of the new species is rounded while that of Area formosa is nearly straight. The new species bears a resemblance to Area con- cinna but it is a higher shell, the anterior end is more obliquely rounded and the basal margin is more swollen and rounded. The new species also bears a close resem- blance to “Barbatia ( Dihivarca ) halidonata oresta” Woodring2 but seems to have wider ribs than the east coast fossil. Area spring - valensis Vokes3 from the Miocene of Trini- dad is also somewhat similar in its general features. Area secticostata Reeve is somewhat simi- lar to A. mazatlamca. Tomlin cited “Ana- dara secticostata” Reeve4 from Coiba Island, Panama. That record can probably be re- ferred to Area formosa or to some similar but different West American species. Area lAnadara) reinharti Lowe. Area ( Anadara ) reinharti Lowe, Trans. San Diego Soc. Nat. Hist., Vol. 8, No. 6, March 21, 1935, p. 16, pi. 1, figs. 3a, 3b, 3c. “Guaymas, 20 fathoms” (type). P. 27, Pun- ta Penasco, Sonora, Mexico, dredged in 10 fathoms. Type Locality: Guaymas, Mexico, in 20 fathoms. Range: Punta Penasco, Sonora, Mexico, to Bahia Honda, Panama. Collecting Stations: Mexico: Manzanillo (184-D-2), 30 fathoms, gravelly sand; Nica- ragua: Corinto (200-D-19), 12-13 fathoms, mangrove leaves on bottom; Costa Rica: Port Parker (203-D-l,-3) , 12-15 fathoms, sandy mud, shelly mud, algae; Panama: Gulf of Chiriqui (221-D-l), 35 fathoms, sandy mud; Bahia Honda. Description: The specimens here referred to Area reinharti appear to be identical with a paratype No. 4697a of that species in the collection of the California Academy of Sci- ences. The present specimens appear to be a little longer than the type of A. reinharti as figured by Lowe. The type was described as possessing about 25 ribs. The specimens in the present collection have about 27 ribs which would appear to fall within the varia- tion of the species. Some specimens are much thinner than others and the ribs are smoother but otherwise they appear to be identical with typical A. reinharti. Area reinharti bears a considerable re- semblance to young specimens of A. multi- 2 Barbatia ( Diluvarca ) halidonata oresta Woodring, Carnegie Inst. Washington. Publ. 366, May, 1925, p. 43, pi. 4, figs. 5 and 6. Bowden, Jamaica, Miocene. 3 Anadara ( Anadara ) spring valensis Vokes, Amer. Mus. Novitat., No. 988, May 16, 1938, p. 10, fig. 3. Springvale, Trinidad. Upper Miocene. 4 Anadara secticostata Reeve, Tomlin, Jour. Conch., Vol. 18, No. 7, May, 1928, p. 189. “Coiba, valves dredged in 10-12f.” See Area secticostata Reeve, Conch. Icon., Vol. 2, Area, February, 1844, sp. 38, pi. 6, fig. 38. “Hab. — ?” costata Sowerby. Both are slightly inequi- valve, the left valve over-lapping the right. The shells of A. reinharti are more elongate and have about 10 less ribs. Distribution: This species is known to occur from the Gulf of California to Pana- ma. In some localities it is fairly abundant in 10 to 35 fathoms, especially off Manza- nillo, Mexico, and off Port Parker, Costa Rica. Area t Anadara I similis C. B. Adams. Plate I, Figures 2 and 5. Area similis C. B. Adams, Ann. Lyceum Nat. Hist. New York, Vol. 5, 1852, p. 485 (separate p. 261). “Panama.” Type Locality: Panama. Range: Corinto, Nicaragua, to Panama. Collecting Stations: Nicaragua: Corinto (200-D-19), 12-13 fathoms, mangrove leaves on bottom; Costa Rica: Port Culebra; Pun- tarenas Lagoon. Description: The shell of Area similis was described by Adams as having 40 to 44 ribs. The specimens which we identify as this species have about 40 ribs. They differ from Area tuberculosa Sowerby in that Adams’s species is not as high in proportion to the length, the umbonal ridges are rounded and the dorsal margin is rounded at each end. The species Area tuberculosa and A. simi- lis occur together and they are quite similar in general featui'es but they appear to be separable at least in adult forms. Carpenter considered A. similis to be “scarcely a va- riety of A. tuberculosa.” Distribution: The present record of this species from Corinto, Nicaragua, is an ex- tension northward of the range. It is also known to occur off Costa Rica and Panama. Area lAnadara I tuberculosa Sowerby. Area tuberculosa Sowerby, Proc. Zool. Soc. London, May 17, 1833, p. 19. “Hab. ad Real Lie j os.” “Found at low water at the roots of the mangrove trees.” — Reeve, Conch. Icon., Vol. 2, Area, 1844, species 18, pi. 3, fig. 18. Original locality record cited. Type Locality: Real Llejos [near Corin- to], Nicaragua, at low water at the roots of mangrove trees. Range: Ballenas Lagoon, Lower Califor- nia, and the Gulf of California, to Tumbez, Peru. Collecting Stations: Costa Rica: Port Culebra ; Culebra Bay ; Ballena Bay ; Golfito Bay; 1 mile south of Golfito Bay; Panama: Isla Partida; Colombia: Gorgona Island. Description: Shell large, elongately ovate, moderately inflated, moderately thick, and ornamented by about 33 to 37 ribs which are tuberculate toward the margin and espe- cially so anteriorly. The dorsal margin of the shell is usually somewhat angulated at either end. 158 Zoologica: New York Zoological Society [XXVIII: 19 Compared to Area similis C. B. Adams, A. tuberculosa is higher in proportion to the length, more angulated at the ends of the dorsal margin, and the posterior um- bonal area is more angular. This species is commonly used for food along the west coast of Central America. At La Union, El Salvador, these shells are known as “coriles” and in Nicaragua as “tuchia.” It is the “concha prieta” of Peru- vian fishermen. Distribution: This species occurs com- monly from the Gulf of California to Peru. It often occurs abundantly in soft mud of mangrove swamps. It has been cited as oc- curring in archaeologic ruins and kitchen middens of southwestern United States. Subgenus Arcopsis von Koenen. Area lArcopsisI solida Sowerby. Byssoarca solida Sowerby, Proc. Zool. Soc. London, May 17, 1833, p. 18. “Hab. ad Pay- tarn, Peruviae.” “Found under stones.” Area solida Sowerby, Reeve, Conch. Icon., Vol. 2, Area, 1844, species 106, pi. 16, fig. 106. Original locality record cited. Type Locality : Paita, Peru, under stones. Range: Asuncion Island. Lower Califor- nia, Mexico, and the Gulf of California, to Paita, Peru, and the Galapagos Islands. Collecting Stations: Nicaragua: Corinto (200-D-19), 12-13 fathoms, mangrove leaves on bottom ; Costa Rica : Port Parker (203- D-3), 12 fathoms, shelly mud. Description: Shell small, somewhat in- flated, subquadrate, with many fine radial ribs. It may be distinguished from other similar shells by the character of the liga- ment which is restricted to a small triangu- lar pit beneath the beaks and by the lack of teeth in the central part of the hinge. Area solida is very similar to A. afra Gmelin, an African species. Lamy consid- ered the two forms to be only subspecifically different. Maury has mentioned a similarity be- tween A. solida and A. adamsi (Shuttle- worth) Smith from the upper Miocene and Pliocene of Trinidad, and Recent from North Carolina to Brazil. Distribution: Area solida occurs common- ly from the Gulf of California to Peru, under rocks between tides. Subgenus Lunarca Gray. Area (Lunarca) vespertina Morch. Plate I, Figures 6 and 7. Area ( Argina ) vespertina Morch, Mala- kozool. Blatter, Bd. 7, 1861, p. 204. “Realejc in Cribrina obtecta valv. sinistra fracta, an juvenilis?.” — - Kobelt, Martini-Chemnitz Conchyl.-Cab., Bd. 8, Abt. 2, Area, 1891, p. 225. Original record cited. Type Locality: Realejo [near Corinto], Nicaragua. Range: Mazatlan, Sinaloa, Mexico, to Cor- into, Nicaragua. Collecting Station: Nicaragua: Corinto (200-D-17, 19), 7-13 fathoms, sand, man- grove leaves. Description: “T. alba quadrangularis postice lata, fere perpendiculariter recte truncata, angulo inferiore rotundato ; margo dorsalis rectus, ventralis antice leviter ad- scendens; costis circiter 38 approximate, unde margine interno prof unde sulcato ; um- bonis coerulescentibus, radiis obscuris tri- bus quorum posticis subconfluentibus. Long. 7 3/4, alt. 6 1/4 mill.” (Original descrip- tion). Morch compared Area vespertina to Area obliqua Reeve, a species from West Africa which it somewhat resembles in shape. The shell described by Morch was appar- ently a juvenile specimen. Specimens studied by us from Corinto, Nicaragua, and from off Mazatlan, Mexico, seem to be referable to Area vespertina. Small specimens agree closely with Morch’s description. Large specimens are much more elongate in pro- portion to the height of the shell. There are from 36 to 38 ribs which are medially sul- cated toward the ventral margin. The color of the shell is white with a black spot on the beak and the shell is covered by a dark brown periostracum. Area brevifrons bucaruana Maury & Shel- don is very similar but is more quadrate in shape, thicker, and with about 33-35 ribs. The three forms, Area brevifrons, A. brevifrons bucaruana and A. vespertina, ap- pear to be closely related, differing in the number of ribs. The Peruvian form brevi- frons possesses 22-23 ribs, bucaruana from Panama with 33-35 ribs, and vespertina from Central America and Mexico with 36- 38 ribs. The species described as Area ( Barbatia ?) melanoderma by Pilsbry & Lowe appears to be referable to Area brevi- frons bucaruana. Distribution: This rare species has been found only occasionally from Mazatlan, Mex- ico, to Corinto, Nicaragua. Subgenus Barbatia Gray. Area (Barbatia ) resveana d’Orbigny. Area helbingii Bruguiere, Reeve, Conch. Icon., Vol. 2, Area, April, 1844, species 90, pi. 14, fig. 90. “Hab. St. Elena and Monte Christi, West Columbia, and island of Cor- rigidor, Philippines (found under stones at low water) ; Cuming.” Area reeveana d’Orbigny, Voy. Amer. Merid., Vol. 5, 1846, p. 635. “a ete rencon- tree a Payta (Perou) par M. Fontaine, et a Monte-Cristi (republique de 1’ Equateur), par M. Cuming.” [Not the record Corrigi- dor, Philippine Islands], New name for Area 1943] Hertlein & Strong: Mollusks from the Pacific 159 helbingii Bruguiere, Reeve, 1844, pi. 14, fig. 90, not Area helbingii Bruguiere, 1792. Type Locality : Santa Elena, Ecuador, here designated as type locality. Monte Cristi, Ecuador, also cited originally. Range : Manuela Lagoon, Lower Cali- fornia, and Punta Penasco, Sonora, Mexico, to Zorritos, Peru. Collection Stations: Mexico: Santa Inez Bay, Gulf of California; Arena Bank, Gulf of California (136-D-4), 55 fathoms, mud; Costa Rica: Ballena Bay. Description: Shell large, variable in shape, attaining a length of 65 mm., oblong, ‘often very sinuate at the ventral margin, moderately thin for the group, beaks often nearly central ; sculpture of radiating threads crossed by concentric lines which, where they cross the radials, cause beading, especially on the center of the valves, ribs usually coarser along the posterior dorsal margin ; hinge arcuate, with the central teeth fine and closely-set, the distal teeth are larger, longer, and oblique; ligament occupying the entire cardinal area; diamond- shaped grooves on the area are closely set. The sculpture of Area reeveana is in its general features similar to that of Area Candida Gmelin of the Caribbean region, and Area velata Sowerby and A. decussata from the Indo-Pacific region. Area platensis Philippi from the Tertiary of Patagonia is a related species as is A. subhelbingi d’Or- bigny from the Miocene of France. Distribution: This species occurs com- monly from the Gulf of California to Peru, and is found attached to the under surfaces of rocks at low tide. Subgenus Calloarca Gray. Area ICalloarcal alternata Sowerby. Byssoarca alternata Sowerby, Proc. Zool. Soc. London, May 17, 1833, p. 17. “Hab. in Columbia occidentali.” “Found attached to stones, on a rocky bottom, in twelve fath- oms.” Area alternata Sowerby, Reeve, Conch. Icon., Vol. 2, Area, 1844, species 88, pi. 13, fig. 88. Original locality record cited. Type Locality: Western Colombia, in 12 fathoms, on rocky bottom. Range: Punta Penasco, Sonora, Mexico, to Ecuador. Collecting Stations: Mexico: Santa Inez Bay, Gulf of California (145-D-l,-3, and shore), 4-13 fathoms, sand; Nicaragua; Costa Rica: Port Parker (203-D-l, 3), 12- 15 fathoms, sandy mud, crushed shell, shelly mud. Description: Shell thin, elongate, some- what contracted medially and with a sharply carinated posterior umbonal slope. It is ornamented by 26 to 28 ribs and those pos- terior to the carinated portion, the three on the anterior end of the shell are coarse and crenulated. On the medial portion of the shell the ribs are flat, smooth, close-set and divided down the middle by a fine in- cised line. The cardinal area is long and narrow, and wider in front of the beaks. The ligament occupies only the posterior part of the area. Distribution: This is not a common species but is known to occur from the Gulf of California to Ecuador. It is also known to occur in the Pleistocene of Magdalena Bay, Lower California. Subgenus Cara Gray. Area I Carol emarginata Sowerby. Area emarginata Sowerby, Proc. Zool. Soc. London, May 17, 1833, p. 20. “Hab. ad littora Maris Pacifici.” “From Atacamas, Real Llejos, Xipixapi, Panama, and the Gulf of California.”- — Reeve, Conch. Icon., Vol. 2, Area, 1844, species 26, pi. 4, fig. 26. Original locality records cited. Type Locality,: Atacames, Ecuador, here designated as type locality. Pacific Ocean littoral, Xipixapi, Ecuador, Panama, and the Gulf of California, also cited originally. Range: Gulf of California, to Guayaquil, Ecuador. Collecting Station: Nicaragua: Corinto (200-D-ll, 19), 8-13 fathoms, mangrove leaves on bottom. Description: Shell thin, elongate, anterior end very short, posterior end produced and with a distinct notch just below the hinge line in the upper posterior margin ; the color is white with usually a black ray on the sulcated beaks; ribs 28 to 30, flat and close- ly-set, divided by very narrow interspaces; anterior ribs more or less medially grooved and crenulate, the ribs over the umbonal ridge are wider ; cardinal area long, narrow, and somewhat wider anteriorly; teeth very fine on the center of the hinge but longer and larger distally, especially toward the posterior end. The shell attains a length of at least 50 mm. This species is somewhat similar to Area esperanza Maury from the Miocene of Trini- dad. Distribution: This species is found oc- casionally from the Gulf of California to Ecuador. It is also known to occur in the Pleistocene of Magdalena Bay, Lower Cali- fornia. Subgenus Cunearca Gray. Key to the Species of Cunearca. A. Shell with 30 or more ribs a. Shell thin, posteriorly produced aequatorialis aa. Shell thick, only slightly posteriorly produced bifrons B. Shell with less than 30 ribs 160 Zoologica: New York Zoological Society [XXVIII: 19 a. Cardinal area with chevron-shaped grooves; 26-28 ribs esmeralda aa. Cardinal area without chevron- shaped grooves b. Beaks and apex of area central, 28 ribs perlabiata bb. Beaks and apex of area anterior, about 22-23 ribs nux Area I Cunearca) aequatorialis d’Ol'bigny. Area ovata Reeve, Conch. Icon., Vol. 2, Area, February, 1844, species 49, pi. 8, fig. 49. “Hab. St. Elena, South America (found in sandy mud at the depth of from six to eight fathoms) ; Cuming.” Area aequatorialis d’Orbigny, Voy. Amer. Merid., Vol. 5, 1846, p. 636. Santa Elena, Ecuador. New name for Area ovata Reeve, not Area ovata Gmelin, 1791. Type Locality : Santa Elena, Ecuador, in 6 to 8 fathoms, sandy mud. Range : Mazatlan, Mexico, to Zorritos, Peru, and the Galapagos Islands. Collecting Stations: Guatemala: 7 miles west of Champerico (197-D-1-2), 14 fath- oms, mud; El Salvador: La Libertad (198- D-l-2), 13-14 fathoms, mud; Nicaragua: Corinto (200-D-10, 19), 7-13 fathoms, man- grove leaves on bottom; Panama: Gulf of Chiriqui (221-D-l,-5) , 35-40 fathoms, sandy mud. Description: The shell of this interesting species is moderately thin and inequivalved. The color is white but often stained with reddish-brown. The ribs are about 30 to 32 in number, those on the left valve are broad- er than those on the right. The ribs on the anterior area are coarsely nodose, those on the posterior area are flat, smooth, and less prominent. Area aequatorialis differs from A. bifrons Carpenter, in the smaller, thinner shell which is more produced posteriorly and ornamented with more coarsely nodose ribs anteriorly. Distribution: This species is known to occur from Mazatlan, Mexico, to Peru. Specimens dredged by the Zaca Expeditions were taken in depths from 7 to 40 fathoms. Area I Cunearca I bifrons Carpenter. Area brasiliana Lamarck, Reeve, Conch. Icon., Vol. 2, Area, 1844, species 17, pi. 3, fig. 17. “San Bias, Bay of California (found on the sands) ; Cuming.” [Not the record “Rio Janeiro, coast of Brazil”]. Area bifrons Carpenter, Cat. Mazatlan Shells, February, 1856, p. 134. “Mazatlan,” Mexico. Area ( Cunearca ) bifrons Carpenter, Maury, Palaeontogr. Amer., Vol. 1, No. 4, 1922, p. 197 (35), pi. 31 (3), fig. 12. Near the mouth of Rio Chepo, Panama. Type Locality: Mazatlan, Mexico. Range: Gulf of California, to Paita, Peru. Collecting Station: Costa Rica: Gulf of Dulce. Description: Shell with about 30 rather broad ribs on the left valve, the anterior ones transvei’sely wrinkled, not nodulose, the posterior ones fiat, subobsolete and separated only by very narrow interspaces. A single left valve referred to Area bifrons is present in the collection from the Gulf of Dulce, Costa Rica. It agrees with Reeve’s figure (pi. 3, fig. 17) and with that given by Maury (pi. 31 (3), fig. 12). We are inclined to question the reference to Sower- by’s figure (Gen. Shells, pi. 217, fig. 3) of A. inaequilateralis to A. bifrons. Apparently the name bifrons proposed by Carpenter has priority over the name cordata of Deshayes which was proposed for A. cardiiformis Sowerby, 1833, not A. cardiiformis Basterot, 1825. The shell of Area bifrons is larger and thicker than that of A. aequatorialis. It also differs from Area aequatorialis in that the posterior end is less produced and the shell slopes more steeply from the posterior umbonal ridge. Area (Cunearca ) esmeralda Pilsbry & Olsson. Area ( Cunearca ) esmeralda Pilsbry & Olsson, Proc. Acad. Nat. Sci. Philadelphia, Vol. 93, September 9, 1941, p. 53, pi. 13, figs. 4 and 5. “Canoa formation, Punta Blanca,” Ecuador, Pliocene. Type Locality: Punta Blanca, Ecuador, Canoa formation, Pliocene. Range : Isabel Island, Mexico, to Panama. Collecting Station: Panama: Gulf of Chiriqui (221-D-1-5), 30-40 fathoms, sandy mud. Description: Shell fairly large and mod- erately thick; valves ornamented by 26 to 28 ribs, those on the left valve strongly nodose, and on the right valve the anterior 8 or 10 are strongly nodose and the remain- der are weakly nodose or smooth; a low wide depressed area extends from the beaks to the posterior ventral margin, most notice- able on large specimens; beaks strongly prosogyrate; the apex of the ligamental area is situated about a third of the length from the anterior end; the area is orna- mented by 2 to 4 chevron-shaped grooves; hinge with a row of teeth which are a little larger at the ends than in the middle, there are about 18 to 20 teeth in the anterior series and 32 to 34 in the posterior series; the margins of the valves are fluted. The dimensions of the type specimen were given as: length, 57 mm., height, 53 mm., semi- diameter, 25.5 mm. The shell of Area esmeralda differs from those of the other West American species of Cunearca in that it possesses chevron- 1943] Hertlein & Strong: Mollusks from, the Pacific 161 shaped grooves on the ligamental area. Small shells of Area esmeralda differ from those of Area nux in that they have chevron- shaped grooves on the ligamental area which is less sloping, in the more numerous ribs, 26 to 28 rather than 22 to 23, the posterior dorsal area is more inflated and less steeply sloping and lacks the strong posterior um- bonal angulation, and the posterior margin is more rounded. Area esmeralda attains a much greater size than does A. nux. Specimens dredged off Mexico and Pana- ma agree so closely with the description and illustrations of Area esmeralda, which was described from the Pliocene of Ecuador, that we have referred them to that species. Distribution: This species occurs from off Isabel Island, Mexico, to Panama. It is also known to occur in the Pliocene of Ecuador. Area tCunearcal nux Sowerby. Area nux Sowerby, Proc. Zool. Soc. Lon- don, May 17, 1833, p. 19. “Hab. ad Xipixapi.” “Found in sandy mud at a depth of twelve fathoms.” — Reeve, Conch. Icon., Vol. 2, Area, 1844, species 1, pi. 1, fig. 1. Original locality record cited. Type Locality: Xipixapi, Ecuador, in 12 fathoms, sandy mud. Range: Concepcion Bay, Lower California, to Zorritos, Peru. Collecting Stations: Mexico: Santa Inez Bay, Gulf of California (145-D-l to 3), 4-13 fathoms, sand; Banderas Bay; Tenacatita Bay; Guatemala: 7 miles west of Champer- ico (197-D-1-2), 14 fathoms, mud; El Sal- vador: La Libertad (198-D-1-2), 13-14 fath- oms; Meanguera Island, Gulf of Fonseca, 16 fathoms, sand, mud; Nicaragua: Corinto (200-D-10-ll,-19) , 7-13 fathoms, mangrove leaves, sand; Costa Rica: Port Parker (203- D-1,-3), 12-15 fathoms, sandy mud, crushed shell, shelly mud; Port Culebra (206-D-1-2- 3), 14 fathoms, sandy mud; Golfito, Gulf of Dulce; Panama: Gulf of Chiriqui (221-D-l- 5), 35-40 fathoms, sandy mud. Description: Shell small, oblique, gibbous, ornamented by about 22 to 23 ribs. Speci- mens of this species are usually not over 15 to 20 mm. in altitude and the ribs are often noded although sometimes only slightly so. The beaks are situated anteriorly about two- thirds to three-fourths the length of the hinge line and the ligamental area is asym- metrical. In Area perlabiata the beaks are nearly central and the ligamentary area is sym- metrical. Area chemnitzii Philippi from the Caribbean region is similar to A. nux, but has a thicker shell, steeper posterior slope and broader ligamentary area. Distribution : This species occurs in fairly shallow water from the Gulf of California to Peru. Area ICunearca) perlabiata Grant & Gale. Area labiata Sowerby, Proc. Zool. Soc. London, May 17, 1833, p. 21. “Hab. ad Real Llejos et ad Tumbez.” “Dredged among sandy mud at a depth of seven fathoms.” Coll. Cuming. — Reeve, Conch. Icon., Vol. 2, Area, 1844, species 7, pi. 1, fig. 7. Original locality record cited. Scapharca ( Cunearca ) labiata Sowerby, Maury, Palaeontogr. Amer., Vol. 1, No. 4, 1922, ‘p. 199 (37), pi. 29 (1), fig. 8. Peru. Area ( Area ) perlabiata Grant & Gale, Mem. San Diego Soc. Nat. Hist., Vol. 1, 1931, p. 141. Earlier records cited. New name for Area labiata Sowerby, not Area labiata So- lander. Type Locality : Real Llejos [near Corinto], Nicaragua, here designated. Tumbez, Peru, also cited originally. Range: Magdalena Bay, Lower California, and the Gulf of California, to Tumbez, Peru. Collecting Stations : Mexico : Arena Bank, Gulf of California (136-D-2), 45 fathoms, Area, conglomerates ; Tangola-Tangola (196- D-17), 23 fathoms, mud; Nicaragua: Corin- to; Isla Cardon; Costa Rica: Port Parker (D-l, 3), 12-15 fathoms, sandy mud, crushed shell, shelly mud; Ballena Bay, Gulf of Ni- coya (213-D-ll to 17), 35 fathoms, mud. Description: Shell heart-shaped, high and fairly thick. There are about 28 ribs which are transversely lightly nodulous. It bears considerable resemblance to Area nux but can be separated by the heavier and much less oblique and larger shell, more numerous ribs, and in that the beak is nearly central in relation to the symmetrically diamond- shaped ligamentary area, while in A. nux the beaks are anterior and the area is asymmetrical. Area chemnitzii Philippi from the Carib- bean has a smaller, less symmetrical liga- mentary area and the beaks are anterior to the center of the shell rather than almost central as in A. perlabiata. Area cacica Olsson from the Miocene of Costa Rica is said to bear some resemblance to A. nux. Distribution: This species occurs from Magdalena Bay, Lower California, and the Gulf of California, to Peru. It has been taken on sand bars, at extreme low tide and has been dredged in rather shallow water. It is also known to occur in the Pleistocene of southern California and LowerCalifornia. Subgenus Larkinia Reinhart. Key to the Species of Larkinia. A. About 25 to 27 radial ribs grandis B. About 31 to 38 radial ribs mvlticostata Area l Larkinia I grandis Bl’oderip & Sowerby. Area grandis Broderip & Sowerby, Zool. Jour., Vol. 4, January, 1829, p. 365. “Hab.” 162 Zoologica: New York Zoological Society [XXVIII: 19 —Reeve, Conch. Icon., Vol. 2, Area, Decem- ber, 1843, species 4, pi. 1, fig. 4. “Hab. Real Llejos, Bay of Guayaquil, etc., western coast of South America; Cuming, Hinds.” Type Locality : Original locality not known. Panama Bay here designated as type locality. Range : Magdalena Bay, Lower California, and the Gulf of California, to Tumbez, Peru. Collecting Stations : Nicaragua: Potosi and Monypenny Point; Castanones Pen- insula, Corinto; Isla Encantada, Corinto; Costa Rica: Port Parker; Ballena Bay; Jasper Island; Golfito Bay; Panama: Bahia Honda. Description : This is a large, very thick, squarish Area which occurs commonly at Panama and other Central American local- ities and is used for food. These big shells are very thick and may weigh as much as two and one-half pounds. They are orna- mented by about 25 to 27 radial ribs which may be slightly nodulous anteriorly. The closest relatives of this species occur in the Miocene of the Caribbean region where a number of names have been applied to various forms such as Area patricia Sow- erby, Area usiaeurii Anderson, Area patri- arclia Anderson, Area patricia waringi Maury, Area grandis colombiensis Weis- bord, and Area grandis cedralensis Weis- borcl. Distribution: This species occurs from Lower California to Peru. It is often taken at extreme low tide on sand bars. Area (Larkiniai multicostata Sowerby. Area multicostata Sowerby, Proc. Zool. Soc. London, May 17, 1833, p. 21. “Hab. ad oras Americae Centralis.” “Dredged from a depth of twelve fathoms in the Gulf of Tehuantepec.”- — Reeve, Conch. Icon., Vol. 2, Area, 1844, species 23, pi. 4, fig. 23. Original locality record cited. Type Locality : Gulf of Tehuantepec, Mexico, in 12 fathoms. Range: Newport Bay, California, to Pan- ama, and the Galapagos Islands. Collecting Stations: Mexico: Santa Inez Bay, Gulf of California (141-D-l to 4, and shore), 7-10 fathoms, sand, sandy mud, crushed shell, weed; Santa Inez Bay (145, shore) ; Santa Inez Bay, Monument Sta- tion; Ceralbo Island, Gulf of California; Arena Point, Lower California; Gorda Banks, Gulf of California (150-D-6), 60 fathoms, muddy sand, rocks; Chamela Bay; Tenacatita Bay; Port Guatulco; Tangola- Tangola (196-D-8), 9 fathoms, sand; Costa Rica: Port Parker; Culebra Bay. Description: Area multicostata Sowerby somewhat resembles A. grandis Broderip & Sowerby but is easily separated by the pres- ence of about 31 to 36 ribs rather than about 25 to 27 in A. grandis which species has a much heavier shell. Area ( Anadara ) reinharti Lowe is somewhat similar to small shells of Area multicostata but it has only about 25 to 27 radiating ribs and is a much more oblique shell. Both Area reinharti and young specimens of A. multicostata are slightly inequivalved, the left valve over- lapping the right. Distribution: This species occurs from southern California to Panama and the Galapagos Islands. In tropical waters it is occasionally taken on sand bars at very low tide and may be dredged in shallow waters. Subgenus Scapharca Gray. Key to the Species of Scapharca. A. Shell with 39 or more ribs obesa B. Shell with about 30 to 32 ribs a. Shell elongate, anterior ribs grooved concinna aa. Shell squarely ovate, anterior ribs not grooved cepoides Area (Scapharca I cepoides Reeve. Area cepoides Reeve, Conch. Icon., Vol. 2, March, 1844, species 66, plate 10, fig. 66. “Hab. San Miguel, South America (found in sandy mud) ; Cuming.” Type Locality: San Miguel, Panama, in sandy mud. Range: Ceralbo Island, Gulf of California, to San Miguel, Panama. Collecting Station: Mexico: Ceralbo Channel, Gulf of California (137-D-3), 46 fathoms, rock. Description: Shell rather large, subquad- rate, ventricose, ornamented by about 32 smooth flat-topped ribs. The specimens ob- tained in the present collection are small, less than 15 mm. in length. Distribution: This species is rather rare- ly taken but is known to occur from the Gulf of California to Panama. Area ( Scapharca i concinna Sowerby. Area concinna Sowerby, Proc. Zool. Soc. London, May 17, 1833, p. 20. “Hab. in America Centrali.” “Found in coarse sand, at a depth of twelve fathoms, in the Gulf of Nocoiyo.” — Reeve, Conch. Icon., Vol. 2, Area, 1844, species 34, pi. 6, fig. 34. Original locality record cited. Type Locality: Gulf of Nicoya, Costa Rica, in 12 fathoms, coarse sand. Range: Santa Inez Bay, Lower California, to Panama. Collecting Stations: Mexico: Santa Inez Bay, Gulf of California (143-D-4-5), 18-25 fathoms, sand; Arena Bank, Gulf of Cali- fornia (136-D-2), 45 fathoms, mud, Area conglomerates; 4 miles south-southwest of Maldanado Point (192-D-l,-3), 26-38 fath- oms, mud; Tangola-Tangola Bay (196-D-l,- 18-19-20), 5-50 fathoms, gray sand, mud; 1943] Hertlein & Strong: Mollusks from the Pacific 163 Costa Rica: Cedro Island, Gulf of Nicoya (213-D-l, 10), 10 fathoms, mud; off Ballena Bay, Gulf of Nicoya (213-D-ll, 17), 35 fathoms, mud; Panama: Gulf of Chiriqui (221-D-1-5), 35-40 fathoms, sandy mud. Description: Shell elongate, and orna- mented by about 30 rather fine ribs, of which the most anterior ones are divided by a medial groove. This shell resembles in general characters A. inaequilateralis Guppy from the Miocene of the Caribbean region. Distribution: This species is known to occur at various localities from the Gulf of California to Panama. Most of the speci- mens dredged by the Zaca Expeditions were from depths of 18 to 50 fathoms. Area IScapharcal obesa Sowerby. Area obesa Sowerby, Proc. Zool. Soc. | London, May 17, 1833, p. 21. “Hab. in Co- lumbia Occidentali.” “A few specimens only i were dredged, in seven fathoms, at Ata- camas.” — Reeve, Conch. Icon., Vol. 2, Area, I 1843, species 3, pi. 1, fig. 3. Original locality cited. Type Locality: Atacames, Ecuador, in 7 fathoms. Range: Off San Jose del 'Cabo, Lower California, Mexico, to Negritos, Peru. Collecting Stations: Mexico: Tangola- Tangola (196-D-17), 23 fathoms, mud; Guatemala: 7 miles west of Champerico (197-D-1-2), 14 fathoms, mud; El Salvador: La Libertad (198-D-1-2), 13-14 fathoms, mud; Nicaragua: Corinto (200-D-19), 12-13 fathoms, mangrove leaves on bottom; Costa Rica: 14 miles southeast of Judas Point (214-D-4), 61 fathoms, mud, rocks; Pana- ma: Gulf of Chiriqui (221-D-1-5), 35-40 fathoms, sandy mud. Description: The ventricose ovate form and large number of fine, close-set ribs are characteristic of Area obesa. The species usually has about 42 to 44 ribs but some specimens from La Libertad, El Salvador, have only 39. Distribution: This species occurs from Lower California to Peru but is not a com- mon species. The specimens from the Zaca Expeditions were dredged from depths of 12 to 61 fathoms. Genus Noetia Gray. Subgenus Noetia s.s. Noetia I Noetia) reversa Gray in Sowerby. Area reversa Gray (MS) in Sowerby, Proc. Zool. Soc. London, May 17, 1833, p. 20. “Hab. in Peruvia.” “Found in soft mud, at a depth of seven fathoms, at Tumbez.” — Reeve, Conch. Icon., Vol. 2, Area, 1843, species 5, pi. 1, fig. 5. Original locality record cited. Type Locality: Tumbez, Peru, in 7 fath- oms, soft mud. Range: Gulf of California, to Peru. Collecting Stations: Guatemala: 7 miles west of Champerico (197-D-1-2), 14 fath- oms, mud; El Salvador: La Libertad ( 198- D-l-2), 13-14 fathoms, mud; Nicaragua: Corinto (200-D-19), 12-13 fathoms, man- grove leaves on bottom; Isia Cardon, Corin- to; Costa Rica: Golfito, Gulf of Dulce; Pan- ama: Gulf of Chiriqui (221-D-1-5), 35-40 fathoms, sandy mud. Description: Shell of medium size, sub- trigonal, inflated, equivalve, ornamented by about 36 ribs. Anterior end rounded, poster- ior end sloping and subtruncated. Beaks large, prominent, posterior, and curved to- ward the posterior of the shell. Distribution: This species occurs at various localities from the Gulf of Califor- nia to Peru. Specimens collected by the Zaca Expeditions were dredged at depths of 12 to 40 fathoms. Subgenus Sheldcnella Maury. Noetia ISheldoneilal delgada Lowe. Area delgada Lowe, Trans. San Diego Soc. Nat. Hist., Vol. 8, No. 6, March 21, 1935, p. 16, pi. 1, fig. 2. “Manzanillo, 20 fathoms.” — MacNeil, U.S. Geol. Surv., Prof. Paper 189-A, 1938, p. 39. Original record cited. Type Locality: Manzanillo, Mexico, 20 fathoms. Range: Manzanillo, Mexico. Collecting Station: Mexico: Manzanillo (184-D-2), 30 fathoms, gravelly sand. Description: Shell ornamented by about 30 ribs. The rounded posterior portion of the shell is expanded and very obliquely produced. MacNeil has suggested the possibility that Area delgada may be related to Noetia ( Eontia ) centrota Guppy which has been reported from the upper Miocene and Plio- cene of Trinidad. However, the general shape of delgada appears to be closer to that of Noetia ( Sheldonella ) maoica Maury from the Miocene of the Dominican Republic. Distribution: This species is known only from the type locality, Manzanillo, Mexico, where it has been dredged in 20 to 30 fath- oms. Subgenus Eontia MacNeil. Noetia I Eontia I olssoni Sheldon & Maury. Noetia olssoni Sheldon & Maury, Palaeon- togr. Amer., Vol. 1, No. 4, 1922, p. 172 (10), pi. 29 (1), figs. 6, 9 (as Area ( Noetia ) olssoni on expl. to plate). “Bucaru,” Los Santos Province, Panama. Eontia olssoni Sheldon & Maury, MacNeil, . U. S. Geol. Surv., Prof. Paper 189-A, 1938, p. 13, pi. 1, figs. 19, 20. Earlier records cited. Type Locality: Bucaru, port of Tonosi, Los Santos Province, Panama. 164 Zoologica: New York Zoological Society [XXVIII: 19 Range : Corinto, Nicaragua, to Negritos, Peru. Collecting Station: Nicaragua: Corinto (200-D-10, 11, 19), 7-13 fathoms, mangrove leaves, sand. Description: The shell of Noetia ( Eontia ) olssoni is easily recognized from that of other west American species of Noetia by the narrow, elongate shell in which the posterior end is not transversely expanded but bears a medial depression. Noetia olssoni is similar to N. centrota Guppy from the upper Miocene and Pliocene of Trinidad, and N. bisalcata Lamarck, Re- cent in the Caribbean and Atlantic coast of South America from Colombia to Uruguay. The west American species differs in the greater length posteriorly, more central beaks and nearly equal rows of teeth and in other details. The largest specimen in the present collection measures 14.6 mm. in length. Distribution : The occurrence of this species at Corinto, Nicaragua, furnishes an extension northward in its range. It occurs south to Peru. It is also known to occur in the Pleistocene of Magdalena Bay, Lower California. Superfamily Pteriacea. Family Pteriidae. Key to the Genera of the Pteriidae. A. Hinge extended posteriorly forming a wing Pteria B. Hinge not extended posteriorly to form a wing Pinctada Genus Pteria Scopoli. Pteria sterna Gould. Avicula sterna Gould, Proc. Boston Soc. Nat. Hist., Vol. 4, November, 1851, p. 93. “Inhabits Mazatlan.” — Gould, Boston Jour. Nat. Hist., Vol. 6, 1853, p. 404, pi. 16, fig. 7. “Inhabits Mazatlan? Lieut. Green. Panama. Prof. Adams, Col. Jewett.” Avicula peruviana Reeve, Conch. Icon., Vol. 10, Avicula, March, 1857, species 53, pi. 14, fig. 53. “Hab. Peru.” Type Locality: Mazatlan, Mexico. Range: Hueneme Point, California, to the Gulf of California, and south to Paita, Peru. Collecting Stations: Mexico: Cedros Is- land, in channel (126-D-19), 25 fathoms, rocks, algae; Santa Inez Bay: Arena Bank; Cape San Lucas, Lower California. Description : Shell oblique and with a pos- terior wing and shorter anterior ear be- neath which is a byssal notch. Rows of spines occur on unworn shells. On worn shells yellowish stripes usually are present on a brown background. The degree of development of a posterior wing, thickness of shell, and obliquity of the hinge line appear to be variable charac- ters in Pteria sterna. Some specimens have a well developed posterior wing while in others this is only slightly developed. In some specimens it can be observed that a long wing was present in the young stage but later the sinus on the posterior margin became more rounded and thus produced only a small wing. Some specimens possess thicker shells of a dark color which are ornamented by fine radial color bands. This appears to be an inconstant character as shown by some specimens on which the earlier part of the shell is dark and the later part is ornamented by yellowish stripes. Odhner pointed out that the width of color bands of Pteria (Electrotoma) zebra Reeve depends upon the situs to which it is at- tached. From a consideration of the characters shown in a series of specimens it appears that there is no satisfactory basis for sepa- rating Pteria peruviana Reeve from P. sterna Gould. Distribution: This species occurs at vari- ous localities from southern California to Peru. It is only occasionally found in south- ern California, but it occurs rather abun- dantly at certain localities in the Gulf of California. It does not occur in the great numbers originally found in the Gulf of California due to pearl fishing. This species is said to have produced literally bushels of pearls. Genus Pinctada BOLTEN. Pinctada mazatlanica Hanley. Meleagrina mazatlanica Hanley, Cat. Rec. Bivalve Shells, Ap., 1856, p. 388, pi. 24, fig. 40. “Mazatlan.” Avicula barbata Reeve, Conch. Icon., Vol. 10, Avicula, March, 1857, species 9, pi. 5, fig. 9. “Hab. Panama (under stones at low water) ; Cuming.” Type Locality: Mazatlan, Mexico. Range : Gulf of California, to Paita, Peru, and the Galapagos Islands. Collecting Stations: Mexico: Santa Inez Bay, near Concepcion Point, Lower Cali- fornia; Banderas Bay; Chamela Bay; Pass- avera Island, Chamela Bay (shore) ; Port Guatulco (195-D-15, also shore), 1.5 fath- oms, coral bottom; Tangola-Tangola Bay; Nicaragua: Isla Cardon, Corinto; Costa Rica : Port Parker ; Potrero Grande ; Cule- bra Bay; Golfito, Gulf of Dulce; Piedra Blanca; Colombia: Gorgona Island. Description: This is the “concha de perla” or “Panama shell,” the common com- mercial pearl oyster of the Gulf of Cali- fornia and Panama. The color of the ex- terior of the shell of Pinctada mazatlanica is grayish-yellow or light brown. On unworn specimens there are l'adial rows of pointed foliaceous spines and often radial rows of 1943] Hertlein & Strong: Mollusks from the Pacific 165 spots a little darker than the ground-color. The nacre is silvery white with a narrow golden or brassy colored margin. This species is closely related to Pinctacla margaritifera Linnaeus, one of the well known pearl oysters of the Indo-Pacific region. Jameson believed that there is in- tergradation between the two species. He also mentioned that in the group related to P. margaritifera, the lightest colored forms, mazatlanica, Zanzibar ensis, persica and erythrceensis, occur on the shores of the great continents, the Australian and Malay shells are intermediate in color, and the oceanic variety cumingii is darkest in color. Compared to Pinctada margaritifera and others of the group, P. mazatlanica is char- acterized by its greater convexity, greater anterior projection, and light brown color. Jameson pointed out that the anterior mar- gin below the byssal notch projects farther forward than in any other related form. A perpendicular to the anterior end of the hinge would cut off about one-third of the valve. The posterior angle of mazatlanica is acute or but rarely a right angle. The posterior margin of the nacre slopes for- ward from the hinge, resembling in this character P. maxima Jameson described from New Guinea. Distribution : This species occurs from the Gulf of California to Peru. It is fairly abundant at various localities in the Gulf of California and Panama where it is fished for pearls. It was formerly very abundant in the Gulf of California but the beds have been greatly depleted due to pearl fishing. Family Pinnidae. Key to the Genera of the Pinnidae. A. Shell triangular; a median groove interiorly dividing the nacreous layer into two parts Pinna B. Shell ham-shaped; no median groove interiorly A Irina Genus Pinna Linnaeus. Pinna rugosa Sowerby. Pinna rugosa Sowerby, Proc. Zool. Soc. London, September 25, 1835, p. 84. “Hab. in Sinu Panamensi. (Isle of Rey).” “They were procured from sand banks.” — Reeve, Conch. Icon., Vol. 11, Pinna, 1858, species 50, pi. 26, fig. 50. Original locality cited. Type Locality : Isle of Rey, Bay of Pana- ma, sand banks. Range : Manuela Lagoon, Lower Califor- nia, and Punta Penasco, Sonora, Mexico, to Panama. Collecting Stations: Mexico: Gorda Banks, Gulf of California (150-D-7), rock, calcareous algae; Acapulco; Nicaragua: Corinto (200-D-1-3), 2 — 6.5 fathoms, man- grove leaves; Castanones Peninsular La- goon, Corinto; Costa Rica: Long Beach N. W. of Port Parker; Golfito Bay; Gulf of Dulce. Description: This species possesses an elongate, triangular, rugose shell. It grows to a length of a foot and a half or more and is ornamented by about eight rows of foliaceous, tubular spines. The spines may be almost obsolete on old specimens. Distribution: This species occurs rather commonly along the west coast of Mexico and south to Panama. Genus Atrina Gray. Key to the Species of Atrina. A. Shell ornamented by 26-38 rows of spines a. About 38 rows of spines; ventral margin broadly rounded oldroydii aa. About 26 rows of spines ; ventral margin slightly rounded texta B. Shell ornamented by about 18 rows of strong, regular, tubular spines maura Atrina maura Sowerby. Pinna maura Sowerby, Proc. Zool. Soc. London, September 25, 1835, p. 84. “Hab. apud Panamam.” Obtained “from muddy banks.” — Reeve, Conch. Icon., Vol. 11, Pinna, 1858, species 54, pi. 29, fig. 54. Orig- inal locality cited. Type Locality: Panama, muddy banks. Range: Santo Domingo (in kitchen mid- dens), and Magdalena Bay, Lower Califor- nia, and Punta Penasco, Sonora, Mexico, to Peru. Collecting Stations : Mexico : Gulf of Cali- fornia, Santa Inez Bay, near Point Concep- cion along shore of Lower California, also at Monument Station; Banderas Bay; Nic- aragua: Monypenny Point, Gulf of Fonseca, 4 fathoms, mud; Corinto (200-D-1-10) , 1/2 to 7 fathoms, mangrove leaves. Description: The shell of this species has a nearly straight dorsal margin. About 18 rows of fairly strong, unusually regular, yellowish-white arched tubular spines occur on a brown or rusty black background ; these spines almost disappear on the anterior side. Some specimens are of a yellowish or dark greenish-olive color. According to Pilsbry and Lowe the shells of Atrina maura are known as “hachas” or hatchets by the native fishermen along west- ern Mexico. The large white muscle which is used for food is said to resemble and taste much like that of the giant scallop of the Atlantic Coast. Winckworth indicated that “Pinna” tu- berculosa Sowerby is identical with “Pinna” maura Sowerby. However, the descriptions of Sowerby and Hanley and the figures given by Reeve seem to indicate that tuber- 166 Zoologica: New York Zoological Society [XXVIII: 19 culosa is a distinct species and that Pinna lanceolata Sowerby, a preoccupied name, can be placed in the synonymy of maura. The shell of Atrina tuberculosa Sowerby is tri- angular, inflated, and quite different in gen- eral appearance from that of A. maura. Atrina rigida Dillwyn from the Antillean region bears a resemblance to A. maura. Atrina oldroydii Dali. Atrina oldroydii Dali, Nautilus, Vol. 14, No. 12, April, i901, p. 143. “Taken alive by fishermen in 25 fathoms, San Pedro Bay,” California.- — Dali, U. S. Nat. Mus., Bull. 112, 1921, p. 17, pi. 2, figs. 4, 5, 6. Off San Pedro, California, in 25 fathoms. Type Locality: San Pedro Bay, California, in 25 fathoms. Range : San Pedro, California, to Cedros Island, and Magdalena Bay, Lower Cali- fornia. Collecting Station : Mexico: east of Ced- ros Island, in Channel (126-D-17), 40 fath- oms, mud. Description: The shell of this large dark colored Atrina from southern California and Lower California is much wider than that of Atrina texta n. sp. and the posterior end is more oblique and less square in out- line than is that of A. maura. There are about 38 radiating ribs which bear rather fine low spines which do not extend over the ventral surface and on the posterior fourth of the shell the ribs and spines are obsolete. The specimens in the present col- lection are broken but the largest one meas- ures 240 mm. from beak to base. Atrina serrate Sowerby from the Atlantic coast is a similar species. Distribution: This species is not common- ly found but is known to occur in the waters off southern California and along the west coast of Lower California, south to Magda- lena Bay. Atrina texta Hertlein, Hanna & Strong, sp. nov. Plate I, Figures 9 and 10. Shell roughly triangular with a curved beak; dorsal side gently convex, the ventral concave behind the beak then gently convex, straightening out to the posterior margin which is gently rounded; shell translucent, showing the external sculpture on the in- terior; color pale horn brown, darker on the beaks and on a large irregular blotch over the muscle scar; posterior margins gaping about 10 mm. and ventral margin ganing about 5 mm. in the concave portion behind the beaks; exterior sculpture con- sisting of about 26 radial rows of short and low triangular spines, slightly recurved; these spines are also regularly arranged in rows concentric with the growth lines; this zone of sculpture completely crosses the ventral margin but the rows become fewer anteriorly as the zone crosses the umbonal ridge; ventral zone marked only by rather coarse growth lines except posteriorly where the radial rows of spines gradually fade away. Length, 141 mm.; width, 80 mm.; convexity, 35 mm. Holotype in Calif. Acad. Sci. Paleo. Type Coll., from Sta. 150-D-4, Gorda Bank area off the southern end of Lower California, Lat. 23°01' N., Long. 109°30' W. ; dredged in 70 fathoms; collected by the Templeton Crocker-Beebe Expedition, April 21, 1936. This interesting species has been com- pared with all the other west coast forms of Atrina and in none of them is there found the combination of characters shown here. Atrina oldroydii Dali is perhaps the closest; that species also has radial and concentric rows of fine spines but the shape is decidedly asymmetrical, the beaks are not hooked and the dorsal margin is scarcely convex. The shell of the new species differs from that of Atrina maura Sowerby in the wider, squarer posterior end and in that it is ornamented by about 26 radial rows of triangular spines rather than about 18 rows of regular, semitubular spines. Family Pedalionidae. Genus Pedalion SOLANDER IN HUDDESFORD. Pedalion chemnitzianum d’Orbigny. Plate I, Figure 8. Perna chemnitziana d’Orbigny, in Sagra, Moll. Cuba, Vol. 2, 1845, p. 346. “Nous l avons recue de Cuba, de la Martinique et de Sainte-Croix ; elle vit attachee aux roch- ers par son byssus au niveau des basses marees ordinaires.” Reference to Concha semiaurita. Chemnitz (not Ostrea semiau- rita Linnaeus), Neues Svst. Conchvl.-Cab., Vol. 7, p. 250, Tab. 59, fig. 580. West Indies. Type Locality: Cuba, here designated. Martinique and St. Croix also cited origin- ally. Range: Coronado Islands, off the west coast of northern Lower California, Mexico, to Chile. Also Atlantic. Collecting Stations: Mexico: Sulphur Bay, Clarion Island; Port Guatulco (195-D-14), 4 fathoms, coral; also on shore; Tangola- Tangola, on beach; Nicaragua: Fumarole Island; Potosi and Monypenny Point; Costa Rica: Piedra Blanca. Description: The shell of Pedalion chem- nitzianum is very variable. It may be quad- rate, elongate, or irregular in shape and it may be fairly thick or thin. It is sometimes oi’namented by fine radial ornamentation but usually this is lacking and the exterior shows only scaly concentric layers. The col- or may be yellowish-brown but often it is partially and occasionally almost wholly, of a purple color. There are usually about 6 1943] Hertlein & Strong: Mollusks from the Pacific 167 to 8 and occasionally as many as 12 liga- mental pits along the hinge line. Right valve with a byssal sinus. There appears to be no satisfactory method of separating the west American forms here referred to Pedalion chemnitzi- cmum from the east American species of this name. Many workers familiar with the At- lantic species have been unable to detect any constant differences between the forms oc- curring on the east and west coast. Pedalion janus Carpenter, a species occurring com- monly from San Ignacio Lagoon, Lower California, and the Gulf of California to Oaxaca, Mexico, usually possesses a thinner shell ornamented by radial sculpture, and it apparently has a more northern distribu- tion; at least it has not been observed by us in collections from Panamic waters. However some specimens seem to show the characters of both P. chemnitzianum and P. janus, as pointed out by Stearns. No doubt the two forms are closely related. Car- penter believed that variants of P. chem- nitzianum approach the Hawaiian species P. incisum Conrad, while those of P. janus approach P. costellatum Conrad. Distribution : This species occurs com- monly attached to rocks or other objects between tides or in shallow water from northern Lower California to Chile, and also Florida and the Caribbean region. 168 Zoologica: New York Zoological Society EXPLANATION OF THE PLATE. Plate I. Fig. 1. Area ( Anadara ) mazatlanica Hertlein & Strong, sp. nov. Holotype, right valve, from Station 153-D-2, dredged in 120 fathoms (218 meters), Lat. 23°06'00" N., Long. 106°47'00" W„ 19 miles west of Mazatlan, Sinaloa, Mex- ico. Length, 62 mm., height, 36.3 mm., convexity (one valve), 15.5 mm. P. 156. Fig. 2. Area (Anadara) similis C. B. Adams. Hypotype, right valve, from Puntare- nas Lagoon, Costa Rica. Length, 50 mm., height, 31.8 mm., convexity (one valve), 15mm. P. 157. Fig. 3. Area ( Anadara ) biangulata Sowerby. Hypotype, left valve, from Station 143- D-5, dredged in 18 fathoms (33 me- ters), Lat. 26°54'00'' N., Long. 111°53'00" W., Santa Inez Bay, Lower California. Length, 49 mm., height, 36 mm., convexity (one valve), 17 mm. P. 155. Fig. 4. Area ( Anadara ) mazatlanica Hertlein & Strong, sp. nov. Holotype. View of the interior of the left valve of the specimen shown in Fig. 1. Fig. 5. Area ( Anadara ) similis C. B. Adams. View of the interior showing hinge of the left valve of the specimen shown in Fig. 2. Fig. 6. Area (Lunarca) vespertina Morch. Hypotype, right valve, from Loc. 27230 (C.A.S.), Petatlan Bay, about 6 miles south of Sihuatenejo, Guerrero, Mex- ico. View of the interior. Length, 36.1 mm., height, 22.5 mm., convexity (one valve), 9.1 mm. P. 158. Fig. 7. Area ( Lunarca ) vespertina Morch. View of the exterior of the specimen shown in Fig. 6. The shell of this species becomes more elongate in large specimens such as this one, but juvenile forms are more quad- rate in shape. Fig. 8. Pedalion chemnitzianum d’Orbigny. Hypotype, right valve, from Loc. 28186 (C.A.S.), Kino Bay, Sonora, Mexico. Length, 20.5 mm., height, 46.5 mm., convexity (both valves), 8 mm. P. 166. Fig. 9. A trina texta Hertlein, Hanna & Strong, sp. nov. Holotype, right valve, from Station 150-D-4, Gorda Banks off the southern end of Lower California, Lat. 23°01'00" N., Long. 109°29'00" W., dredged in 70 fathoms (128 meters). Length, 141 mm., width, 80 mm., con- vexity (both valves), 35 mm. P. 166. Fig. 10. Atrina texta Hertlein, Hanna & Strong, sp. nov. Left valve of the specimen shown in Fig. 9. All the specimens illustrated on this plate are in the type collection of the Department of Paleontology of the California Academy of Sciences. PLATE I HERTLEIN & STRONG. MOLLUSKS FROM THE WEST COAST OF MEXICO AND CENTRAL AMERICA. NEW YORK ZOOLOGICAL SOCIETY General Office: 630 Fifth Avenue, New York 20, N. Y. OFFICERS Fairfield Osborn, President Alfred Ely, First Vice-president Laurance S. Rockefeller, Second Vice-president Harold J. O’Connell, Secretary Cornelius R. Agnew, Treasurer William Bridges, Editor and Curator of Publications Claude W. Leister, Education Donald Marcy, Associate, Education Zoological Park Lee S. Crandall, General Curator & Curator of Birds Leonard J. Goss, Veterinarian Claude W. Leister, Associate, Mammals Harry C. Raven, Associate, Anatomy John Tee-Van, Associate, Reptiles Grace Davall, Assistant to General Curator Charles M. Breder, Jr., Director Christopher W. Coates, Aquarist Ross F. Nigrelli, Pathologist Myron Gordon, Research Associate in Genetics G. M. Smith, Research Associate in Pathology Homer W. Smith, Research Associate in Physiology SCIENTIFIC STAFF General John Tee-Van, Executive Secretary Jean Delacour, Technical Adviser Aquarium Department of Tropical Research William Beebe, Director Jocelyn Crane, Research Zoologist Henry Fleming, Entomologist William K. Gregory, Associate Gloria Hollister, Associate John Tee-Van, Associate Mary VanderPyl, Associate Editorial Committee Fairfield Osborn, Chairman William Beebe Charles M. Breder, Jr., William Bridges Jean Delacour Claude W. Leister John Tee-Van ZOOLOGICA SCIENTIFIC CONTRIBUTIONS of the NEW YORK ZOOLOGICAL SOCIETY VOLUME XXVIII Part 4 Numbers 20-23 Published by the Society The Zoological Park, New York 60, N. Y. December 31, 1943 CONTENTS 20. Chemical Sensory Reactions in the Mexican Blind Characin. By C. M. Breder, Jr., & Priscilla Rasquin. (Plates I-III; Text- figures 1-9) 169 21. The Avian Genus Zosterops in Siam, with Description of One New Race. By H. G. Deignan 201 22. Causes of Diseases and Death of Fishes in Captivity. By ROSS F. Nigrelli. (Plates I- VI) 203 23. Display, Breeding and Relationships of Fiddler Crabs (Brachyura, Genus TJca) in the Northeastern United States. By Jocelyn Crane. (Text-figure 1) 217 Index to Volume XXVIII 225 Breder & Rasquin: Mexican Blind Characins 169 20. Chemical Sensory Reactions in the Mexican Blind Characins. C. M. Breder, Jr. & Priscilla Rasquin* New York Zoological Society (Plates I -III; Text-figures 1-9). Introduction. As in all studies concerned with the chem- In the attempt to understand more fully the various features of the behavior and possible evolution in progressive stages of the changes in the Mexican blind fish of the Astyanax mexicanus (Philippi) — Anoptich- thys jordani Hubbs and Innes series it is, of course, necessary to obtain some evaluation of the various sensory systems. The present contribution is therefore such an attempt to analyze the differences, if any such be present, in the chemical sensory equipment of these fishes. Both the structural and be- havioristic elements involved are herewith discussed. Recently a new form has been discovered in another cave by one of Dr. Hubb’s col- lectors, as noted by Tafall (1942 and 1943) and Breder (1943). This form is evidently still further advanced in eye and pigmenta- tion loss. We have been fortunate in obtain- ing some of these alive through the good offices of Mr. B. Dontzin who visited this cave, Cueva de los Sabinos, in 1942 for that express purpose. Tafall (1943) has made an important contribution to the ecology of both La Cueva Chica and Cueva de los Sabinos. In his study of conditions he enum- erates the species of aquatic organisms and discusses at length the fishes of these caves, making comparisons with the fauna of the cenotes of Yucatan. The taxonomic consid- erations involved are under study by Dr. Hubbs, while the present paper discusses some of the anatomical and behavioristic items of this form compared with fishes from La Cueva Chica and normal-eyed river fishes. The details of the ocular anatomy and reactions to light in this form are discussed by Breder and Gresser (Ms.) A popular version of the history of these studies may be found in Bridges (1940 and 1943). * The studies embodied in this paper were started as part of the work in the graduate course in ichthyology given by the senior author at New York University. Subsequently the studies were expanded to their present form. ical senses of fishes it was found much more difficult to obtain definitive results than in the earlier studies on reactions to light of Breder and Gresser (194a and b). We feel, however, that enough data have been accumulated to indicate the basic nature of these features in reference to the behavior of the forms involved and the apparent nature of the evolutionary changes in prog- ress. The work was carried on in the Depart- ment of Animal Behavior of the American Museum of Natural History. We are grate- ful to Miss Annette Bacon for advice in connection with the mathematical treatment and for editorial assistance and to Lt. James W. Atz for numerous helpful comments. The Nasal Capsule. Serial sections and gross dissections of the nasal capsule of each of three types under comparison showed that these were rather typical characin structures, but that each form showed a slight but signifi- cant morphological difference, in which La Cueva Chica material was intermediate between the eyed river fish from the Rio Tampaon and Cueva de los Sabinos speci- mens. These differences were chiefly in the reduction of the number of sensitive lamellae and a decrease in the depth of the pit, pro- ceeding from the river to the Cueva de los Sabinos stocks. The functional significance of the sense of smell involved in these changes is not entirely clear on an anatomi- cal basis. The funnel-like flap that serves to direct water into the anterior opening increases in size slightly in this series as is indicated in Plate I, wherein the right hand flaps are shown in profile. No histological difference could be found in the olfactory epithelium from one form to another. Text-figure 1 shows a conven- tionalized reconstruction of the ridges in relation to the positions of the anterior and 170 Zoologica : New York Zoological Society [XXVIII: 20 posterior nares of each type. As is indicated, in the river fish there are eight pairs of ridges, in La Cueva Chica material seven and in the Cueva de los Sabinos only five. A se- lected, approximately homologous section of each of these is shown photographically in Plate II. Here the raising of the floor of the pit and the more fully exposed positions of the finger-like lamellae sections in the Cueva de los Sabinos fish is indicated. It is also evident from this plate that in the river fish some of these lamellae become anchored at their distal ends more anteri- orly and are more numerous than in either cave fish. The nasal openings of the Cueva de los Sabinos fish are larger and offer more direct exposure of the ridges to the exterior. vulgaris Cuvier to number about ten pairs, figures the nasal pits of Gadus morrhua Lin- naeus ind Mugil chelo Linnaeus and discus- ses nasal anatomy in general. Adrian and Ludwig (1938) indicate about 31 pairs for Ameiurus nebulosus. Busnita (1932) dis- cusses the histological details of the sensi- tive epithelium of Carassius, while Laibach (1937) shows that in development the nasal lamellae increase in number from four to thirty-six pairs as Anguilla vulgaris Lin- naeus transforms from the leptocephalus stage and grows to adult size. That possible size or age difference could not have any- thing to do with the differential counts in the characins has been checked. Specimens long before reaching the size of a small adult Text-fig. 1. Diagrammatic representation of the sensory ridges in the nasal capsules of Mexican characins. Left: from Rio Tampaon. Middle: from La Cueva Chica. Right: from Cueva de los Sabinos. Based on reconstruction from serial sections and gross dissections. In each case the left nostril is represented in plan view with the fish facing upward. The dotted outlines represent the anterior and posterior external openings. See also Plates I and II. Whether these features are to be inter- preted as a reduction of nasal acuity, on the basis of a reduction of sensitive areas, or an increase, based on the greater expos- ure of the remaining areas and a larger tunneling flap, is something that must wait on the availability of a larger stock of the Cueva de los Sabinos fish suitable for a more extended experimental analysis. There is considerable variation in the nasal anatomy of teleosts as is well indica- ted by Liermann (1933) who figures the lamellar construction of Carassius carassius Nordmann, Anguilla vulgaris Flemming, Perea fluvitalis Linnaeus, Pleuronectes fle- sus Linnaeus, Ammodytes tobianus Lin- naeus, Zoarces viviparous Linnaeus, Gas- 'terosteus aculeatus Linnaeus and Syphono- stomum typhle Linnaeus. Tretajakoff (1930) shows the lamellae of Scomber scombrus Linnaeus to be practically radially arranged and the platelets to number about twenty. Matthes (1934) shows the lamellae of Tinea show a fixed and constant number. If such changes take place with the development of these characins it must be at a size below which we have comparative material. As would be expected the nasal capsules of the cyprinids more closely resemble those under discussion than do the other species more remotely related. Many other authors have described the nasal anatomy of a variety of fishes includ- ing figures of Zeus faber Linnaeus and Gadus morrhua by Berghe (1929), Protop- terus by Fullarton (1933) and various Het- erosomata by Chabanaud (1927 and 1936). Derscheid (1924) discusses at length the nasal capsules of the Isospondyli with fig- ures of the anatomy of no less than thirty species well distributed throughout the order. Most of these have numbers of lamel- lae much higher than our characins but the following approach them in this sense: Chaidiodus sloani Bloch and Schneider with five, Pantodon bucholtzi Peters with ten 1943] Breder & Rasquin: Mexican Blind Cliaracins 171 pairs, Heterotis niloticus Cuvier and Valen- ciennes with sixteen radially arranged, while seven species of Mormyridae of the genera Marcusenius, Mormyrus, Genyomy- rus and Mormyrops range from six to eight pairs of lamellae. In general terms these Isospondyle nasal pits resemble those of the Ostariophysi about as closely as these two orders resemble each other in most respects. Frisch (1941a) indicates ten pairs of lam- ellae in the nasal pit of Plioxinus laevis Agassiz. It is evident from this brief survey of the literature, moreover, that the sensi- tive lamellae of these characins approach the lower limits in number much closer than they do the upper. From this and general considerations we cannot but conclude that the fishes in question are not nearly as well provided for in regard to olfaction as are many other species, including a number with especially well developed eyes. The Taste Buds. Complete serial sections of one of each of the three forms has made possible the recon- struction of the distribution of the taste bud tracts on the exterior surfaces of the fishes. The distribution of these in each is indicated in Text-figure 2. As each sectioned fish varied slightly in bodily proportions, all have been reduced to a standard outline in order to facilitate comparison. The ac- tual outline of each fish is indicated by the small inserts. The dots on the surface of the outline are indicative of the positions of gustatory areas and are not intended to represent individual taste buds in terms of absolute number. It is clear from this figure that there is a marked increase in the areas of gustation from the river fish to those from Cueva de los Sabinos. This is the first unequivocal positive change that has been found in this series. Other changes found, excepting perhaps only the nasal pits, have all been in the nature of reductions, Breder and Gresser (1941a) and Breder (1943). It is to be noted in this figure that the optic area in La Cueva Chica stock, 7th generation of aquarium-reared fishes, is scarcely encroached upon by taste buds whereas in the more advanced Cueva de los Sabinos material this area is well covered by them. The former had a sunken eye, un- covered, lacking a lens and similar to the fish of Plate II of Breder and Gresser (1941a). While the general form of the chief tracts of taste buds is fairly constant it is to be noted that with the increasing number there is evidently some slight translocation of some of the areas of taste. Also, along with this increase in taste areas, chiefly on the head, there is evidently some slight re- duction of them on the caudal fin, but this is perhaps little more than a matter of in- dividual variation. These findings have been reinforced by the examination of other spec- imens in less detail. The taste buds within the oral cavity, not figured, showed no significant change. In all three specimens and others the taste buds themselves showed no evident morph- ological change from one fish to another. There is, however, a marked change in the form of the taste buds of each fish, iden- tical in nature, from one region of the body to another. These changes of form are in- dicated in Plate III. Those within the mouth are raised on little papilla-like mounds in the corium. Those embedded in the rela- tively thick epidermal covering of the head are merely pushed up to the general surface with which they are flush. Those on the body in the thin epidermal covering of the scales are broadened and flattened out in a manner suggestive of the neuromasts of the lateral-line canal but are much larger and histologically more clearly defined. So far as can be determined at this time these modifications of the exterior taste buds are merely an expression of mechanical re- sponses to the thickness or other features of the epidermal layers in which they hap- pen to be imbedded, excepting possibly those within the oral cavity which appear to be more highly developed. Comparative Behavior. In order to attempt an evaluation of the significance of the chemical sensory appara- tus of these fishes a series of experiments was undertaken on the reactions of the three types to various chemical stimuli. For this purpose normal fish from Cueva de los Sabinos and La Cueva Chica stock were employed as well as blinded river fish and anosmic La Cueva Chica specimens. Inci- dental to this it was found that river fish, in which the optic nerve had been severed, on recovery took on the essential wandering behavior of the naturally blind fish. That is, they commenced an incessant wandering very like that characteristic of the blind fish as reported by Breder and Gresser (1941a) and Breder (1943). The former, in studying light reactions, did not blind their river fish, with the result that they could not treat them in the same statistical ’man- ner as the cave fish. It would have been of no significance to their purposes, for when blinded, these fish become entirely indiffer- ent to light as may be shown in a gradient trough. On the contrary, in present connec- tions it is obvious that both experimental and control animals must operate without visual cues. These blinded river fish quickly became accommodated to their new status and lived on as well as the cave fish and found their food in an identical manner. The anosmic fish were produced by means of electro-cautery of the nasal capsules. On recovery these showed little basic differ- Text-fig. 2. Distribution of exterior taste buds on Mexican characins. Upper: from Rio Tampaon, 34 mm. in standard length. Middle: from La Cueva Chica, 29 mm. in standard length. Lower: from Cueva de los Sabinos, 32 mm. in standard length. Re- constructed from serial sections. The small inserts represent the true outline of the fish in each case. The large outlines of identical nature are based on a typical form to which the taste bud distribution has been referred in order to facilitate comparison. See text for full explanation. 1943] Breder & Rasquin: Mexican Blind Characins 173 ence from the normal fish except that they did not find their food as quickly. Beigel-Klaften (1913) discussed the regen- eration of the nasal organ in Tinea vulgaris and in Cyprinus carpio Linnaeus. With the electro-cautery method employed in the pre- sent experiments, examination as well as the subsequent behavior which remained the same after several months showed there was clearly no reconstruction of the apparatus or ability to smell. Much of the work on comparative be- havior was undertaken in a trough 22 inches wide by 39 inches long with water to a depth of two inches. This was optically divided into two like compartments by means of a string drawn taut above the surface of the water so that there were, in effect, two parts, each I8V2 inches by 22 inches. Burettes were suspended, one over the end of each, five inches from the end of the trough along its mid-line, so that the two sources of chemicals were 29 inches apart. In these troughs the positions of the fishes, in reference to the two compart- ments, were checked at five second inter- vals in a manner similar to that used by Breder and Gresser (1941a and b) for their studies on light reactions. In most of the experiments four fish were used at a time as more than that number were generally too difficult to keep accurate check on si- multaneously. The end at which the chem- icals were administered was alternated with each experiment in order to overcome any possible inherent bias in the trough. For studies involving the circling move- ments of normal and semi-anosmic fishes another trough of 39 inches by 17 inches was employed, but which was otherwise similar. After each experiment, the trough was flushed and tempered water was substituted for the contaminated water of the last test. The temperature of the water was main- tained between 21° and 25° C. throughout the experiments described. Chemical Repellents. In the early exploratory part of the ex- periments it was found that acetic acid operated well as a repellent for these fishes and as a consequence it was largely used in these studies. Citric acid was also used but was found to show no advantage. Am- monium carbonate was also employed. The simplest and most satisfactory method of application tried was to allow dilute solu- tions to drip from a burette at as nearly a uniform rate as possible at one end of the trough while plain water from the trough dripped at a similar rate at the opposite end. Any possible, but not readily detec- table, mechanical disturbance due to the slow inflow was thus compensated. The basic data of these experiments are given in Tables I, III and IV together with perti- nent data on quantities administered and pH values reached. Table I gives the chemical and related data as well as the positions of the fish in percentage at the repellent end of the trough. Every experiment is designated by a number and letter. Each number indi- cates the use of a different substance or quantity. The letters indicate separate ex- periments, of which, in most cases, there were four of one kind. Each period of ob- servation, of which there are eight, indi- cated at the right of the table, represents 100 observations at 5-second intervals. These values indicate the number of fishes recorded, expressed in percentage. Thus in experiment “la” under period “1,” the value 40 means that during the 100 observations made, 160 fishes were counted out of a possible 400 (4 fishes moving at random under such conditions would show 200 in one compartment or 50% of the total pos- sible). All periods for each experiment, or horizontal row of figures, represents a pe- riod of 58 minutes and 20 seconds of ob- servation. The even-numbered periods are consecutive with the preceding odd-num- bered periods while an interval approxi- mating a half hour represents the spacing of the beginning of each odd-numbered pe- riod. The means of each experiment thus represent an actual observation time of nearly two hours (1 hour, 53 minutes, 20 seconds) with 3,200 observations at five- second intervals. Table III, together with other data to be discussed later, gives continuations of five of the tests that were carried on for longer periods, the longest reaching 1,900 obser- vations over a period of more than two and one-half hours (2 hours, 37 minutes, 30 seconds), the latter of which were spread over longer units of time as is indicated in the table. Table IV gives details of periods “5” and “6” of Tables I and II in eight smaller units together with other data. The horizontal rows of figures in this table are all con- secutive and represent 200 observations with the means representing 800 observa- tions. A digest of this data is given in Text- figure 3. Graph A shows the retreat of the fishes during the actual acid-dripping pe- riod, based on Table IV, and indicates clearly the similarity in behavior between the blinded river fish, the normal fish from La Cueva Chica and those from Cueva de los Sabinos. It also indicates the difference in behavior of the anosmic fish as com- pared with the normals and the difference between the reactions of the normals to ammonium carbonate as compared to acetic acid. Graph B treats the data of Table I in a 174 Zoologica: New York Zoological Society [XXVIII: 20 H 05 < Oh 05 Q xn S' lo © cs cd © lo »o t- i wo n H lOHCOCO ©3 LO ^ lO | COCO ^ CO ^7- 03 © -fa 03 WHCOH N t>- 00 LO Lo OilMNCO Oj 'fa I Oo ZD 03 'fa ^ t-H LO t-h CD t— I CO 03 Co 03 lO | TfTf ^ CO xn H Z w J H w cu w K X H 05 w cu X W O £ h c ^ o r 05 § % « £§g P> PQ C/2 O • fa OIOCOC3 ^ Tj^OOt- O 0 05^00 OOOOCOCD OI, rf c] -'t *-fa ZD tH 03 T — | t-H tH t>h t— I t-H CO 03 03 r— I -fa LO CO -cj- £— 03 CO 03 O Hf 03 lO LO Os lOHI>^ ^ H05t>0 ^ ^qCOO »0 zQ CO 03 -fa ’ — < Co TfLOHrJ< 05 CO ^ CO Go -fa 03 03 -fa G-?, LO lo ^ ^ L^ lO “ -Q P 05 ■2 H ■ -» P la .2 > T3 S.£ m & “ O 03 g ft ^ S w X 4J 05 P „ 8g £ 00 a - 05 W pH P p rP 0> W c Cp P <+H 0 3 c -o O s o _g 8 £ 2 .2 £ T3 _ T3 2 £ rP 03 p 0 -fa t— 'fa CO 00 >H CD CD "fa t-H Oi 03 O 03 O i-H OH-fa03 >H 03 03 00 00 §3 00 w g -fa LO 'fa LO ^3 'fa CD CO LO LO LO LO LO LO CD 'fa 'fa CD LO LO LO LO "fa LO -fa Xfl 0 Ci LO 03 fa 5 Oh LO ^7- LO LO r ^ fa fa n 05 CO C3 00O r- 00 LO -fa O 03 O 0 03 CD 03 'fa 03 C5 00 CD CD O O} 0 oPh P LO 'fa LO LO -fa CD 03 CD ^ 03 LO LO CD 03 CO C— LO LO LO LO LO lO fa fa U 05 Ph 03 C3 'fa 03 0- CO 03 CD C3 Oo I O CO CD Co LO t-H [>■ T—i CO 0^030 T-H I 'fa -fa LO LO LO ■fa CD 03 LO ■<*■ 1 CO LO 'fa CD CD 'fa LO LO LO LO LO LO LO 1 fa O 05 y >H 'Co >H 05 LO LO ^7- LO LO fa Ph rH O LO LO C- Oi O L- 03 00 I OOH 03 03 LO 03 os 'fa 03 CD CO t-h I 'fa LO 'fa LO LO CO LO LO | CO LO CD LO LO CO 'fa ^7- LO x, 'fa LO LO £ Q OOOO^ ^ 00 cDci ^ £ < ffi o „ K ^ 5 £ oqoqo t-h ^ H CO 'fa 'fa 'fa 4 T-j CO 00 00 °0 cd lo cd -fa id cq rjjqoo >h LO LO LO -fa LO 00 03 03 00 O ZD I> t> ZD hi cq zd q 03 00 LO LO* 'fa 'fa -d ZD 00 ZD 'fa LO LO © "fa © LO LO -fa GO id °o *4 03 *fa 'fa "fa LO t> t> o CD oq 03 Lq LO 'fa CO "fa -4 00 ^ cq oq Lq rfa ZD LO LO LO Tf q q o oq ■-fa LO "fa LO ^ ZD 03 03 -q- lo so c4 cc O NOq 06 06 06 00 cq -rq (5^ t> 06 06 06 00 rf cq cq ^ t> 06 t> ^ O CO CD 03 00 00 00 00 Oo W fa CQ < H bjO o ?h r^l S “ >i pH Ph 3 C3 60 g qz; .S HH rfa A o c3 q w5 Q d fa r< CO O p fa 05 < fa OJ <3 fa PQ CQ P Xfl 03 03 03 Go L'- L~ L — d Tt O CD O Oo CD O 00 CD ^ CDOOLOD O* OOfafaW oq 03 03 03 03 ©4 t> C> IT- ^ GO CD 'fa O OMOO ^ CO 00 CD 00 t: — ir — ig — ir — Jq 03 "fa1 'fa L- d id 03 r— I 03 LO t-h 00 03 CD t- 00 O O CO 03 -<> fafaLOCD Co ^ 03 o 03 O) td L-- l>- d N. LO O 00 oo 00 O 00 id 00 03 CD 03 — | 03 [>00^ qJ 03 03 03 Ifa l> I> I> OOOOt- s, 03 O 00 00 00 *0 t-H •O t-H ch 0 05 00 0 5:, H fa g o ^ Xfl fa g§ Q fa o tc | "fa ‘fa M zn d 03 c3 a a; o> q s” n S o •r* o> pH 0) Ph Tfl cjfa oT g . r— I r*1 c^3 S c c3 h* r-1 S'Mft g S 6^ g .§ 8 tS V h Ph ZO j"g 2 tS- c3D OTi 03 03 03 03 O 05 T3 <3 a3 o3 05 s fa >» Cv a3 K* ^H X— s fa '■p c q 0 fa .rft a m fa 05 o cum a3 c3 2 i-J*? 'faO c3D C5T! CO CO CO CO 05 '-a <3 c3 05 c/2 c^3 o q q fa ri.fa ^ >-g 05 C^3 C3 If} ° r-1 as o .§ 8 fa1 rH 05 Ph^ Cp c3D C5p -fa 'fa 'fa Ttf P P 05 .q h-> C aS O £ £ O g'S# hSS as . > «- 2 r3- 6^ 2 ^ as fl P o fa I_J -p o rP q 'faO- M3 fa f P P ^ 05 q £ p 05 fa -4-> O | g 2J § q rH 05 p p g <3 05 fa Ph P P 'P m3 dr p 2 ^ 2 s fa 05 g H q; Pl, CO o £ co OJO 2 p >i a3 fa-O S P £ o aSP gu LO LO LO LO 1 Fish showed distress. 2 Fish showed distress, with some mucus coagulation. "> Fish showed distress, one fish died and experiment was discontinued. 1 Fish showed distress, three fish died, just at end of experiment. 1943] Breder & Rasquin: Mexican Blind Characins 175 Table II. Experiments with Attractants. Each figure concerning the distribution of fish represents 100 observations, the primary means, 400 and the secondary means, 800, except as indicated. Exp. Fish used Substance Cc. no. AND NO. and % SOL. used Sub. Blank 7a 100 86 7b 4 blinded Cane 100 96 7c river fish sugar — 40% 100 96 7d 120 93 Primary means 105 93 Secondary means 8a 70 71 8b 4 La Cueva Cane 90 93 8c Chica fish sugar — 20% 93 87 8d (normal) 100 90 Primary means 88 85 Secondary means 9a 84 99 9b 4 La Cueva Cane 70 56 9c Chica fish sugar — 40% 96 62 9d (normal) 92 68 Primary means 85 71 Secondary means 10a 100 96 10b 2 La Cueva Cane 93 69 10c Chica fish sugar — 40% 98 79 lOd (anosmic)1 100 86 Primary means 98 82 Secondary means 11a 100 76 lib 4 Cueva de Cane 100 100 He los Sabinos sugar — 40% 85 66 lid fish (normal) 100 100 Primary means 96 85 Secondary means Percentage of Fish in Substance pH at start and Compartment Periods of Observation Pre-drip Sub. drip. Post-drip END2 1 2 3 4 5 6 7 8 7.0 51 53 55 52 53 45 53 62 7.0 52 46 52 48 60 51 64 62; 7.2 45 45 49 51 51 43 47 34 7.2 49 53 50 50 55 53 44 51 7.1 50 49 49 52 51 50 55 51 4-8 52 52 52 6.8 51 45 53 50 56 53 7.0 52 54 46 53 53 52 54 53 7.0 50 52 54 55 62 53 48 45 7.0 54 52 52 54 58 56 56 59 6.9 52 52 53 51 51 52 56 54 53 54 53 52 7.0 60 51 58 69 76 76 73 78 7.0 50 48 52 52 54 52 53 55 6.8 47 43 43 45 48 49 42 43 7.0 50 48 53 55 53 ■58 54 57 6.9 52 50 47 51 53 55 58 58 59 55 57 58 7.2 51 51 51 49 51 55 53 51 7.2 50 51 49 49 50 53 56 52 7.1 50 52 48 49 55 52 61 58 6.8 50 53 50 57 57 50 53 62 7.1 50 51 52 49 50 52 53 53 53 56 56 56 7.2 28 45 56 51 65 67 61 53 7.2 47 37 45 41 40 49 55 49 7.2 50 49 51 54 60 62 54 50 7.4 46 49 35 34 35 39 29 31 7.3 43 44 45 47 46 45 50 52 54 50 48 46 1 Each part of this experiment represents the mean of two experiments, combined here for purposes of this table, as to be comparable with the other experiments. 2 Cane sugar would not change the pH values but they were checked at the end of each experiment for possible variations due to unknown causes. In all cases the pH values did not vary. similar manner and serves to indicate the extent of the normal departure from the expected 50 percent, distribution that is found in these fishes by means of the four pre-drip or control periods in each case. Graph C compresses the data of graph B to double periods and to that extent acts as a smoothing method. In other words, in this graph all the consecutive observations are treated as a unit. This serves to give a more general representation and reduces the incidental details more fully expressed in graph B. Text-figure 5 show graphically the longer term experiments, the data of which are given in Table III. It is evident from Tables I and III that some of these experiments ran to the point of lethality. These effects are indicated in foot notes to the tables. The word “dis- tress” indicates that the fish began moving at higher than normal speeds with their snouts out of water and . gave every evi- dence of being in serious difficulty. In cer- tain cases the mucus became whitish as though coagulated and fell away in small pieces. In all, four fish died either during or immediately after experiments, as is in- dicated. All the rest survived and at this writing are still living. They gave evidence of rapid and complete recovery in all but the four cases noted. The significance of these graphs is treat- ed in the discussion. Chemical Attractants. Experiments with attractants were han- dled exactly as were those of the chemical repellents. The basic data of these are given Text-fig. 3. Reactions to chemical repellents. Graphic representation of the data of experiments 1 to 5 inclusive, based on Tables I and IV. Exp. 1 — Blind river fish, 10% acetic acid. Exp. 2 — La Cueva Chica fish, normal, 10% acetic acid. Exp. 3 — La Cuevn Chica fish, anosmic, 10% acetic acid. Exp. 4 — Cueva de los Sabinos fish, 10% acetic acid. Exp. 5 — La Cueva Chica fish, normal, 10% ammonium carbonate. Graph A: distribution of fishes during periods of the dripping administration of substances. Each point repre- sents 100 observations. Graph B : distribution of fishes before, during and after the dripping administration of substances. Each vertical division four times the size of those in graph A. Each point represents 400 observations. Graph C: distribution of fishes as in graph B but with all consecutive observations represented as a single point. Each point represents 800 observations. OBSERVATION PERIODS 1943] Breder & Rasquin: Mexican Blind Characins 177 in Tables II, III and IV, and the remarks under the previous head apply equally here. Dissolved prepared dry fish food, the juice of horse meat and cane sugar were used in these experiments. As no statistical differ- ence was suggested in the early work, cane sugar was used for most of the experiments because of ease in standardizing. A digest of these data is given in Text- figure 4 and is handled in a manner exactly comparable to that of the repellent agents. Text-figure 5 shows graphically the data of Table III. Table V gives data on the circling move- ments of normal, anosmic and unilaterally anosmic fishes from La Cueva Chica stock. These experiments were performed in a simple trough. The stimuli, prepared dried food and cane sugar, was given in the solid form by allowing a small quantity to pass down a large glass tube with its end held near the floor of the trough at a time when the fish was at the far end of the trough. In each case a single fish was employed at a time. This was done because it was evi- dent that the excited movements of a feed- ing fish served to attract the attention of others. This may be a lateral-line effect or could be due to the sound of chewing. The chewing sounds of the pharyngeal teeth of larger species may be heard quite distinctly through a simple “submarine listening tube” made of a stethoscope over the mouth of which is placed an ordinary unperforated nursing nipple. The essential similarity in the behavior between the normal fish and the fully anos- mic in respect to right and left hand turns as compared with the semi-anosmic is indi- cated in Text-figure 6. In conducting these experiments two fish were found, one in the normal group and one of the fully anosmic group, which showed a strong bias to circle to one side. Evidently there was some asym- metrical influence operating in the sensory mechanism of these individuals. For this reason they were omitted from the main body of the data and are so indicated in Table V and Text-figure 6. The significance of these data is treated in the discussion. Calculation of Results. The numerical data obtained in the be- havior experiments are capable of being analyzed in a number of respects additional to the simple estimation of the quantitative nature of the repellent or attractive stimu- lation of the chemical quantities employed. Considering only the control periods which preceded the experiments, it is de- sirable to examine them in order to deter- mine if, in spite of experimental precau- tions, any bias existed in the trough or in the behavior of the fish that might have a bearing on the results of the experiments. Since it is expected that if there were no bias the fishes should be moving at random, and since in each case here considered there were four specimens in an optically divided trough of two like compartments, it follows that the numerical values of the observa- tions should approximate 50 percent. The actual values observed for each of the four types of fishes used are given in Table VI. None shows any statistically significant de- parture and the mean of all was found to be 49.88+, that is, the fish were at one end of the tank as often as they were at the other. Another measure of the presence of a possible bias is to consider the frequency of occurrence of the possible aggregation numbers as compared with the calculated values of the binomial distribution of four items. Since in any observation at one end there can be any of five possible numbers of fish, 0, 1, 2, 3 or 4, it follows from the equation for the binomial formula (p+q)n, expanded for each term, that the observed occurrence of 0, 1, 2, 3 and 4 fish together at one end at one time should be 1, 4, 6, 4 and 1 or one chance out of 16 that there will be none, 4 out of 16 that there will be one and so on. The observed values are given in Table VI, from which it is clearly evident that they closely approach the cal- culated expectancy: 6.25, 25, 37.5, 25, 6.25, being 7, 25, 36, 25, 7. Since the experiments were all carried out using first one and then the other end of the trough, any bias in it would appear if the values obtained at one end be compared with those of the other. Designating one end as “A” and the other as “B,” this comparison is also made in Table VI. That there is no significant dif- ference from one end to the other is evi- dent. In this same table there is- also given a comparison of all the observations for each form studied. These calculations are shown graphically in Text-figure 7, both with regard to each form of fish and with regard to the trough end. The slight diver- gencies are clearly indicated. Only in those experiments involving small numbers of ob- servations are any notable divergencies from the binomial values to be seen. Such as are present have a distinct bearing on the problem, but only become clear in the following further analysis. It was evident during these control pe- riods that each type of fish did not act ex- actly as every other in respect to their attitudes to their fellows. The above calcu- lations give scant indication of this, how- ever, since whether the fish were acting strictly independently or with some slight degree of unanimity could hardly be expect- ed to show clearly in such a treatment. If the fish were fully tied together, optically or by other means, such behavior would, of course, be evident at once from these fig- 178 Zoologica: New York Zoological Society [XXVIII: 20 Table III. Long Term Experiments. Period numbers continue from Tables I and II. This is a continuation of the experiments indicated of those tables and the notation is exactly the same. Experiment Percentage of fish in substance compartment NUMBER 9 10 11 12 13 14 15 16 17 18 19 20 La Cueva Chica 2a (acid) 31 2 24 22 35 25 20 30 24 26 11 22 2b 30 36 511 65 51 46 40 44 39 43 51 53 2c 2 0 19l 8 39 34 40 45 52 41 55 62 Pri. means 12 13 31 32 A2 35 33 AO 38 37 39 A6 Sec. means 12 32 39 37 38 A2 La Cueva Chica 8a (sugar) 51 56 36 44 51 55 41 40 47 40 Sec, means 5A AO 53 AO AA River Fish la (acid) 44 44 452 50 54 68 Sec. means AA A9 61 Difference in minutes between start of readings. Even numbered readings continuous with preceding odd numbered periods. Odd numbered periods in Tables I and II started about 30 minutes apart, with the following even numbered periods continuous with them. Experiment NUMBER 1 3 5 7 9 11 13 15 17 19 2a 25 40 25 40 65 55 60 60 60 2b 30 40 30 30 75 45 60 60 60 2c — 25 40 30 30 70 50 60 60 60 Means 27 AO 28 33 70 50 60 60 60 8a — 45 30 30 70 50 60 60 60 la — 25 40 30 30 65 55 i Fish showed distress. - Fish showed distress, with some mucus coagulation. ures, for instead of a binomial distribution of four items one should obtain a distribu- tin of one item. In other words the figures would indicate the behavior of a school, of four fishes in this case, which under the conditions would give a 1 to 1 distribution value. If the linkage were not as perfect as this case, which might be expected to obtain in a school of herring or mackerel, the values should approach one of the in- termediate binomial distributions. Thus with four fish it could be possible to have any of the following frequencies depending on the cohesiveness of the group : 1-1 ; 1-2-1; 1 -3-3-1 ; 1-4-6-4-1 displaced on the scale of 0 to 4 appropriately. That there is no evidence of a tendency toward the first three possibilities is apparent. The follow- ing tabulation of these possibilities and their significance should make this entirely clear. Possible FREQUENCIES SIGNIFICANCE 1-1 1 aggregation. Four fish stay close together. 1-2-1 2 aggregations. Groups of 2 and 2 or 3 and 1. 1-3-3-1 3 aggregations. Groups of 2, 1 and 1. 1-4-6-4-1 4 aggregations. Four “groups” of 1 fish each. Attacking the problem in another way, it became apparent in the original pages of data that in certain cases consecutive obser- vations showed the same number of fishes together for more or less extended time. These sequences of consecutive observations of certain numbers of fish indicated that when the fishes came close together they tended to remain there, which gives some measure of a slight and otherwise scarcely detectable aggregating tendency. This tend- ency appeared to differ in the various forms. Although this part of the analysis was under- taken because of what appeared to be be- havior differences noted during the carry- ing out of the experiments, the statistical analysis led to a different interpretation of the condition than that based on simple ob- servations. The mathematical analysis of the occur- rence of these sequences of observations of the same number of fish is slightly more complicated than the preceding study of the distribution of the aggregations but is a further rendering of the binomial equation. Since the likelihood of a given event re- curring in sequence in such a study is pn, it is possible to compare the observed se- quences in regard to both their duration in time and number of times they appear in the observations to the probability values obtained for each of the five possible num- 1943] Breder & Rasquin: Mexican Blind Characirth 179 Table IV. Short Term Periods During the Administration of Substances Each of these periods are one-quarter the length of those of Tables I and II. They are indicated as lettered parts of periods 5 and 6 of those tables. Each figure represents 25 observations, the means 100. Exp. No. 5a 5b Drip 5c 5d PERIODS 6a 6b 6c 6d Exp. No. 5a 5b Drip periods 5c 5d 6a 6b 6c 6d la 56 57 27 17 13 6 3 2 7a 50 56 52 54 56 40 34 48 lb 46 16 19 10 13 3 2 2 7b 68 54 50 67 56 54 44 52 lc 52 50 43 25 11 15 18 7 7c 60 46 52 47 45 49 52 27 Id 47 21 8 3 40 37 11 0 7d 52 59 57 50 55 52 49 56 Means 57 36 24 n 19 16 8 3 57 .54 53 54 53 54 45 U6 2a 56 . 37 49 31 17 18 15 7 8a 56 56 54 59 44 49 64 54 2b 48 60 63 39 19 9 5 5 8b 52 61 50 48 54 41 58 54 2c 33 13 2 12 9 11 9 9 8c 59 69 61 59 55 56 54 46 2d 60 64 31 23 4 10 5 9 8d 53 62 58 59 52 49 53 71 Means 4.9 44 36 26 12 10 8 7 55 62 56 56 57 47 57 57 3a 54 47 70 75 65 73 49 38 9a 75 66 85 77 77 81 80 67 3b 75 42 31 17 19 30 17 8 9b 56 57 46 57 57 57 44 49 3c 50 40 24 35 26 40 37 35 9c 50 51 42 47 46 56 45 50 3d 53 50 45 26 31 31 24 24 9d 58 56 47 59 59 59 56 59 Means 58 45 42 38 38 US 32 42 60 57 56 57 60 63 56 56 4a 54 50 33 27 15 3 6 9 10a 42 53 54 56 56 54 52 58 4b 53 40 16 8 17 5 8 6 10b 43 58 49 51 33 59 56 66 4c 34 32 26 16 16 11 13 11 10c 56 50 62 55 56 55 52 43 4d 76 49 24 12 12 4 4 4 lOd 58 49 60 60 48 44 61 50 Means ■54 42 25 16 15 6 8 7 50 52 56 55 U8 53' 55 52 5a 55 64 48 49 42 39 41 52 11a 66 69 60 64 66 63 63 77 5b 62 60 60 53 41 47 41 48 lib 48 28 38 46 34 43 53 64 5c 80 74 68 56 68 55 35 52 lie 50 57 57 75 66 65 64 55 5d 38 36 44 33 27 27 48 33 lid 33 33 35 38 48 30 37 41 Means 59 58 55 4 8 44 42 47 U6 49 4 7 42 56 53 50 52 5.9 bers of fish in a group occurring n times in succession. Divergencies from the calcu- lated values of significant magnitude then appear in two ways: (1) as numbers of times that a sequence of any given length occurs above expectation and (2) actual length of the longer sequences compared with the probability of such a length of se- quence occurring at all. The data so ob- tained for the various experiments of each possible observation in regard to the occur- rence of sequences of various lengths is given in Table VII. In this table each ex- periment is indicated together with the number of times of occurrence of each se- quence for every length of sequence ob- served. In Experiment 1, for example, no fish were at one end of the trough for one observation only 52 times, for a sequence of two observations 17 times and so on, while all four fish were together for one ob- servation only 54 times and for two obser- vations 24 times. Since, obviously, for each set of readings, given in this table, there is another, at the opposite end of the trough, which represents the difference between the total number of fishes, four, minus those indicated in Table VII, it is possible to combine these values and reduce the data to the form given in Table VIII in which only three sets of figures for each experi- ment need be handled. Reducing the calculated probability val- ues to terms in which the first power is equated to unity for each of the different kinds of combinations possible (2 fishes, the mean of 1 and 3 fishes and the mean of 0 and 4 fishes), it is possible to compare these calculated values directly with the observed values by equating the number of single observations (sequences of one) to unity. These figures, based on both observed and calculated values, are given in Table IX, in which the calculated values for each class of data are directly comparable to the ob- served values for that class. The evident differences show the increase of the ob- served values both in number of sequences and their length beyond expectancy. It is evident that while the river and Cueva de los Sabinos fish are closely similar and differ only slightly from binomial ex- pectancy, La Cueva Chica material shows a greater divergence and the anosmics show extreme variance. If the calculated values are subtracted from the observed, then these differences, if positive, show the ex- tent of the observed over the expected and, if negative, the reverse. These values are given in Table X and graphically shown in 180 Zoologica: New York Zoological Society [XXVIII: 20 «0 N-* lO in * OS id m m O Is W £ © c < > Pi W EG C2 CO o w > s 00 p 0 w m 1 - o Pu o OS CO CO (M CO I I H Ol C£> H 1— iLOOlOri CO CO 00 ,-H LCD CO (M eg TJ1 | If X O ffl X li'LO rH t— ( tH t— I H H tH t-H i— I i— ( i—t OQ rH < — I o a a M O H N 03 © — O] CO O HN CO -f O H N CO f © r— I N CO f O H N CO f ID o — in 03 f ©1-11 55 E a. d X £ 00 213 1943] Breder & Rasquin: Mexican Blind Characins 187 ©H H W H I© H Tf Q CO COOHIO W i lO H H H CO 1C W©N t— ( CO 05 CO 1— I N O] H 'Ot- r}< lO t> COHLOHt- )OH ^ 05 (M d SO t- 1C i 00 t- t-OOO CO 05 h © i t- LQ C0©U5 OOt-OOlO U H COCOCO H (M H C/3 z o ►H H a 23 B H C/2 w w ffi 2 CO fa rH fa o PS W d CO tH s 23 z tH S3 o 0 si 03 W to a O a o to w u z w D O’ W W Q U 2 s s o u w J a < H w 00 U) CO c fa > *> fa CJ> fa CO <£> 5? o fa fa o _ 1C w fa CO fa £ <3 p< X W d ^ O X) fa '3 ^ IC P3 LO LC co CD 'Tf ic ic ?0t>© O T_- 1 ^ CO *H 1C 02 © 1C CO H © 05 ^ oo CO 00 " lO N C-Ort rH a x H os a iS oj IM u IS o !? ho® > to CO HCD IC o rH ac 4 ood ic IC 1C 1C 05 co 05 9—1 tP © T3 WH fa > 3 ^ O cl 00 rH SC ^ rH 1C 1C rji d 1C o co 1C o i> CD ^ CO CO iq 05 t- 00 00 05 4-H CO Cl rH ICOtJ! OH(M O H Cl CO rf fa O o CO ^ I I (MHO fa fa fa 03 O § CO ^ MHO fa fa fa fa 0> 05 ss Table VII. Observed Distribution of Sequences. Number of consecutive observations for each number of fishes 5 fi 7 8 9 10 11 12 13 14 16 20 Table VIII. Combined Sequences of Observed Distributions. Number of consecutive observations for each number of fishes Mean 0 - 4 406 171 317.5 125 106.5 37.5 665 513 168 389 297 118.6 Blinded river fish (Exp. 1 & 7) 65 30 12 7 1 45.5 16.5 4 1 3.5 1.6 0.6 La Cueva Chica (Exp. 2, 5, 8 & 9) 21 5.5 Anosmics (Exp. 3) 15.5 12 4.5 21 13 13 33 0.5 Anosmics (Exp. 10) 96 59 34 37.5 8.5 _ 0.5 — — 17 — 0.6 154 104.5 32.6 Cueva de los Sabinos (Exp. 4 & 11) 62 21 11 44 — 48 14.5 8 6 2 0.6 11.5 6 2.6 1.6 0.5 0.5 — — 0.5 — — — — — 0.5 Zoolo O) — — — — “““ Text-fig. 7. Dates of description of the 106 species of sunbirds recognized in the present list. Mayr have as usual been of the greatest help, and have kindly read my manuscript. I thank them most sincerely. My experience with sunbirds in life has been extensive. I probably have kept more species in captivity than any other avicul- turist; and I have collected and observed many forms in the wild state, including a number not previously described, mostly in Indo-China but also in India, in Madagas- car and in various parts of Africa. This should justify my endeavor to revise the family in the light of recent researches. I. Genus Anthreptes. Tongue ending in two half-cylindrical lobes. Bill slightly curved, equal to the head or shorter. Nasal operculum naked. Mandi- bles finely serrated or only roughened. First primary short. Both sexes sometimes with metallic colors, or both without; or males with, and females without. Pectoral tufts present in all males except in gabonica and platura, less often in females. The plainest species, gabonicus, is olive brown above, whitish below, with a pale Text-fig. 8. Genera of the Family Nectariniidae. 22 Zoologica: New York Zoological Society [XXIX: 4 eyebrow; it probably represents the most primitive sunbii'd now existing. Other spe- cies can be divided into three groups: two African and one Indo-Malayan, as shown in the diagram. The less specialized forms of each group have a duller plumage and a longer tail. In the most elaborate species, platura, the two central rectrices of the male in breeding plumage are narrow and greatly elongated. With the exception of two species ( ga - bonicus and fraseri ) all males show metallic colors. Except in 'reichenowi, rectirostris and platura, the females of all other Afri- can species have some metallic feathers, dif- fering only slightly from the males in sev- eral cases. In all the Indo-Malayan species, males have metallic feathering but females have none. In simplex, however, the metallic pur- ple-blue is reduced to the forehead of the male. Anthreptes is found over most of the range of the family, from Africa to the Philippines and Celebes. Some of the spe- cies inhabit rain forest, others, dry jungle and gardens. Text-fig. 9. Species of the Genus Anthreptes. 1. A. gabonicus. Brown Sunbird. Gambia to the Lower Congo, on the coast. 2. A. fraseri. Scarlet-tufted Sunbird. Upper Guinea to eastern Belgian Congo and Uganda. 3. A. reichenowi. Gunning’s Sunbird. Coast of Kenya Colony to Portuguese East Africa (lower Zambesi Valley). 4. A. anchietae. Anchieta’s Sunbird. An- gola to Northern Rhodesia and Mozam- bique. 5. A. simplex. Plain-colored Sunbird. Ma- lay Peninsula, southern Burma and peninsular Siam; Sumatra, Borneo. Nias and Natuna Islands. 6. A. griseigularis. Gray-throated Sun- bird. Philippine Islands. 7. A. malacensis. Brown-throated Sun- bird. Southern Burma, Siam and Indo-China; Malay Peninsula, Su- matra, Java, Philippines, Celebes and neighboring islands. 8. A. rhodolaema. Rufous-throated Sun- bird. Tenasserim, Malay Peninsula, Sumatra, Borneo. 9. A. singalensis. Ruby-cheeked Sunbird. Burma, Indo-China, Siam, Malay Pen- insula, Sumatra, Java and neighboring islands. 10. A. longuemarei. Violet-backed Sun- bird. Savannas of the Ethiopian re- gion from Senegal to Abyssinia, south to Angola, Mashonaland and Portu- guese East Africa. 11. A. neglecta. Uluguru Sunbird. Eastern Tanganyika Territory. 12. A. aurantium. Blue-backed Sunbird. Forested Cameroon, Gaboon and Bel- gian Congo. 13. A. pallidig aster. Moreau’s Sunbird. Northeastern Tanganyika Territory. 14. A. rectirostris. Shaw’s Sunbird. West Africa, from Gambia to Fernando Po and the Congo, Uganda and Usambara in Tanganyika Territory. 15. A. collaris. Collared Sunbird. The greater part of the Ethiopian region from Senegal and Kenya Colony south to eastern Cape Province. 16. A. platura. Pygmy Long-tailed Sun- bird. Senegal, east across the Sudan to Philae on the Nile, Abyssinia, Erit- rea, Somaliland and southwestern Arabia, south to the northern Congo and northern Uganda. Notes: A. axillaris, from Uganda and the Bel- gian Congo, is certainly conspecific with fraseri from Fernando Po and West Africa, differing only in its gray instead of olive head. A. yokanae Hartert (Bull. B.O.C., XLI, 1921, p. 63, Rabai, N. of Mombasa) is either identical with or a very indistinct race of A. reichenowi Gunning (Ann. Transvaal Mus. 1, 1909, p. 173) from the lower Zam- besi Valley, Portuguese East Africa. I could not examine the latter, but Grant and M.- Praed share my opinion (Bull. B.O.C.., LXIV, 1943, pp. 11-12). 1944] Delacour: Revision of the Sunbirds 23 A. griseigularis, A. malacensis and A. rhodolaema, although very similar, must be considered separate species, as two of them coexist in many localities of Malaysia and the Philippines. A. singalensis is certainly related to the three preceding species, which it closely recalls in its plumage pattern. But it ap- pears to be a regressive species in that its bill is particularly short, almost straight, and devoid of real serrations, the edges of both mandibles being only roughened. I do not think, however, that such a character warrants the separation of this species in a special genus (Chalcoparia) , even less in a special family, as Stuart Baker proposed (Fauna of Br. India: Birds, III, 1926, p. 368). It is much too near other Anthreptes in every other way. Serration is very variable in sunbirds, as the microscopic examinations of many hundreds of speci- mens has convinced me. It is fairly constant in each species, but only within certain limits. It can differ in closely allied forms. Such a character cannot therefore have much taxonomic meaning. A similar state of things is found in the African species of Pycnonotus of the subgenus Andropadus (cf. Revision of the Pycnonotidae, Zoo - logica, XXVIII, 1943, p. 18) to a much higher degree. All other species of Anthrep- tes have weak serrations, sometimes almost invisible, particularly in some specimens of platura and malacensis. The peculiar sun- birds of the genus Neodr epanis, long said to lack serrations, have in fact none on the upper mandible, but some very small ones on the lower. For all these reasons I in- clude singalensis in the genus Anthreptes. I never found, as Stuart Baker asserts, that it differs in any way in habits from other sunbirds. The question of the Violet-backed Sun- birds in East Africa has been discussed by J. Vincent ( Ibis , 1936 pp. 72-74) and by R. E. Moreau (Ibis, 1937, pp. 335-337). It seems that the best solution is to recognize A. neglectus, in which the female possesses an entirely metallic mantle, as a full species, and to consider all other forms, including orientalis, as subspecies of A. longuemarei. It does not seem probable that the ranges of any of them really overlap. If they do, orientalis will have to stand as a full species. I have not been able to examine specimens of A. pallidi g aster , but from its description it seems to be not very far from A. aurantia. A. tephrolaema is certainly a southern race of A. rectirostris, only with less yellow pigment. A. platura is the most specialized species of the genus, but is fairly closely related to A. collaris, as shown by the color pattern of the males. The bill is very short and only indistinctly serrated. As it is found in other genera of sunbirds (N ectarinia and Aetho- pyga), the lengthening of the tail of the male is but an extreme specific differentia- tion. A. metallicus is merely an eastern race of platura, with the purplish-blue color ap- pearing on the breast and extending farther on the back. II. Genus Nectarinia. This genus, by far the largest of the family, differs from Anthreptes mainly in its longer and more curved bill. It is there- fore more highly specialized and, contain- ing a far greater number of species, it in- cludes forms of more varied and elaborate color pattern. On the whole, however, it follows closely the same variations as An- threptes. In the majority of species the nasal oper- culum is naked, but in a small group it is covered to a varying degree with short metallic feathers like those of the forehead. The bill is always serrated, but more or less finely and deeply. It varies greatly in length and curvature, even in subspecies of the same species. In the more primitive forms both sexes are dull and alike, as in Anthreptes. Only in a few slightly less primitive ones, females show metallic colors like the males, or are slightly duller. But in all the more advanced species which form the very great majority, males always possess metallic colors, while females lack them. In N ectarinia the tail varies from short and square, with or without elongated central rectrices, to strongly graduated. Its length often varies greatly among races of the same species. As I have noted above, it is possible to distinguish four subgenera in the genus Nectarinia; and in the larger of them, we can group species for practical purposes according to the plumage pattern of both sexes. The adoption of these minor group- ings makes it easier to understand the par- ticular relationships of a large number of species which, however, are not sufficiently differentiated to be considered as forming genera. Text-fig. 10. Subgenera of Nectarinia. 24 Zoologica: New York Zoological Society [XXIX: 4 Subgenus Cyanomitra. This subgenus is composed of primitive species which have several characters in common. The plumage of the more general- ized forms is olive all over, darker above; then gradually there appear yellow or red pectoral tufts, metallic fringes to some feathers, and finally patches of metallic colors. But these solid patches are always dark purplish-blue or green, and they are confined to the head, throat and upper breast, one aberrant species of uncertain status ( hypogrammica ) showing purplish- violet on the nape and upper tail-coverts. Except for pectoral tufts, the only bright non-metallic color present is yellow; but brown, gray and olive are always dominant. In the majority of species, females re- semble males, but in four ( hypogrammica , hartlaubi, newtoni, cyanolaema ) they are almost as strikingly different as in the other subgenera. In verticalis, the female also differs from the male, but retains a metallic cap. These better evolved species resemble some of the more advanced groups. N. cyanolaema, although close to verticalis, plainly recalls N. ( Chalcomitra ) fuliginosa in both sexes, while N. hartlaubi and N. newtoni are not very far from N. ( Nec - tarinia ) dussumieri and N . (N.) jugularis. In the majority of species of Cyanomitra, however, the bill is proportionately coarser, broader and deeper at the base. But it never is so thickened as in Anthreptes. Pec- toral tufts are present in males, except in those of seimundi, batesi, hypogrammica, hartlaubi, newtoni and thomensis. Females of urstdae, verreauxi, reichenbachi and oritis are also tufted. In olivacea, they have tufts or not, according to the subspecies. All species of the subgenus Cyanomitra are African, with the exception of one from Socotra and one from Indo-Malaysia. 1. N. seimundi. Little Olive Sunbird. Liberia to Fernando Po, east to Uganda. 2. N. batesi. Bates’s Sunbird. Fernando Po, southern Cameroon, and Belgian Congo. 3. N. olivacea. Olive Sunbird. Practically the whole Ethiopian region, in forests. 4. N. ursulae. Little Mouse-colored Sun- bird. Fernando Po. 5. N. verreauxi. Mouse-colored Sunbird. South and East Africa, from eastern Cape Colony to Lamu Island. 6. N. balfouri. Socotra Sunbird. Socotra Island. 7. N. hypogrammica. Blue-naped Sun- bird. Burma, Siam, Indo-China, Malay Peninsula, Sumatra, Natuna Is., Borneo. 8. N. reichenbachi. Reichenbach’s Sun- bird. West Africa, from Gold Coast to Gaboon and western Belgian Congo. 9. N. hartlaubi. Principe Sunbird. Prin- cipe Island. 10. N. newtoni. Newton’s Sunbird. Sao Thome Island. 11. N. thomensis. Sao Thome Sunbird. Sao Thome Island. 12. N. oritis. Western Blue-headed Sun- bird. Highlands of Fernando Po and Cameroon. 13. N. alinae. Blue-headed Sunbird. Ru- wenzori and mountains of the Kivu District. 14. N. verticalis. Green-headed Sunbird. Senegal to the Congo, Uganda, south- western Tanganyika Territory and Nyasaland. 15. N. cyanolaema. Blue-throated Sunbird. West Africa from the Gambia and Fernando Po, east of Uganda. Notes: N. seimundi is undoubtedly the most primitive species of the genus, and one of the most primitive of the family. It has been considered an Anthreptes ; but on account of its slender, rather long bill, of its similiarity to N. batesi, and of its strik- ing differences in proportions from the primitive Anthreptes gabonica and A. fra- seri, it is preferable to place it in the pres- ent genus. N. batesi, which is found in the southern Cameroon alongside N. seimundi, is so simi- lar as to be difficult to distinguish. However, it has a slightly longer and more curved bill, is a little deeper in color, and the inner parts of its rectrices are darker.^ This is another cajse of two very closely' allied species living in the same territory. N. ursulae is an island representative of N. verreauxi, much smaller and approaching olivacea in its greenish back. I consider it a full species. [The spelling of verroxi, being evidently an error in transcription of Verreaux’s name, must be altered to verreauxi .] N. hypogrammica, as stated before, is a very puzzling species, not closely related to any other. The place given it here is purely tentative. Because of its general coloration and of its apparently primitive nature, it fits better in Cyanomitra than in any other subgenus. N. balfouri is also peculiar in its pale gray, streaked and spotted plumage, but the dark marks of its feathers seem to represent an alteration of a metallic pattern through adaptation to desert conditions. It is not really very different from other forms of Cyanomitra. 1944] Delacour: Revision of the Sunbirds 25 N . hartlaubi and N. newtoni are island representatives of N. reichenbachi, but they show further specialization in the very dif- ferent plumages of the two sexes. They re- call the color pattern of N. jugularis, from the Indo-Malayan-Australian regions. N. thomensis, remarkable for its large size, long graduated tail and purple-edged plumage, represents a much transformed offshoot of the same group, the product of an early invasion. N. newtoni, very small and different in pattern, is probably a later invader of Sao Thome which has evolved in an opposite direction. The last four species : oritis, alinae, verti - calis and cyanolaema, show a gradual evo- lution toward the more advanced groups. Subgenus Chalcomitra. The sunbirds of this group are easily dis- tinguished from all others by their rather long bill, short, square tail, long body, and by the general velvety brown or black color of their plumage. Metallic colors are con- fined to the crown, throat and chest, wing and upper tail-coverts. Females are rather dark, with mottling or stripes on the breast. According to climatic conditions, they may or may not be tinged with yellow under- neath. The nasal operculum is naked. The most primitive species, fuliginosa, has yel- low pectoral tufts, in the male only. All five species are African. 16. N. fuliginosa. Carmelite Sunbird. West Africa from Senegal to the west- ern Congo. 17. N. rubescens.* Green-throated Sun- bird. Fernando Po, Cameroon, Gaboon and N. Angola to Uganda and Kavi- rondo. 18. N. ametliystina. Amethyst Sunbird. Southern and eastern Africa from Cape Province to Angola, northeastern Uganda and Lamu Island. 19. N. senegalensis. Scarlet-chested Sun- bird. Savannas of the Ethiopian re- gion, from Senegal to Eritrea and Somaliland, south to Natal. * Replaces angolensis auct. 26 Zoologica: New York Zoological Society [XXIX: 4 20. N. adelberti. Buff-throated Sunbird. West Africa, from Senegal to Nigeria. Notes: The five species of the subgenus Chal- comitra are closely connected. The first three are certainly closer together than the other two, but still represent separate species. The range of N. senegalensis covers prac- tically all the suitable parts of the Ethiopian region. Its subspecies vary in their size, in the depth of their general color, going from smoky brown to black, in the tone of the scarlet of their chest, where the metallic blue middle bar of the feathers is more or less visible, and in the shade and extent of their metallic patches. These are confined to the crown, moustachial streaks, chin and upper throat in the majority of subspecies, but in others : gutturalis, cruentata, hunteri, the lesser wing-coverts are bright violet, and in the last-named, the rump and upper tail-coverts as well. Similar variations are found in N. amethystina. N. cruentata is merely the Abyssinian subspecies of senega- lensis differing from the others in its black chin, and from most of them in its violet lesser wing-coverts, similar to those of gut- turalis. N. hunteri, found in Somaliland and the drier parts of southern Abyssinia and Kenya, west -to northeastern Uganda, is also a subspecies of senegalensis, as its territory does not seem to overlap with that of any others. It differs most from cruen- tata in being blacker, in having the pos- terior part of the green crown shading into violet, the rump and upper tail-coverts violet blue, and in lacking the blue middle band to the scarlet feathers of the middle of the breast. N. adelberti, a rare species, shows a defin- ite affinity in its color pattern to N. zeylonica and so links this subgenus with the follow- ing one. Subgenus Leptocoma. This Indo-Malayan group of sunbirds is certainly related to the African Chalco- mitra. Like them, the members have a great deal of velvety feathering, brown, black, or even red. Their metallic colors, never exten- sive, are confined to the crown, throat, upper breast, lower back, wings and tail, reaching the lower breast only in the most specialized species. The mantle and sides of the head are always velvety. They differ from Chal- comitra in several particulars; they are smaller and shorter ; their bills are weaker ; 1944] Delacour: Revision of the Sunbirds 27 the nasal operculum is covered with small feathers; their tail varies from short and square to rounded or graduated and fairly long. The females are of a very different pat- tern, grayish-olive above, uniform yellow below, more or less pure, the throat being usually pale. In the majority of forms the head is gray and the throat white. Only one species (IV. chalcostetha ) shows pectoral tufts, in the male. Sunbirds of the subgenus Leptocoma range from India and Ceyion to the Philip- pines, Celebes, the Moluccas, New Guinea and the Bismarck Archipelago. 21. N. zeylonica. Amethyst-rumped Sun- bird. Ceylon and India, north to Bom- bay, Calcutta and Assam. 22. N. minima. Small Indian Sunbird. Ceylon and western India, north to Bombay. 23. N. sperata. Van Hasselt’s Sunbird. Assam, southern Siam, southern Indo- China, Malay Peninsula, Sumatra, Borneo, Java, the Philippines, Celebes and neighboring islands. 24. N. sericea* Black Sunbird. Celebes, Moluccas, New Guinea and neighbor- ing islands, Bismarck Archipelago. 25. N. chalcostetha. Macklot’s Sunbird. Southern Burma, Siam and Indo- China; Malay Peninsula, Sumatra, Borneo, Palawan, Java and neighbor- ing islands. Notes: We have noted above the relative simi- larity in general pattern of N. zeylonica and N. adelberti. I include in N. sperata all the forms usually referred to N. braziliana, and I also consider as conspecific N. grayi of Celebes. Peters {Bull. Mus. Comp. Zool. LXXXVI, No. 2, Nov. 1939, p. 121) had pointed out, quoting Chasen, that braziliana and sperata, together with henkei and juliae, constitute one species. The varia- tions of the velvety portions of the plumage from black to maroon above, deep maroon to red and yellow below, with several com- binations, are very interesting. N. sericea is near N. sperata in its gen- eral pattern, but all the velvety parts of its plumage are always pure black. The metallic patches vary in hue with the numerous subspecies. Females are of the usual type, a few having a yellow instead * Nectarinia sericea meyeri ( Hermotimia meyeri Sharpe, Handl. Gen. Sp. Birds, V, p. 43, 1909), new name for scapulata Meyer and Wiggles worth, Abh. Mus. Dresden, 1896, p. 16, Celebes (East), is preoccupied by Nectarinia jugular is meyeri ( Cinnyris jrenata meyeri Hartert, Nov. Zol. IV, p. 156, 1897; N. Celebes). I propose for it the new name: Nectarinia sericea wig - glesworthi. of a white throat, and an olive head instead of a gray one. As to N. chalcostetha, it is just a long- tailed, more brightly colored form of this group, in which the metallic purple invades the whole breast and yellow pectoral tufts appear. The female resembles closely that of N. sericea, differing only in its longer, graduated tail and in its large size. It is a lowland species, common in rubber and coconut plantations. Subgenus Nectarinia. This large group includes all the species in which metallic colors are the most ex- tensive, with no velvety feathers except in two species, johannae and superba, which have a maroon red band across the breast. The more primtive forms have the upper parts dull olive or brownish-gray, and the metallic color reduced to a purple patch on the throat and breast. It gradually invades the forehead, the crown, the back and the breast, and changes to various shades of blue, green, violet and coppery red, first dark and sombre, becoming brighter until, in the end, almost the whole plumage is glistening with iridescence. Females are always dull-colored and dif- fer slightly in pattern in the various groups. The strength, weakness, or absence of the yellow pigment varies even among geo- graphical races and depends upon their damper or drier habitat. The tail may be short and square, or long, graduated, or with narrow and length- ened central rectrices; there are all sorts of intermediate stages. Long-tailed birds are found in different groups and represent the extremes of evolution in several lines, having no very close connection with one an- other. Size is variable, as are the length and curvature of the bill. Pectoral tufts are present in males of approximately seven- tenths of the species, never in the females, with the exception of one {N. johnstoni) . Nectarinia occurs throughout the whole range of the family. Group A. Sunbirds of this group are small, square- tailed and the least brilliant of the sub- genus. The most primitive males have me- tallic purple only on the throat and chest, breast gray or yellow and upper parts dull. The most specialized are metallic above and below down to the upper breast; but the colors are always dark, never very bright: purplish violet, copper, steel blue or bronze green. The feathers of the lower breast and abdomen are never shiny, varying from white and yellow to orange, brown and black; the upper tail-coverts are not en- larged. 28 Zoologica: New York Zoological Society [XXIX: 4 Females have gray or olive upper parts, light gray or yellow under parts, either uniform or slightly mottled, but never striped or spotted. The males have pectoral tufts, which are absent in females. Species of Group A are found all over the range of the genus. 26. N. dussumieri. Seychelles Sunbird. Sey- chelles Islands. 27. N. lotenia. Loten’s Sunbird. Ceylon and Southern India, north to Ratnagiri and Madras. 28. N. jugularis. Olive-backed Sunbird. Bur- 1944] Delacour: Revision of the Sunbirds 29 ma, Andaman and Nicobar Islands, Siam, Indo-China, Hainan, southern China, Philippines and neighboring is- lands, Malay Peninsula, Sumatra, Java, Borneo, Lesser Sunda Islands, Celebes, Moluccas, New Guinea and neighboring islands to the Solomon Islands, N. Aus- tralia. 29. N. Solaris. Orange-breasted Sunbird. Lesser Sunda Islands. 30. N. asiatica. Purple Sunbird. Afghanis- tan, Baluchistan, India, Ceylon, Assam, Burma, southern Siam and Indo-China. 31. N. souimanga. Souimanga Sunbird. Ma- dagascar, Assumption and Aldabra Is- lands. 32. N. humbloti. Humblot’s Sunbird. Great Comoro and Moheli Islands. 33. N. comorensis. Anjouan Sunbird. An- jouan Island. 34. N. coquereli. Mayotte Sunbird. Mayotte Island. 36. N. venusta. Pale-bellied Sunbird. Al- most the whole Ethiopian region, from Senegal and Eritrea south to Natal, but not in lowland forests. 36. N. fusca. Dusky Sunbird. From the Kar- roo through southwestern Africa to Benguella. Notes : The Seychelles sunbird provides a good link between the African subgenus Cyano- mitra and the birds of the present group, which seem to have expanded first from Africa to the Orient, then returned to Africa via Madagascar. N. souimanga, in fact, differs from N . jugularis rhizophorae, from Hainan and Annam, only by its me- tallic green instead of grayish-olive mantle. The crown and underparts are completely similar. On the other hand, N. humbloti from Moheli and Great Comoro Islands has the dull mantle of jugularis, while the birds from the other two Comoros are much brighter and more saturated in both sexes ; they still stem from souimanga, and their nearer African relatives are the birds of the widespread species venusta. The black abdomen of comorensis is no more signifi- cant of its relationships than is that of asiatica, or of the black-bellied subspecies of jugularis. The white-bellied form of venusta in the northeast, albiventris, is linked to the yel- low and orange-bellied forms ( fazoqlensis , falkensteini, igneiventris ) through the very pale yellow blicki. The intensity of the color depends upon the habitat, turning white in dry districts (Friedmann, U. S. Nat. Mus. Bull., 153, pp. 356-360, 1937). The white-bellied forms of the south and southwest, talatala and oustaleti, on the con- trary, come into abrupt contact with the yellow-bellied falkensteini and kuanzae. The distribution of oustaleti remains con- fused on account of the mention by C. Grant and Mackworth-Praed {Bull. B.O.C., LXIII, p. 70, 1943), of an immature specimen col- lected by A. M. Chapman at Mwenzo, Northern Rhodesia, the identification of which may not be absolutely certain. So far it had only been found in Angola. It does not seem that talatala really coexists any- where with falkensteini ( nyassae ) in Nyasaland and Rhodesia. If this were the case, N. talatala would have to be considered a full species, with oustaleti as a subspecies. But it remains certain that sunbirds of this whole group are all very closely related. N. jugularis has an extraordinarily wide range, but the conspecificity of all its forms cannot be doubted, even that of the black- bellied forms (teijsmanni, clemenciae, keien- sis and buruensis) from the Moluccas, Djampea and Kalao. This was proved by the discovery of the black-bellied N. j. ulenburgi Rand in New Guinea, where otherwise only yellow-bellied forms occur. The bright orange-bellied green-crowned birds of the Lesser Sunda Islands, which coexist with yellow-bellied ones, must be considered as a separate species (solans). They are probably the reult of an earlier invasion. I prefer to consider biittikoferi, from Sumba, as a race of jugularis on ac- count of its yellow underparts. N. asiatica is evidently also an off -shoot of jugularis, com- pletely saturated with melanin. N. lotenia, with its very long curved bill, and its dark highly metallic coloration, is difficult to place. I have, however, no doubt that it is the highly specialized product of an early invasion from Africa, and that N. dussumieri is its nearest, if still distant, relative. N. fusca is a curious regressive desert species, but it appears to be near the ve- nusta group, as shown by the female plum- age pattern and the white lower abdomen of the male, as well as by its general shape. The bird described as Cinnyris picta by Hachisuka from Basilan (Proc. Biol. Soc. Washington, 54, p. 52, 1941), seems to be a hybrid between N. jugularis jugularis and N. sperata juliae (See Zimmer and Mayr, Auk, 60, p. 259, 1943). Group B. In this group, birds vary from small to fairly large for the family. The tail is square, rounded or graduated. Males are always bright metallic green on the head, neck, mantle, throat. Underparts are gray, olive, or yellow, usually with a broad scarlet band across the chest, spreading to the greater part of the belly in two cases (regia, rockefelleri) and replaced by orange or olive yellow in two others (loveridgei and vio- 30 Zoologica: Neiv York Zoological Society [XXIX: 4 lacea) . Females very plain, unmarked, gray to olive green, paler below than above. All the males have pectoral tufts, which are absent in females. These sunbirds are purely African. 37. N. chalybea. Lesser Double-collared Sunbird. South Africa north to Angola and Lake Tanganyika. 38. N. afra. Greater Double-collared Sun- bird. South and southeast Africa, to Angola, Lake Nyasa, Ruanda and Kivu Highlands, and Ruwenzori. 39. N. mediocris. East African Double-col- lared Sunbird. Highlands of Kenya and Tanganyika, south to Nyasaland. 40 ,'N. lovemdgei. Loveridge’s Sunbird. Uluguru Mountains. 41. N. preussi. Mountain Double-collared Sunbird. High mountains of West Ken- ya, Ruwenzori, Kivu, Cameroon and Fernando Po. 42. N. neergaardi. Zambesi Double-collared Sunbird. Coast of southeast Africa, from northern Zululand to northern Portuguese East Africa. 43. N. chloropygia. Olive-bellied Double-col- lared Sunbird. West Africa and Fer- nando Po, to northern Angola, Uganda and southwestern Abyssinia. 44. N. minulla. Tiny Sunbird. Cameroon and Belgian Congo. 45. N. regia. Regal Sunbird. Ruwenzori and Kivu highlands, and Mt. Kungwe. 46. N. rockefelleri. Rockfeller’s Sunbird. Mt. Kandashomwa, eastern Belgian Congo. 47. N. violacea. Wedge-tailed Sunbird. Western Cape Province to Little Nama- qualand, South Africa. Notes: All the species of this group with few ex- ceptions resemble one another in such a way that it is difficult sometimes to distinguish them. A close study of the birds and of their geographical distribution enables one, however, to understand their relationships, which often have been confused. N. chalybea is a moderate-sized bird with a comparatively narrow scarlet pectoral band and a very pale gray abdomen. The blue band on the unner breast and blue patch on the upper tail-coverts are reduced in size and of a steel blue; the latter is even absent in the race intermedia of Angola. They are southern birds of low or moderate altitudes. The recognizable races seem to be chalybea, subalaris, zonaria , mano°nsis, gertrudis and intermedia, but of some I have not been able to examine specimens. N. afra is a bigger bird, with a longer, more graduated tail, a large bill, a wide scarlet band on the chest and a slightly olive yellow darker shade on the abdomen. The blue of the chest-edging and the upper tail-coverts is more violet. Its green color is less golden. It is a bird of the lowlands or moderate elevation in south and southeast Africa (afra) and Angola (ludovicensis) , with two mountain subspecies in central eastern Africa (stuhlmanni and graueri) . The last-named has a longer and more grad- uated tail than all the others. Quite inde- pendently, C. Grant and Mackworth-Praed (Bull. B. O. C., LXIV, pp. 9, 10, 1943) have come to the same conclusions. Contrary to the opinion of several others, N. chalybea in- termedia and N. afra ludovicensis, which occur together in Angola, are two very dis- tinct birds. N. mediocris occupies a central position, being smaller than chalybea, but larger than preussi. It has golden green upper- parts, a steel blue band on the chest and upper tail-coverts, like chalybea, and strong- ly olive yellow lower breast and belly. It is a highland species of East Africa. Two sub- species, mediocris and fulleborni, occupy a large territory in E. Keya and in Tangan- yika. According to C. Grant and Mack- worth-Praed (Bull B. O. C., LXIV, p. 10, 1943) N. m. usambarica (Grote) is a sy- nonym of mediocris, and N . m. moreaui a snyonym of fulleborni. I have been unable to examine specimens of moreaui, and have seen only one of usambarica, which does not aopear to differ appreciably from a series of mediocris. I am also in accord with Grant and Mack- worth-Praed in considering the very pe- culiar N. loveridgei as a separate species. The scarlet chest-band is replaced in this bird by a small central patch, gradually merging with the intense olive yellow which extends all over the lower parts. N. love- ridgei has been considered at times a form of mediocris or of regia, but it is much too different, in both sexes, to be included in either species. (See Grant and Mackworth- Praed, ibid., pp. 10, 11). The female of love- ridgei is peculiar in having a slightly me- tallic gloss on the upperparts. N. preussi * is still smaller than medio- cris, to which it is similar, but very bright and with a dark olive brown abdomen. The band on the chest and the upper tail-coverts are violet blue and the scarlet extends far down on the breast. It is a mountain bird confined to hmh ground in East Africa (kikuyensis) , Fernando Po, and Cameroon (preussi, genderuensis) . * Cinnyris reichenowi Sharpe, Ibis, 1891, p. 444, is ante- dated by Drepanorhynchus reichenowi Fischer, Journ. f. Orn., 1884, p. 56, as both species are here included in the genus Nectarinia. The next available name for this species is Cinnyris preussi Reichenow, Journ. f. Orn., 1892, p. 190: Mt. Cameroon. The East African subspecies must now be called N. preussi kikuyensis Meams, Proc. U. S. Nat. Mus., XLVIII, p. 388, 1915: Escarpment Station, Kenya. 1944] Delacour: Revision of the Sunbirds 31 N. neergaardi, which I have not been able to examine, seems to be very similar to N. preussi, but larger with a still darker abdomen. A lowland bird of the Southeast coast, it can scarcely be a representative of the montane species. N. chloropygia is a small, widespread species resembling preussi, but with no blue band on the breast nor blue patch on the upper tail-coverts. It lives at low and moderate altitudes. Its recognizable races seem to be kempi, chloropygia, insularis, luhderi and orphogaster. The status of bine- chensis still remains doubtful. N. minulla closely resembles chloropygia, with which it coexists in Cameroon and the eastern part of the Belgian Congo. It is smaller and the scarlet chest-band is slightly mottled with metallic blue, this effect being produced by a narrow subter- minal blue band in the feathers. It has white instead of gray under wing-coverts. N. regia is a little more like mediocris. It has a long, graduated tail ; its underparts are bright scarlet in the middle, widely golden yellow at the sides and olive yellow on the lower abdomen ; under tail-coverts scarlet; chest-band and upper tail-coverts violet blue. It is a high altitude bird. From the eastern Belgian Congo another bird, N. rockefelleri, was described by Chapin in 1932. It is very near regia but differs in having the whole breast scarlet without yellow borders, and has a longer bill. It seems strange that two distinct species of such similar appearance can live on the same mountain (Mt. Kandashomwa, west of the Ruzizi Valley) ; but until more is known about it we must list rockefelleri as a full species. N. violacea is the most aberrant of the group. It is highly specialized in its long tail and orange-yellow breast, but regressive in its dull olive lower back and tail. How- ever, it belongs distinctly in the present group, as the female plumage shows. Group C. These sunbirds vary from small, square- tailed forms to large ones with long central rectrices. Males have extremely brilliant metallic colors, ranging from golden green to shinv violet and coopery red, with a black or metallic lower breast and abdomen. Females are gray or olive above, paler below, not at all or only faintly marked underneath. The males of three species, habesshvca, bouvieri. osea, are tufted. Members of this group inhabit Africa, Arabia, Palestine and southern Syria. 48. N. habessinica. Shining Sunbird. Red Sea Province. Abyssinia, Somaliland, northeastern Uganda, and southwest- ern Arabia. 49. N. bouvieri. Bouvier’s Sunbird. Savan- nas of Loango Coast, Cameroon, north- ern and southern Congo, base of Ruwenzori and Uganda. 50. N. osea. Orange-tufted Sunbird. Southern Syria, Palestine, Arabia, Chad Territory, Darfur, Lado district of the Sudan and the northern Belgian Congo. 51. N. cuprea. Coppery Sunbird. Savannas from Senegal to Abyssinia, south to the Congo, Angola and Nyasaland. 52. N. tacazze. Tacazze Sunbird. High- lands of Eritrea, Abyssinia, Kenya Colony and Kilimanjaro. 53. N. bocagei. Bocage’s Sunbird. High- lands of Angola. 54. N. purpureiventris. Purple-breasted Sunbird. Ruwenzori and Kivu Moun- tains. Notes: It is mostly by the simliar plumage of the females that these seven species show their affinities. The males vary greatly in colors, as could be expected of such bright birds. The resemblance in color of N. cuprea and N. tacazze is extremely striking, and their great difference in size, as well as the elongated central rectrices of the male tacazze, are the main distinctions between them. N. bocagei seems to be a dull, regressive offshoot of tacazze, while N. purpureiven- tris, which tops the group, is one of the most elaborately beautiful species of the whole family. Group D. This is another group of small and large species, short or long-tailed, much like the preceding one, to which it is closely re- lated. Males also have very bright metallic colors, but always of different shades of green and blue above. Abdomen black or metallic. The females of the short-tailed species differ from those of Group C in being darker above and more or less strongly mottled below. Those of the long- tailed forms gradually become paler and plainer. Males are without pectoral tufts except in three extreme species (coccinig aster, famosa, johnstoni). The female of johns- toni is the only tufted one of the whole subgenus. All species are Ethiopian. 55. N. shelleyi. Black-bellied Double-col- lared Sunbird. Lower Zambesi Valley to northern Rhodesia. 56. N. bifasciata. Purple-banded Sunbird. Southern and eastern Africa, north to the Loango Coast, Lake Edward, East 32 Zoological New York Zoological Society [XXIX: 4 Africa, Zanzibar and Pemba Islands, Abyssinia and Eritrea. 57. N. coccinig aster. Splendid Sunbird. From Senegal through West African savannas to northern Belgian Congo; also Gaboon. 58. N. erythroceria. Red-chested Sunbird. Upper Nile Valley to Uganda, the country around Lake Victoria, andi Congo Valley west to Lukolela. 59. N. pulchella. Beautiful Sunbird. Sene- gal to northern Nigeria, Abyssinia, Eritrea, Kenya Colony, southern Soma- liland and northern Tanganyika Terri- tory. 60. N. famosa. Malachite Sunbird. East and South Africa from Abyssinia to Cape Province (highlands). 61. N. johnstoni* Scarlet-tufted Sunbird. Mountains of Kenya Colony, Kiliman- jaro, Ruwenzori and northern Nyasa- land. Notes: This group includes some of the most beautiful sunbirds. N. shelleyi, a species restricted to a rather small territory, is evidently related to bifasciata, and to some extent also to habessinica and to neergaardi. It lacks, however, the yellow pectoral tufts of both those species, and the female is mottled below, showing a close affinity to N. bifasciata. N. bifasciata is a widespread and highly variable species. The small western race bifasciata reaches Lake Tanganyika, and is replaced along the coast of East Africa, up to southern Kenya Colony, by the almost identical microrhyncha, which is also pres- ent on Zanzibar Island. In South Africa and on the highlands of East Africa and Abys- sinia live four larger subspecies: mari- quensis, suahelica, osiris and hawkeri. All these forms have a definite dark red band on the breast between a violet band and the black belly. In the drier parts of coastal Kenya Colony and southern Somaliland one finds two races : tsavoensis and chalcomelas. Although closely related to those just men- tioned, they differ considerably in the purer black of the lower breast and abdomen, and in the absence of the dark red band. They have a wide bright violet band above the black lower parts. The local race of Pemba Island, pembae, still accentuates this characteristic. It is bluer above and in its coloration matches almost completely the Grand Comoro form mobii, of the Madagascan species N. notata, a very much larger bird. * Nectarinra johnstoni s^lvadorii ( N . salvadorii Shelley Bull B.O.C., XIII, p. 61, 1903. Kachere, Nyasaland) is pre- occupied by Nectnrinia sericea Salvador n ( Cinnvris s^l- o'n Jr'rii Mnrine-y. Sunbirds. p. 105, ul. 35. 1877. Jobi). I propose for it the new name: Nectarinia johnstoni nyikensis . N. coccinig aster, one of the brightest species, suggests a glorified N. bifasciata. Females of both species are very similar. The three other species of the group have elongated central rectrices, still fairly short in certain races of erythroceria and pul- chella. I consider congensis as a subspecies of erythroceria, both sexes being generally similar. They have fairly long first pri- maries and no eclipse plumage. The eastern and southern representatives of N. pulchella are melanogaster, nectarini- oides and erlangeri. They differ mainly in their dull black instead of metallic green belly, the varying extent of yellow on the sides, and the length of the central rec- trices. From the examination of good series, I have come to the conclusion that melanogaster and nectarinioides are prob- ably conspecific, the latter representing a small, saturated form of pulchella, with short bill and tail. From all available in- formation, I cannot find that the ranges of the two species really overlap. According to van Someren both have been found at Magadi, but this may have happened just at the limits of their respective habitats, where wandering birds might meet during the off-season in their search for flowering trees. It seems that nectarinioides is adapted to the semi-desert eastern country of southern Kenya Colony and northeastern Tanganyika Territory, while melanogaster prefers the less dry, open scrub and bush- velt to the west. I agree with Grant and Mackworth-Praed {Bull. B.O.C., LXIV, p. 8, 1943) that nec- tarinioides is not so closely related to erythroceria as to melanogaster, but I do not think that it ought to be considered a valid species. It is true, however, that nec- tarinioides almost links the pulchella group with erythroceria, since its female ap- proaches that of the latter species by its dark throat. Grant and Praed are mistaken when they mention “tufts” in pulchella, melanogaster and nectarinioides. None of these sunbirds possesses any real pectoral tufts of erectile fluffy feathers. They merely have normal yellow feathers on the sides of the breast, which are few in number in nectarinioides. They are as completely devoid of pectoral tufts as erythroceria. The difference in the length of the first primary in nectarini- oides and in melanogaster is proportionate to the size of the two birds; erythroceria and congensis have decidedly longer ones. N. famosa is a much larger relative of pulchella, and N. johnstoni is a handsome high mountain form, in which the female is red-tufted like the male, a strange reap- pearance of a character found in some primitive species of the genus. 1944] Delacour: Revision of the Sunbirds 33 Group E. The birds of this last group are all large and very brightly colored. The males have either short, square tails or long rectrices. The short-tailed species are iridescent green above, the two long-tailed ones being cop- pery or bronze ; the abdomen in all of them is black. The females differ from those of other groups in their yellow underparts streaked with dark olive gray. In two species ( superba and reichenowi ) these markings are faint; in the other four they are conspicuous. The female of superba is unusual in its orange under tail-coverts. In only one species, johannae, the male has yellow pectoral tufts. These birds occur in Africa, Madagascar and the Comoro Islands. 62. N. notata. Angaladian Sunbird. Mada- gascar, Comoro Islands. 63. N. johannae. Johanna’s Sunbird. West Africa from Sierra Leone to the Congo mouth, east to the Ituri district of the Belgian Congo. 64. N. superba. Superb Sunbird. West Africa from Sierra Leone to northern Angola and east to Uganda. 65. N. kilimensis. Bronzy Sunbird. Ituri highland, Uganda and Kenya Colony, south to Mashonaland and Angola. 66. N. reichenowi. Golden-winged Sunbird. Highlands of Kenya Colony and Uganda, south to Ngorongoro in Tan- ganyika Territory, west to the south- ern Kivu district. Notes: I have pointed out above the similiarity in color between the males of certain races of N. notata and N. bifasciata. On the other hand, females of notata resemble closely those of N. johannae, so that notata tends to link the two groups. N. suverba, with its gorgeous male and pale-bellied female, represents the extreme of a line of evolution, and is certainly close to johannae. These two species are the only ones of the subgenus Nectarinia to have a velvety dark red breast, suggestive of that found in the subgenus Leptocoma, particu- larly in some races of N. sperata. Males of N. kilimensis, long-tailed and coppery green in color, are very different from those of the foregoing species, but the general similarity of the females shows their relationship. The female of N. reichenowi resembles that of kilimensis in shape and color. The strongly curved bill of reichenowi is almost matched by that of N. kilimensis arturi. This evident proof of near relationship between the two species makes it impossible to admit generic distinction (Drepanorhyn- chus) for reichenowi. This wonderful species, in which both sexes have the rec- trices, primaries and secondaries bordered with golden yellow, represents the most specialized and elaborate form of the group, and perhaps of the whole genus. III. Genus Neodrepanis. The small Madagascan sunbirds of this genus are remarkable in the large fleshy blue wattle around the eye of the breeding male and in the shape of the first primary. This feather is extremely long, scarcely shorter than the second, which equals the seventh in length, the third to fifth being the longest; and it is emarginate at the tip. The tail is extremely short and soft. The bill is very long and strongly curved, with very slight serrations on the lower mandible and none on the upper one. In both species, which are closely related, there is an eclipse plumage in the males, which in breeding dress are dark metallic blue above and yellow below. Females and off-season males are olive green above, yellow mixed with olive brown below. On account of its form, size and color pattern, Neodrepanis seems clearly connected with N ectarinia, particularly the subgenus Cyan- omitra. There are no pectoral tufts. The genus is confined to Madagascar. 1. N. hypoxantha. Slender-billed Wattled Sunbird. Highlands of central eastern Madagascar — probably extinct. 2. N. corruscans. Wattled Sunbird. East- ern Madagascar at moderate altitudes. Notes : F. Salomonsen has contributed an im- portant paper on the sunbirds of this genus. ( L’Oiseau , 1934, pp. 1-9, col. pi.). IV. Genus Aethopyga. The sunbirds of this genus differ consid- erably from those of the three preceding genera in the structure of their tongue, which shows a flat terminal lobe between two semi-cylindrical ones. They possess that important characteristic in common with the following genus, Arachnothera, from which they differ, however, in other struc- tural features. All species of Aethopyga are small. They have a rather short and curved bill, longer however than the head, and variable, with a distinctly ridged cul- men. Nasal operculum bare. Wing rounded; the first primary short; third, fourth and fifth the longest. Tail rounded to much graduated; the central rectrices pointed in several species, very long in others. Males possess metallic colors, but they are never extensive, and are confined to the whole or portions of the head, neck, throat 34 Zoologica: New York Zoological Society [XXIX: 4 Variations of Colors of Breast and Lower Back in Males, Aethopyga gouldiae, Ae. nipalensis and Ae. saturata. Species and Geographical Subspecies Distribution Aethopyga gouldiae gouldiae E. Himalaya, W. Burma isolata S. W. Burma annamensis S. Indo-China harrietae, ) E. Burma, N. Siam, N. dabryi ) Indo-China, S. C. W. China Aethopyga nipalensis nipalensis, 1 E. Himalaya, Burma, W. victoriae, I Yunnan, Tonking, Malay australis J Peninsula horsfleldi W. Himalaya bland C. Laos ezrai S. C. Annam angkanensis N. W. Siam Aethopyga saturata saturata, Himalaya, N. Burma, W. Yunnan S. E. Burma, S. E. Yun- sanguinipectus, } nan, N. Siam, N. C. and W. Indo-China; Malay Peninsula wrayi ) anomala Peninsular Siam johnsi S. Annam * Some specimens show a few faint red streaks, t Many specimens have a small yellow patch, others none. and tail, and, exceptionally, to the wing- coverts. In the most primitive species, boltoni, there are only faint traces of me- tallic sheen on the crown and tail. The sides of the head are never metallic. There is always a good deal of silky red or olive yellow in the plumage. The lower back is usually bright yellow; but this patch disap- pears in certain subspecies of saturata, nipalensis, and gouldiae, and in flagrans. A very important and characteristic feature of the genus is the long, fluffy feathers of the sides of the lower back. There are no pectoral tufts. Females are dull olive above, with a grayish hood in several species. The species of Aethopyga are found from India to central and southeastern China, the Philippines, Sumatra, Java, Borneo, Sanghir Island and Celebes. 1. Ae. boltoni. Mindanao Sunbird. Min- danao. 2. Ae. flagrans. Flaming Sunbird. North- ern Philippines (Luzon, Negros, Panay, etc.). 3. Ae. pulclierrima. Sharpe’s Sunbird. The Philippines. Color of Breast Color of Lower Back Yelloiu Red Streaked Yellow Red Black Yellow Not Yelloiu I I I I I I I I I I I* I I I I I I I I It I I I I I I 4. Ae. duyvenbodei. Duyvenbode’s Sun- bird. Sanghir Island. 5. Ae. slielleyi. Lovely Sunbird. Palawan and the Philippines. 6. Ae. gouldiae. Mrs. Gould’s Sunbird. The Himalayas, Assam, Burma, Siam, central and southern China, Indo- China. 7. Ae. nipalensis. Yellow-bellied Sunbird. East Himalaya, Assam, Burma, Siam, Indo-China, southern China, Malay Peninsula (mountains). 8. Ae. eximia. Kuhl’s Sunbird. Java. 9. Ae. christinae. Sharp-tailed Sunbird. Southeastern China, eastern Indo- China, Hainan. 10. Ae. saturata. Black-breasted Sunbird. The Himalayas, Burma, Siam, Yunnan, Indo-China, Malay Peninsula. 11. Ae. siparaja. Yellow-backed Sunbird. From India to southern China, Siam, Indo-China, Philippines, Malay Penin- sula, Sumatra, Java, Borneo, Celebes. 12. Ae. mystacalis. Scarlet Sunbird. Malay Peninsula, Sumatra, Borneo, Java. 1944] Delacour: Revision of the Sunbirds 35 13. Ae. ignicauda. Fire-tailed Sunbird. Himalaya, Assam, Burma, Yunnan. Notes: The simplest forms of Aethopyga are found in the Philippines. Ae. boltoni has a graduated tail and a yellow back, but the metallic sheen on the dark gray crown is hardly noticeable. Ae. flagrans, rightly ascribed by Oustalet to Aethopyga, has since been wrongly placed in the genus Cinnyris ( Nectarinia ) by several authors. It is difficult to account for such a mistake, as it has all the im- portant characteristics of Aethopyga, par- ticularly the texture of the plumage. The alleged resemblance to Nectarinia jugularis aurora is entirely superficial. Ae. flagrans has a metallic crown, chin and tail, but lacks the yellow patch on the lower back and has a rounded tail. Ae. pulcherrima has broad metallic bor- ders to all the wing-feathers except the primaries, a feature which it shares with Ae. duyvenbodei, and a short, almost square tail. Ae. duyvenbodei, however, is closely allied to Ae. shelleyi in its general color pattern. The small Philippine forms usually re- ferred to bella are nonspecific with Ae. shelleyi from Palawan, which is merely larger and more highly colored, and has a longer tail. Ae. gouldiae is certainly the continental representative of the preceding species, larger and brighter, with a very long tail. Ae. dabryi and Ae. harrietae, from China and northern Indo-China, which have red breasts, are nothing but subspecies of gouldiae. Ae. g. annamensis, from South Annam and South Laos, lacks the yellow patch on the lower back. Both Ae. gouldiae and Ae. shelleyi have very small bills. Ae. nipa'lensis also has two Indo-Chinese races with no yellow patch on the lower back ( ezrai and blanci ), both very rare and found on high mountains. On the isolated mountain of Doi Angka, in Siam, lives the peculiar race angkanensis with an un- streaked grenadine red chest. Ae. n. ezrai and horsfieldi have it plain yellow, while all other forms have a yellow chest streaked with red to a greater or lesser degree. The amount of maroon red on the upper back varies considerably. Ae. eximia and Ae. christinae show a cer- tain agreement in colors. Ae. saturata includes Ae. sanguinipectus and its allies as subspecies. Typical saturata often lacks the yellow patch on the lower back, or has it much reduced, while anomala from peninsular Siam has none. The form johnsi, isolated on the Langbian highlands of Southern Annam, is the most distinct of all the subspecies of saturata, having even been considered often as a full species. It is small in size, with a comparatively short tail ; the black pectoral band is almost absent, and the whole chest is grenadine red faintly streaked with yellow. In the species gouldiae, nipalensis and saturata, isolation in highland areas favors variation in the color of breast and back. Yet when two or more of the species live together in the same locality they do not exhibit any parallelism in these respects. Curiously enough, when males lack the yel- low rump-patch, females of the same race appear to retain it (Ae. saturata anomala, Ae. nipalensis blanci, Ae. gouldiae anna- mensis) . Ae. siparaja is a common and widespread lowland species found from southern and western India ( vigorsi ) throughout Burma, South China, Siam, Indo-China and the Malay countries, to the Philippines (mag- nifica ) and Celebes ( fl av o striata. ) . Ae. mystacalis and Ae. temmincki are conspecific in spite of the difference in the color of their long rectrices (glossy violet and vermilion scarlet). Ae. ignicauda, re- markable for its long vermilion red tail, is the only species of the genus to have an eclipse plumage. V. Genus Arachnothera. The birds of this genus differ consider- ably from all other sunbirds, but they are nearest to Aethopyga. They are large and coarse with a massive body, a short, rounded tail; they have a long and very strong bill, short tarsi but powerful toes and sharp nails. Nostrils naked, and wings as in Aethopyga. They show no metallic colors in the plumage, which is olive, gray, and yellow — brown in one case. Three spe- cies possess pectoral tufts in the males. Their tongue is similar to that of Aetho- pyga. and they also have long, fluffy feathers on the sides of the lower back. In theh’ diet, these birds are much more insecti- vorous, less nectar-sucking, than the other sunbirds. Arachnothera is an Indo-Malayan genus. Two species extend to India, Siam, Burma and Indo-China, while others barely reach the extreme south of some of these coun- tries. It is also represented in the Philip- pines, but the Malaysian region is its stronghold. 1. A. longirostris. Little Spider-hunter. India, Burma, Yunnan, Siam, Indo- China, Malay Peninsula, Sumatra and neighboring islands, Borneo, Java, Palawan, Philippines. 2. A. crassirostris. Thick-billed Spider- hunter. Malay Peninsula, Sumatra, Borneo. 36 Zoologica: New York Zoological Society [XXIX: 4 3. A. robusta. Long-billed Spider-hunter. Malay Peninsula, Sumatra, Borneo, Java. 4. A. flavigaster. Great Yellow-eared Spider-hunter. Cochin-China, Malay Peninsula, Sumatra, Borneo. 5. A. chrysogenys. Yellow-eared Spider- hunter. Tenasserim, Peninsular Siam, Cochin-China, Malay Peninsula, Su- matra and neighboring islands, Borneo, Java. 6. A. philippinensis. Naked-faced Spider- hunter. Philippines. 7. A. af finis. Gray-breasted Spider- hunter. Tenasserim, Peninsular Siam, Cochin-China, Malay Peninsula, Su- matra, Borneo, Java, Bali. 8. A. magna. Streaked Spider-hunter. 1944] Delacour: Revision of the Sunbirds 37 Himalaya, Assam, Borneo, Yunnan, Siam, Indo-China, Malay Peninsula. 9. A. juliae. Whitehead’s Spider-Hunter. Borneo (mountains). Notes: The first three species are the nearest to the other sunbirds. The males have orange pectoral tufts, and the feathers on the sides of the lower back are very long and fluffy. The species longirostris, the smallest, with a slim but very long bill, is the most widespread; diluta from Palawan and flammifera from the Philippines are among its subspecies. A. flavigaster, chrysogenys and philip- pinensis are well differentiated species. A. af finis, magna and juliae are certainly related and very similar in shape and size, but constitute separate species, although juliae is the representative of magna on the high mountains of Borneo and belongs to the same superspecies. But it is too dif- ferent in its color pattern (brown streaked with white, and golden yellow on under tail-coverts) to be considered a subspecies of magna. A. af finis coexists with magna in the Malay Peninsula and Tenasserim. Eclipse Plumage. In a certain number of species of sun- birds, males have a double annual molt and take on a dull plumage resembling that of the females, keeping it for a period of several months after the breeding season. The presence or absence of this eclipse plumage seems to have little systematic significance, since it varies in closely allied forms, even in different subspecies of the same species. It appears that birds from the dry and colder parts, where sharply con- trasted seasons alternate, often possess an eclipse plumage. Those living near the equator in tropical forests seldom show it. But there are certainly many exceptions to the rule. Comparatively little is known on the sub- ject. The study of museum collections is of help only in the case where it is certain that males are molting from bright into dull dress. The occurrence in a series of many males in transition plumage may only mean that a number of young birds, just assuming their adult dress, have been se- cured. I myself have been almost misled by a large series of Aethopyga gouldiae har- rietae, collected in Laos during the months of November and December, 1938. I had good reason to believe that our very numerous males in transition plumage meant that this form had an eclipse dress. But the observation of captive specimens living in my tropical conservatory at Cleres, proved that it had none. Careful and prolonged observation of live birds at liberty or in captivity, under suitable con- ditions, can make certain that a seasonal eclipse plumage is assumed by a given form. We still are in doubt as to many species and subspecies. As an encouragement for further investigations, I am giving here the list of records which are definitely established, by my personal observations and by those of reliable naturalists. 1. Species of Sunbirds in which the Male has an Eclipse Plumage. Anthreptes platura N ectarinia amethystina amethystina lotenia jugularis (Indo-Malayan races) “ asiatica souimanga ? venusta kuanzae* “ fusca a fra (southern races) “ chalybea “ cuprea “ pulchella “ famosa “ reichenowtf Neodr epanis hypoxantha “ coruscans Aethopyga ignicauda II. Species of Sunbirds having no Eclipse Plumage. All Anthreptes except platura All N ectarinia of the subgenera Cyano- mitra, Chalcomitra (except N. a. amethystina ) , Leptocoma N. venusta (except kuanzae ) N. mediocris N. chloropygia N. minulla N. violacea N. tacazze N. bifasciata N. erythroceria N. notata N. coccinigaster N. johannae N. superba N. kilimensis All Aethopyga (except Ae. ignicauda) All Arachnothera The occurrence of an eclipse plumage in all other species remains unrecorded. * In a large series in the American Museum, collected by Arlfeorge in July and August, not a single male is in breeding plumage. • t The eclipse plumage of this species is partial: the male retains its yellow wings and tail, but the coppery lustre of -its -'hea<|J neck, upper back and breast is replaced by a dull black. 38 Zoologica: New York Zoological Society Alphabetical List of Generic Names. In bold face, valid genera; in Small Capitals, subgenera; in italics, synonyms. Adelinus Bonaparte, 1854. (A. verreauxi ) = N ectarinia. Aethocinnyris A. Roberts, 1922. (A. afra) — N ectarinia. Aethopyga Cabanis, 1850. (Ae. siparaja). Aidemonia Reichenbach, 1854. ( N . cuprea ) = N ectarinia. Anabathmis Reichenow, 1905. (A. reichen- bachi) = N ectarinia. Anagaladiana Reichenbach, 1854. ( N . nota- ta) — N ectarinia. Anthobaphes Cabanis, 1850. ( N . violacea) — N ectarinia. Anthodiaeta Cabanis, 1850. (A. collaris ) = = Anthreptes. Anthophagana Strand, 1928. (A. olivacea) = A ectarinia. Anthophagus Jennings, 1829. ( N . olivacea ) = A ectarinia. Anthothreptes Cabanis, 1850. (A. malacen- sis ) = Anthreptes. Anthreptes Swainson, 1837. (A. malacen- sis ). Arachnecthra Cabanis, 1850. (A. lotenia ) = N ectarinia. Arachnocestra Reichenbach, 1854. (A. crass- irostris ) = Arachnothera. Arachnophila Salvadori, 1874. (A. simplex ) = Anthreptes. Arachnoraphis Reichenbach, 1854. (A. flavi- gaster ) -== Arachnothera. Arachnothera Temminck, 1826. (A. chry- sogenys) . Baptothorax A. Roberts, 1922. ( N . sene- galensis gutturalis ) = N ectarinia. Carmelita Reichenbach, 1854. ( N . fuli- ginosa ) = N ectarinia. Chalcomitra Reichenbach, 1854. ( N . ame- thystina ) subgenus of Nectarinia. Chalcoparia Cabanis, 1850. (A. phaenicotis ) — Anthreptes. Chalcostetha Cabanis, 1850. ( N . chalcoste- tha) — Nectarinia. Chromotophora Reichenbach, 1854. (2V. su- perba) = N ectarinia. Cinnyris Cuvier, 1817. (N. superba ) = N ectarinia. Cinnyricinclus Lesson, 1840. (A. longue- marei ) — Anthreptes. Cyanomitra Reichenbach, 1854. (N. ver- ticalis cyanocephala ) subgenus of Nec- tarinia.... Cyrstostomus Cabanis, 1850. (N. jugularis ) = Nectarinia. Drepanorhynchus Fischer and Reichenow, 1354. (A. reichenowi ) = N ectarinia. Dreptes Reichenow, 1914. (N. thomensis ) = Nectarinia. Duyvena Mathews, 1925. (Ae. duyvenbodei) = Aethopyga. Eleocerthia Reichenbach, 1858. ( N . ver- reauxi) — Nectarinia. Eremicinnyris A. Roberts, 1922. (N. fusca) = Nectarinia. Euchoridia Reichenbach, 1854. (A. rectiros- tris) = Anthreptes. Eucinnyris A. Roberts, 1922. (A. venusta talatala ) = N ectarinia. Eudrepanis Sharpe, 1877. (Ae. pulcherrima ) = Aethopyga. Gunningia A. Roberts, 1922. (A. reiche- nowi) = Anthreptes. Haagneria A. Robei'ts, 1925. (A. olivacea) — N ectarinia. Hedydipna Cabanis, 1850. (A. platura) — Anthreptes. Helionympha Oberholser, 1905. (A. pul- chella nectarinioides) = A ectarinia. Hermotimia Reichenbach, 1854. (A. sericea) = A ectarinia. Hypogramma Reichenbach, 1854. (A. hypo- grammica) = N ectarinia. Lampr-othreptes A. Roberts, 1922. (A. longuemarei) — Anthreptes. Leptocoma Cabanis, 1850. (A. sperata braziliana) subgenus of A ectarinia. Leucochloridia Reichenbach, 1858. (A. ver- ticalis) = Nectarinia. Mangusia Bonaparte, 1854. (A. rectirostris) = Anthreptes. Maricornis A. Roberts, 1922. (A. bifasciata mariquensis) = Nectarinia. Microcinnyris A. Roberts, 1922. (A. chaly- bea) — A ectarinia. Nectarinia Illiger, 1811. (A. famosa). N ectarophila Reichenbach, 1854. (A. spera- ta braziliana) — N ectarinia. Neodrepanis Sharpe, 1875. (A. coruscans). A otiocinnyris A. Roberts, 1922. (A. afra) = A ectarinia. Panaeola Cabanis, 1850. (A. pulchella) = N ectarinia. Platydipna Cabanis, 1850. (A. platura) — Anthreptes. Sclaterornis A. Roberts, 1922. (A. kilimen- sis arturi) = Nectarinia. Shelleyia A. Roberts, 1922. (A. shelleyi) = N ectarinia. Tephrolaema Heine, 1860. (A. rectirostris tephrolaema) = Anthreptes. Urodrepanis Shelley, 1876. (A. christinae) — Aethopyga. Schultz: New Fishes ( Gymnotidae , Loricariidae) from Venezuela 39 5* Two New Species of Fishes (Gymnotidae, Loricariidae) from Caripito, Venezuela.1 Leonard P. Schultz Curator of Fishes, United States National Museum. (Plate I; Text-figures 1-2). [This is a contribution from the Forty-third or Venezuelan Expedition of the Department of Tropical Research of the New York Zoological Society made under the direction of Dr. Wil- liam Beebe. The expedition was sponsored by grants from the Committee for Inter-American Artistic and Intellectual Relations and from four trustees of the Zoological Society, George C. Clark, Childs Frick, Laurance S. Rockefeller and Herbert L. Satterlee, and by invaluable assistance from the Standard Oil Companies of New Jersey and Venezuela. For maps and ecological data, see Zoologica, Vol. XXVIII, No. 9, pp. 53-59, 1943.] Dr. William Beebe, Director, Department of Tropical Research, New York Zoological Society, kindly loaned to me for study his collections made during 1942 in the region of Caripito, Venezuela, and 1 take this op- portunity to express my thanks to him for the privilege of studying these specimens. Among this interesting lot of fishes I found what I consider to be two new spe- cies. The first one described belongs to the family Gymnotidae, and to the not well known genus Hypopomus Gill. The second species belongs to the family Loricariidae and was received too late to include in my recent publication entitled “The Catfishes of Venezuela, with descriptions of thirty- eight new forms,” Proceedings of the United States National Museum, vol. 94, pp. 173- 338, figures 1-5, plates 1-14, 1944. No doubt when more extensive collecting is done in the Caripito region of Venezuela additional undescribed species of fishes will be found. Family Gymnotidae. Genus Hypopomus Gill. Hypopomus Gill, Proc. Acad. Nat. Sci. Phila., p. 152, 1864. Genotype: Rhamphich- thys mulleri (Kaup). After examining the material in the na- tional collections along with 4 specimens 1 Contribution No. 684, Department of Tropical Re- search, New York Zoological Society. Published with the permission of the Secretary of the Smithsonian Institution. collected by Dr. William Beebe at Caripito, Venezuela, and comparing these with figures and descriptions of the already described species, considerable doubt must be cast on the identifications made by Ellis in his re- view of the family Gymnotidae and by Eigenmann in his British Guiana fishes, since both accounts are identical with but few exceptions. The specimens from Cari- pito have a very bluntly rounded snout and shorter head and the pore above the poster- ior nostril differs in position when com- pared with forms from other localities. Unfortunately, Kaup did not show the posi- tion of that pore in reference to the poste- rior nostril, but his measurements of, and his figure of, artedi indicate that this spe- cies has a pointed snout with the rear mar- gin of the eye behind the middle of the length from snout to occiput, while in other forms it is an equal distance as shown in figures and in the specimens before me. Steindachner’s figure of brevirostris fortu- nately shows the position of the pores in reference to the posterior nostril and these are the same as in the specimens that I am referring to occidentalis Regan from Pan- ama and the Maracaibo Basin, but the spe- cies must be different since brevirostris has 259 or 260 anal rays instead of fewer than 240 in the other species. Because of the above differences, it appears probable that the specimens from Caripito represent an undbscribed species, while those from the Maracaibo Basin are so close to those from Panama that I identify them as the same form. Measurements made on available specimens are recorded in Table I. Key to the Species of Hypopomus Gill. la. Anal rays about 259 or 260 ; the pore above the posterior nostril (see Text- figs. 1 and 2) lies behind a vertical line through the rear edge of the posterior nostril and this pore is more remote from nostril than nostril is from the edge of the eye; tail behind the anal fin 40 Zoologica: Neiv York Zoological Society [XXIX: 5 rounded, tapering to a point and length of tail is contained about 4 times in the total length; distance from posterior nostril to eye contained about 10 to 15 times in snout to occiput; rear margin of eye near middle of length from tip of snout to occiput. (Rio Guapore). Hypopomus brevirostris (Steindachner) lb. Anal ravs fewer than 240, usually from 204 to 238. 2a. Rear margin of eye at least one-half to an eye diameter behind middle of length of distance from snout to occiput; snout contained 2% to 3 times in the head and about 1 2/3 in postorbital length of head; distance from posterior nostril to eye contained about 15 times in length from snout to occiput; pore above the posterior nostril lies behind a vertical line through rear edge of nostril. (Rio Mona, French Guiana) Hypopomus artedi (Kaup)2 2b. Rear margin of eye midway between tip of snout and occiput; snout contained more than 2 times in postorbital length of head. 3a. Distance from posterior nostril to eye contained from 9 to 14 times in length from snout to occiput; the pore above the posterior nostril lies behind a ver- tical line through rear of nostril or this line bisects the pore (Text-fig. 2), the latter is separated from the nasal open- ing by an isthmus of skin; tail behind ..anal compressed and ending rather abruptly, not gradually tapering to a rounded point and contained about 5 to 6 times in the total length; snout 2 1/5 to 2 1/3 into postorbital length of head. Rio Condoto; Panama; Maracaibo Ba- sin) Hypopomus occidentalis Regan. 3b. Distance from posterior nostril to eye contained about 25 to 32 times in dis- tance from snout to occiput; the pore above posterior nostril lies close to mar- gin of that nostril and is bisected by a - line through middle of posterior nostril or the pore is just in front of this line (Text-fig. 1 ) ; snout very bluntly rounded, 2% to 2 4/5 times in postorbital length of head ; tail very little compressed, tapering to a point and contained about 5 1/3 to 6 times in total length. (Cari- pito, Venezuela). Hypopomus beebei new species. - Rhamphichthys mulleri Kaup is referred to this species as a synonym. I have examined a specimen of artedi (I. U. No. 12620) kindly loaned by Dr. J. L. Kask, Cali- fornia Academy of Sciences, and refer it to this species. Hypopomus beebei, new species. Text-fig. 1; Plate 1, Fig. 4. Holotype: U.S.N.M. No. 120753, a speci- men, 136 mm. in total length and 112 mm. from snout tip to end of anal fin, collected by Dr. William Beebe at Caripito, Venezu- ela, during 1942. Paratypes: Three specimens, 129 and 150 mm., the other specimen with tail broken off, taken along with the holotype and bear- ing same data (Cat. No. 30,040 in the col- lections of the Department of Tropica' Research, New York Zoological Society). Description: The holotype and 2 para- types were measured, and these data, along with that recorded for other species, are presented in Table I. Text-fig. 1. Arrangement of pores on head of Hypopomus beebei. Text-fig. 2. Arrangement of pores on head of Hypopomus occidentalis. Body compressed, tail slightly com- pressed and tapering to a point; head blunt- ly rounded; snout short, about equal to interorbital space, contained about 4 1/5 times in head; jaws without teeth; length of pectoral fin 2 in head; lateral line straight, the 3 rows of scales below and about 4 rows above enlarged; scales along back and ventrally on body much smaller in size; head a trifle longer than greatest depth ; origin of anal fin about opposite tips of pectorals; anal papilla present, its base under middie of opercle or a vertical line 1944] Schultz : New Fishes ( Gymnotidae , Loricariidae) from Venezuela 41 Table I. Counts and Measurements Made on Species of Hypopomus, Expressed in Hundredths of the Length from Snout Tip to End of Anal Fin. beebei occidentalis artedi brevirostris after Stein- Holo- Para- Para- Maracaibo after dachner’s Characters type type type Basin Kaup figure Length to end of anal fin in millimeters. . . 112. 106. 124. 100. 137. 244. 252. Length of anal fin base . 83.0 84.0 83.9 85.0 81.7 86.5 Length of head . 12.1 12.3 11.7 13.1 12.4 13.5 11.7 Length of snout . 3.21 3.11 3.06 4.0 3.87 5.12 2.78 Greatest depth . 11.06 11.8 9.68 12.0 13.1 9.30 9.33 Width of interorbital space . 3.03 3.30 3.14 3.00 2.72 Postorbital length of head . 8.48 8.20 8.14 8.30 8.03 7.78 7.34 Snout to occiput . 8.57 8.49 8.06 10.0 9.20 9.02 7.74 Diameter of eye . 1.25 1.42 1.29 1.50 1.24 0.98 1.59 Distance from anterior to posterior nostril . 2.23 2.08 2.26 2.70 2.12 1.90 Distance from eye to posterior nostril. . . . . 0.28 0.27 0.32 0.80 0.88 0.56 Width of gill opening- . 2.41 2.73 2.58 2.70 3.22 2.85 Snout to anus . 8.48 9.34 8.14 9.50 8.61 7.10 8.92 Snout to anal origin . 17.4 16.6 16.1 16.2 16.8 20.7 15.5 Anus to anal origin . 9.64 7.83 8.39 7.50 8.39 7.54 Snout to pectoral insertion . 11.2 11.3 11.3 12.5 12.1 11.7 Longest ray of pectoral fin . 5.35 5.47 6.20 5.84 5.29 5.36 Longest ray of anal fin . 4.02 5.00 4.60 3.77 Length of tail beyond anal fin . 21.0 22.3 20.1 20.5 25.1 20.9 32.9 Width of head at eyes . 4.46 5.19 4.92 4.30 3.87 Number of anal rays .214 228 217 204 223 220 or 259 or 223 260 through occiput passes through base of anal papilla; lower jaw very slightly shorter than upper; mouth terminal, small; cephalic canals and pores prominent; ihucus pores numerous on head ; gill opening extending a little above and below pectoral fin base and more or less enclosing it, except pos- teriorly ; margin of eye not free ; eye small, a little over two times in the interorbital space; interorbital space convex, about 3 times in distance from snout tip to occiput; fontanel present from between eyes to oc- ciput. Color: Body light brownish in alcohol with 17 narrow dark brown bars across sides to end of anal fin; sometimes an in- complete or broken bar occurs between most or all the nearly complete bars; pectoral fins and anal fin with numerous dark brown pigment specks; tail beyond anal fin with about 3 more brown bars more or less ob- scure or absent. Named beebei in honor of Dr. William Beebe, collector of this new species, who so kindly loaned to me his fish specimens from Caripito, Venezuela. Family Loricariidae. Corymbophanes venezuelae, new species. Text-fig. 2; Plate 1, Figs. 1-3. Holotype: U.S.N.M., No. 120752, a speci- men 72.5 mm. in standard length, 94.5 mm. total length, collected by Dr. William Beebe in the Rio Caripe, Caripito, Venezuela, dur- ing 1942. Paratype: A specimen 32.5 mm. in stand- ard length, 45 mm. total length, collected with the holotype and bearing same data (Cat. No. 30064, in the collections of the Department of Tropical Research, New York Zoological Society). Description: Detailed measurements were made on the holotype and paratype and these data, expressed in hundredths of the standard length, are recorded in Table II, along with comparative data taken from the photographs of Corymbophanes andersoni Eigenmann. The following counts were made, respec- tively, for holotype and paratype: Dorsal rays I, 10; I, 9; anal rays ii, 4; ii, 5; pelvic rays I, 5-1, 5; pectoral rays I, 6-1, 6; I, 6-1, 6 ; branched caudal fin rays 14 ; 14 ; series of scutes along lower sides 24; 24; pores in lateral line 25; 25; plates in front of dorsal fin 4-3; 3-3; scutes between anal and caudal fins 11; 11; spinuels on preopercle 7 to 11; 2 hooked spines on interopercle. Head depressed, body depressed forward, caudal peduncle a little compressed, triangu- lar in cross section; rami of jaws long, the 42 Zoologica: New York Zoological Society [XXIX: 5 Table II. Measurements, Expressed in Hundredths of the Standard Length, for Two Species of Corymbophanes Eigenmann. andersoni venezuelae Characters From figures of type Holotype Paratype Standard lengths in millimeters 86. 72.5 32.5 Width across base of pectorals 30.0 37.2 32.9 Greatest depth 20.4 18.6 18.2 Snout 19.6 24.3 21.8 Interorbital space 11.7 10.3 12.3 Diameter of the eye 3.83 4.96 6.46 Length of ramus of lower jaw 8.33 11.0 10.8 Distance from nostrils to snout tip 13.8 18.2 14.8 Distance from nostrils to eve 4.58 4.14 3.39 Greatest width of lower lip 7.92 7.86 7.08 Tip of snout to gill opening 27.6 25.6 Tip of snout to occiput ?25.8 35.2 34.2 Distance from eye to rear of temporal plate. . 8.08 9.66 9.85 Length of caudal peduncle 28.8 24.1 27.1 Least depth of caudal peduncle 11.2 13.8 14.8 Length of first dorsal ray 19.7 20.7 26.2 Length of last dorsal ray 10.4 14.1 14.2 Length of pectoral spine 24.6 26.6 26.7 Length of upper caudal ray 26.7 21.4 28.6 Length of lower caudal ray 29.2 2 7.9 31.7 Length of longest anal ray 7.92 9.10 6.16 Snout to dorsal origin 45.0 46.9 44.6 Snout to anal origin 64.7 71.7 71.0 Anus to anal origin 9.38 10.2 Length of base of dorsal fin 19.6 26.2 24.9 ramus of the lower jaw contained 0.9 and 1.0 times in the interorbital space; head (to end of temporal plate) 2.8 and 2.7, depth 5.5, both in standard length; eye 3 and 4V2 in snout; eye 4% and 6 in head to end of opercle and 5 and 7% to end of temporal plate; eye 1.5 and 1.7 in interorbital space; lips of oral disk papillate, the papillae larger near margin of upper lip; length of free portion of maxillary barbel about 2/3 eye diameter; ramus of upper jaw not quite as long as ramus of lower jaw; both jaws with numerous very fine teeth with bifid tips, the inner lobe longest; ventral surface naked from anal fin region forward; anterior por- tion of head from just in front of nostrils and eyes naked, but with small, firm, em- bedded nodules ; this naked area extends along sides of head to include interopercle and base of opercle; interorbital space slightly convex and free of prickles except a few over orbit, top of head to occiput also free of prickles ; predorsal scutes in 3 or 4 pairs; dorsal surface of head evenly convex; no ridges or grooves on the head; none of the plates is keeled on the head or sides of body; each side of dorsal fin base naked; adipose fin completely lacking and no ridge where this fin might be expected; origin of dorsal fin an eye diameter in front of a vertical line through pelvic insertion; in- sertion of pelvics a trifle closer to tip of snout than midcaudal fin base; first two or three branched rays of dorsal fin longest; caudal fin a trifle concave; lower lobe long- est; margin of dorsal fin a little rounded; third branched ray of pelvics longest; a vertical line through origin of anal fin is a little closer to base of last dorsal ray than to its tip; pectoral fins reach to or just past pelvic base and pelvic fins just past anal fin base; the opercle and interopercle are separately movable, the latter with two hooked short spines, their bases covered by thick skin; there are from 7 to 11 small spinules on the opercle. Color: Caudal and dorsal fins barred; gen- eral coloration in alcohol brownish. No black spots along base of dorsal fin or on upper or lower surfaces of the body. Remarks: This new species, Corymbo- phanes venezuelae, is a Chaetostoma lacking an adipose fin. If certain other characters, when compared with various species re- ferred to Chaetostoma, were not different, too, I would be inclined to refer it to that genus since certain species, as Chaetostoma anomala, occasionally lack an adipose fin. The chief difference between these genera, besides the adipose fin, is in the degree of nakedness of the snout and the backward extension of this naked area to include the interopercle and most of the opercle. In Chaetostoma usually not quite half the snout is naked, but in Corymbophanes the snout is naked all the way to the eyes and to the anterior margins of the nostrils. Another difference is the lack of prickles in Corym- bophanes on the supraoccipital area forward to and including the interorbital space. 1944] Schultz: New Fishes ( Gymnotidae , Loricariidae ) from Venezuela 43 The two species now referred to this genus may be distinguished from each other by means of the following key: la. Dorsal rays I, 9 or I, 10; eye 1 ^ to 1 % times in the interorbital space; length of base of dorsal fin much longer than distance to dorsal origin. (Rio Caripe). Corymboplianes venezuelae, new species, lb. Dorsal rays I, 7 ; eye in interorbital space 3; length of base of dorsal fin much shorter than eye to dorsal origin. (British Guiana) Corymboplianes andersoni Eigenmann 44 Zoologica: New York Zoological Society EXPLANATION OF THE PLATE. Plate I. Figs. 1, 2, 3. Corymbophanes venezuelae, new species. Holotype, U.S.N.M. No. 120,- 752. Standard length 72.5 mm. Rio Caripe, near Caripito, Venezuela. Fig. 4. Hypopomus beebei, new species. Holo- type, U.S.N.M. No. 120753. Total length 136 mm. Caripito, Venezuela. SCHULTZ. PLATE I TWO NEW SPECIES OF FISHES (GYMNOTIDAE. LORICARIIDAE) FROM CARIPITO, VENEZUELA. Nigrelli & Gordon : Tumor in the Silverside 45 6® A Melanotic Tumor in the Silverside, Menidia beryllina peninsulae (Good and Bean). Ross F. Nigrelli & Myron Gordon New York Aquarium. (Plate I; Text-fig. 1). Introduction. Melanotic tumors in fishes living in a feral state have been reported by several investi- gators. Osburn (1925) described a tumor of this type from several common bullheads ( Ameiurus nebulosus ) taken from a pond near Falmouth, Massachusetts. This inves- tigator showed that the melanotic tumor was induced by a black pigment-producing coccus. The bacteria was isolated, cultured and the disease reproduced in the skin of apparently normal catfish by Injection of the cultivated microorganisms. Other para- sites are capable of causing a proliferation of melanophores. Thus, Hsiao (1941) re- ported a condition of melanosis in a cod (Gadus collaris ) which was induced by the metacercariae of a heterophyid fluke. Hun- ter (1941) made further studies on melano- phores associated with cysts of similar flukes ( Cryptocotyle lingua) in the skin of the dinner ( Tautogolabrus adspersus) . That melanophore proliferation may be correlated with wound healing is indicated by the experiments of Smith (1931, 1932a, 1932b) on goldfish ( Carassius auratus). He pointed out that mechanical injuries to the skin and exposure to X-rays may cause a response of melanophores which in some cases may be so great as to cause a general but temporary melanosis. There are other records of melanotic tu- mors of fishes in the literature, but the etiology of the majority of them is not known. Thus, Ingleby (1929) reported a melanotic neoplasm in an angler ( Lophius piscatorius) . It consisted of chromatophores which were of the same type as those found under the normal epithelium. The pigment cells showed a tendency to spread laterally and to deeper areas of the skin. From this description, it is apparently similar to the melanoma of platyfish-swordtail hybrids, and as will be seen shortly, to that of the silver- side as well. Similar cutaneous tumors were reported by Haddow and Blake (1933) in the thornback ray. In 1931, Mr. Stewart Springer caught in Biloxi Bay at Iberville, Mississippi, a repre- sentative collection of fishes containing many individuals and species. The collec- tion was sent to Dr. Carl L. Hubbs, Cura- tor of Fishes in the Museum of Zoology at the University of Michigan. In sorting and cataloguing the specimens, Dr. Hubbs dis- covered a silverside with an apparent melan- otic tumor. He kindly sent it to us for detailed study. We are indebted to him for the op- portunity of making a comparative study of this tumor, which represents a spontane- ous growth in a feral fish, with the melano- mas of the aquarium-reared platyfish- swordtail hybrids. As Gordon (1941) has pointed out, the melanotic tumors of the platyfish-swordtail hybrids are produced experimentally by genetic methods, specifi- cally by mating a platyfish carrying macro- melanophores with a swordtail. Under na- tural condition in their native habitat, in the rivers of Mexico and Guatemala, Platy- poecilus maculatus may occasionally be found living side by side with Xiphophorus hellerii — yet in more than 10,000 specimens examined, no hybrids and no tumors were found. Regardless of the mode of origin, it will be shown that the silverside and platyfish-swordtail hybrid melanotic tumors have many characteristics in common. Description of the Tumor. The silverside measured 47 mm. in total length after its preservation in alcohol. The tumor growths covered an area of 5 X 1.5 X 1.0 mm. (Text-fig. 1) and they extended along the sides of the body in the dorsal region of the caudal peduncle and upper lobe of the tail fin. An examination of all ex- ternal and internal parts of the body re- vealed no free or encysted protozoan or metazoan parasites, nor were they discov- ered after the tumor was sectioned. Except 46 Zoologica: New York Zoological Society [XXIX: 6 for the melanotic tumor of the tail region, the fish appeared normal. In examination of the pigmentary system of the silverside, it was noted that in the region of the lateral line a broad band of melanophores appeared to form a syncy- tium. Immediately above this line, large melanophores formed a striking reticular pattern merging into an even distribution of melanophores near and at the mid-dorsal line. The melanotic neoplasm was confined chiefly to the lateral line area, extending only slightly below this region. Whether or not the tumor cells were derived from the lateral line melanophores could not be de- termined. Text-fig. 1. Lateral and dorsal view of the tide-water silverside showing position and extent of melanoma. Slightly less than nat- ural size. (Drawing by C. Clark.) The tail region of the fish bearing the tumor was severed from the body and pre- pared for microscopical examination. After embedding in paraffin, sections were cut at 6 microns. They were stained by a va- riety of methods, both for bacteriological and general histological detail. A modified method of Ziehl-Neelson for the detection of acid-fast organisms and the Brown and Brenn technique for the Gram’s reaction, as outlined by Gradwohl (1938), were used. Our studies showed that the tissues appeared to be free of bacteria except for an isolated Gram negative bacil- lus here and there which was not regarded as significant. Some sections were stained with Ehr- lich’s hematoxvlin. others with Mallory’s triole stain. These proved to be of greatest value although Giemsa’s stain was em- ployed as well. Histologically, the pigment cell hvperplasia in Menidia appeared quite like the melanotic tumors reported for the platyfish-swordtail hybrids by Reed and Gordon (1931) and indicated by them as being in the second state of melanosis. It also resembled the tumor described by Gor- don and Smith (1938) in other platyfish species hybrids. The silverside tumor also has many of the features mentioned by Ingleby (1929) for the melanoma in Lovhius. The tumor of the silverside contained sev- eral sizes of melanin-bearing cells, corre- sponding somewhat to the size variation found among the melanophores in the cor- ium in normal regions of the fish’s body. The pigmented cells in the tumor mass dif- fered from those in the normal corium in that they were not of the precise stellate form of the corial melanophores; rather they appeared amoeboid in shape. The corium of the tumor masses was al- most completely replaced by proliferating pigment-bearing cells and was considerably thickened (PI. I, Fig. 2; B) as a result of their growth. The boundaries of the me- lanin-containing cells were not seen clearly owing to the presence of dense deposits of pigment granules. From our study, these ap- parently active cells seemed to spread lat- erally in the corial layer and thus might have given rise to other loci with the result that a number of pigmented nodules ap- peared (PI. I, Fig. 1; 1, 2, 3). In response to the hypertrophied corial tissues, the epi- dermis in several places was penetrated by the tumor cells and destroyed (PI. I, Fig. 2; A). This left the tumor naked at the surface. There was no indication of hemor- rhages as described by Reed and Gordon in the platyfish-swordtail melanoma. Sur- rounding the region of the break-through, the epithelium of the silverside tumor was somewhat thickened and keratinized (Plate I, Fig. 2; C). The epidermal tissue con- tained many cells full of large clusters of melanin granules, which had some resem- blance to engorged macrophages as illus- trated and described by Grand, Gordon and Cameron (1941) in their study of fish me- lanomas in tissue cultures. The presence of large, pigment-filled cells in the epidermis appeared similar to the condition described by Smith (1932a) in the goldfish fol'owing injury and healing, and interpreted by him as being part of the process of melanin elimination by macrophages. Another in- stance of this phenomenon was reported by Gordon and Lansing (1943) in platyfish hybrids. The silverside melanoma cells arising from the corial nodule penetrated (PI. I, Fig. 2; D), in several noints, the underly- ing fascia and invaded the subcutaneous areas attacking musc’e and bone tissues. In the invasion process the muscle fibers were SDlit and eventually broke down, as indi- cated bv the loss of striation and hyaliniza- tion (PI. I, Fig. 2; E). The invasion route of the tumor cells appeared to follow the path established by the intermuscular tis- sues. In localized areas, where bone was reached by the tumor cells, the periosteum was also attacked, destroyed and replaced PI. I, Fig. 2; F). In our study of sections stained with Mal- lory’s, sinusoids and capillaries were evi- dent; they were abundant in regions of par- 1944] Nigrelli & Gordon : Tumor in the Silverside 47 ticularly large tumor masses. No inflamma- tory cells were recognizable and a similar condition was reported by Gordon and Smith (1938a) for platyfish-swordtail me- lanomas. In the melanotic tumor of Lophius, however, Ingleby reported the presence of inflammatory cells in the tissues underlying the tumor, but separated from the tumor proper by a broad layer of connective tissue. Summary and Conclusions. A melanoma in the caudal peduncle and tail region of a feral silverside ( Menidia ) is reported, described and compared with melanotic neoplasms of Lophius and platy- fish-swordtail hybrids. In several important respects they are quite similar. The lesions are characterized by an over- growth of melanin-containing cells which are capable of infiltrating, destroying and replacing epithelium, corium, fascia, mus- cle and periosteum. The etiology of the tumor is not known but parasitological and bacteriological techniques employed indicate that a para- sitic causative agent is probably not in- volved in this melanoma. References. Gordon, Myron 1941. Genetics of melanomas in fishes. V. The reappearance of ancestral macro- melanophores in offspring of parents lacking these cells. Cancer Research, 1: 656-659. Gordon, Myron, and Winifred Lansing 1943. Cutaneous melanophore eruptions in young fishes during stages preceding melanotic tumor formation. Jour. Morphology, 73: 231-245. Gordon, Myron, and G. M. Smith 1938a. Progressive growth stages of a her- itable melanotic neoplastic disease in fishes from the day of birth. Amer. Jour, of Cancer, 34: 255-272. 1938b. The production of melanotic neoplas- tic disease in fishes by selective matings. IV. Genetics of geograph- ical species hybrids. Amer. Jour, of Cancer, 34: 543-565. Gradwohl, R. B. H. 1938. Clinical Laboratory Methods and Diagnosis. C. V. Mosby Co. Pub. 1-1607. Grand, C. G., Myron Gordon and Gladys Cameron 1941. Neoplasm studies. VIII. Cell types in tissue culture of fish melanotic tu- mors compared with mammalian me- lanomas. Cancer Research, 1: 660- 666. Haddow, A., and Isobel Blake 1933. Neoplasm in fish: A report of six cases with a summary of the litera- ture. J. Path. Bad., 36; 41-47. Hsiao, Sidney C. T. 1941. Melanosis in the common cod, Gadus collaris L., associated with trematode infection. Biol. Bull., 80: 37-44. PI UN TER, III, G. W. 1941. Studies on host-parasite reactions. VI. An hypothesis to account for pig- mented metacercarial cysts in fish. J. Para., 2t: 33 (suppl.). Ingleby, Helen 1929. Melanotic tumor in Lophius pisca- torius. Arch. Path., 8: 1016-1017. Osburn, Raymond C. 1925. Black tumor of the catfish. Bull. Bu- reau of Fisheries, 41 : 9-13. Reed, H. D., and Myron Gordon 1931. The morphology of melanotic over- growths in hybrids of Mexican killi- fishes. Amer. Jour, of Cancer, 15: 1524-1549. Smith, George M. 1931. The occurrence of melanophores in certain experimental wounds of the goldfish ( Carassius auratus). Biol. Bull., 61: 73-84. 1932a. Eruptions of corial melanophores and general cutaneous melanosis in the goldfish (Carassius auratus) fol- lowing exposure to X-ray. Amer. Jour, of Cancer, 16: 863-870. 1932b. Melanophores induced by X-ray com- pared with those existing in patterns as seen in Carassius auratus. Biol. Bull., 63: 484-491. 48 Zoologica: New York Zoological Society Fig. 1 Fig. 2 EXPLANATION OF THE PLATE. (Photomicrography by S. C. Dunton). Plate I. Sagittal section through the caudal peduncle and tail of the silverside showing melanotic nodules (1, 2, 3,). H-E. About 25 X. Higher magnification of a sagittal section through one of the melanotic nodules showing the extent of the le- sion. A, point where melanin-bearing tumor cells had broken through the epidermis; B, thickened corium; C, slightly thickened epidermis with massed macrophages containing mel- anin granules; D, the fascia separat- ing corium from muscle is broken and the pigmented cells invade the deeper layers of tissues; E, muscle fibers are split and the muscle tissue loses its striation and becomes hya- linized; F, bone showing the melano- phores massed around the periosteal area; melanophores may also be seen in the region around the dorsal aorta; G, Scale. H.-E. 250 X. NIGRELLI & GORDON. PLATE 1. FIG. 1. FIG. 2. A MELANOTIC TUMOR IN THE SILVERSIDE, MENIDIA BERYLLINA PENINSULAE (GOOD AND BEAN). k NEW YORK ZOOLOGICAL SOCIETY General Office: 630 Fifth Avenue, New York 20, N. Y. OFFICERS Fairfield Osborn, President Alfred Ely, First Vice-president Laurance S. Rockefeller, Second Vice-president Harold J. O’Connell, Secretary Cornelius R. Agnew, Treasurer William Bridges, Editor and Curator of Publications Zoological Park Lee S. Crandall, General Curator & Curator of Birds Leonard J. Goss, Veterinarian Claude W. Leister, Associate, Mammals John Tee-Van, Associate, Reptiles Grace Davall, Assistant to General Curator Aquarium Christopher W. Coates, Curator and Aquarist Ross F. Nigrelli, Pathologist Myron Gordon, Assistant Curator C. M. Breder, Jr., Research Associate in Ichthyology G. M. Smith, Research Associate in Pathology Homer W. Smith, Research Associate in Physiology SCIENTIFIC STAFF General John Tee-Van, Executive Secretary Jean Delacour, Technical Adviser Claude W. Leister, Education Department of Tropical Research William Beebe, Director Jocelyn Crane, Research Zoologist Henry Fleming, Entomologist William K. Gregory, Associate Gloria Hollister, Associate John Tee-Van, Associate Mary VanderPyl, Associate Editorial Committee Fairfield Osborn, Chairman William Beebe William Bridges Lee S. Crandall Jean Delacour Claude W. Leister John Tee-Van ZOOLOGICA SCIENTIFIC CONTRIBUTIONS of the NEW YORK ZOOLOGICAL SOCIETY VOLUME XXIX Part 2 Numbers 7-9 Published by the Society The Zoological Park, New York August 22, 1944 CONTENTS Page 7. The Effects of Steroids on the Skeleton of the Poecillid Fish Lebistes reticulatus. By Joseph L. Scott. (Plate I) 49 8. The Function of Secondary Sexual Characters in Two Species of Dynastidae (Coleoptera) . By William Beebe. (Plates I-V).. 53 9. Eastern Pacific Expeditions of the New York Zoological Society. XXXIII. Pacific Myctophidae. (Fishes). By William Beebe & Mary Vander Pyl. (Text-figures 1-25) 59 Scott: Effects of Steroids on Lebistes reticulatus 49 7. The Effects of Steroids on the Skeleton of the Poecillid Fish Lebistes reticulatus. Joseph L. Scott Department of Biology, Washington Square College of Arts and Science, New York University. (Plate I). The effect of the sex hormones and ex- tracts of the anterior lobe of the pituitary upon the structure of bone has been studied by various investigators. Estrogen-treated mice, guinea pigs, pigeons and chickens have shown, upon histological study of the skeleton, an increased ossification, at least in certain regions. In order to help com- plete the vertebrate phylogenetic picture, the work reported here was designed to ex- tend this study to fish. The effects of preg- neninolone (Pranone, ethinyl testosterone, anhydrodroxy-progesterone) and alpha es- tradiol (Progynon DH) upon the skeletal structures of Lebistes reticulatus (Peters), a fish in which the secondary sex characters are highly responsive to these substances, were determined and are reported here. In this species, the effects of pregneninolone are purely androgenic, but in mammals this steroid substance has various other actions. The author wishes to express his appre- ciation to Drs. Robert Gaunt and C. M. Breder, Jr., for suggestions and assistance. He is also indebted to Dr. Max Gilbert of the Schering Corporation for supplying the hormones used in this work. The data pre- sented here are taken in part from a thesis for the degree of Master of Science at New York University. Ten to fifteen fish were kept in two and a half gallon aquaria. Several plants were grown in each aquarium in an inch of sand. The temperature of the water was con- trolled thermostatically and ranged from 740-78° F. Each tank was fitted with a por- ous stone through which was pumped a steady stream of air. Both treated and con- trol fish were fed, daily, a diet of dried tropical fish food which was supplemented by Daphnia and Tubifex once a week. Lebistes reticulatus, a viviparous poecillid with marked sexually dimorphic character- istics, was used as the experimental animal. In this species the anal fin of the adult male is modified to form a gonopodium, the body size of the male is normally smaller than that of the female, and the male only is bril- liantly colored. The reaction of these sexu- ally dimorphic traits of Lebistes to the two steroids used here have been previously de- scribed by Berkowitz (1) and Eversole (6). Since their results were fully confirmed in the course of our observations no dupli- cate description is given of them. A total of 67 fish were treated. Thirty received estradiol, contained in tablets made for human oral use which were powdered and sprinkled on the water. Total dosage varied from 6.5 to 15 milligrams per tank given in- equal weekly portions for periods varying from 18 to 110 days. A similar pro- cedure was followed in treating fish with 5-milligram doses of pregneninolone, except that the effects of a single treatment were so long-lasting that monthly administration was more than sufficient to maintain mas- culinization of all secondary sex characters. In fact, to eliminate effective amounts of the steroid from the tanks a very thorough cleaning had to be done even six months after a dose of the compound was added to the tanks. Histological studies were made on a seg- ment cut transversely just caudal to the pec- toral fin and between the posterior end of the body cavity and the caudal fin. The whole skeleton was studied in total mounts. The material was fixed in Bonin’s fluid, washed in 70% alcohol, dehydrated by the dioxan method, and embedded in tissue-mat. Serial sections were cut at 10 microns in thickness and were stained either with Har- ris’ hematoxylin and counterstained with eosin, or with Mallory’s triple stain. Whole mounts were prepared by clearing freshly killed fish in KOH until the muscular and skeletal portions were translucent and then staining with Alizarin dye as described by Hollister (10). Since the treatments given either mascu- linized or feminized the external sexual characteristics of the animals, the original sex was determined at the end of the ex- periment by histological examination of the gonads. These were routinely sectioned in the course of other observations. The gonad is itself affected by these treatments (1, 6), "k 25 >44 50 Zoologica: New York Zoological Society [XXIX: 7 but whether it was originally a testis or ovary was generally easily discernible by its histological appearance, the essential game- togenic elements of which were not changed. The sex ratio of the strain of fish used here is 1:1. Observations. I. Sexual dimorphism in the skeleton. Cleared whole mounts of normal male and female Lebistes showed a skeletal dimorphic character, aside from the difference in size, only in the region of the anal fin or gono- podium (.male). The skeleton which sup- ports the anal fin of the adult female (Fig. 2) and of immature fish is composed of a group of ten separate interhaemal bones which project dorsaily from the anal fin to- ward the haemal spines of the first three caudal vertebrae. Three of the interhaemal bones, the sec- ond, third and fourth, are fused as one bone in the adult male (Fig. 1). In addition the first three caudal vertebrae are markedly enlarged and make a sharp angle in the di- rection of the three fused interhaemal bones. Most bones in this species are made up of a central cartilaginous mass surrounded by a crust-like ossified layer. No sexual di- morphism in the histology of these struc- tures was noted, except again in the anal fin region where the interhaemal bones of the male contained much thicker ossified layers. II. The effects of hormones upon the skeleton. All fish, regardless of sex, fed 5 milli- grams of pregneninolone at birth, developed typical male structures precociously in the anal fin region described above. There was a fusion of four and sometimes five of the interhaemal bones in the treated animals (Figs. 3, 4, 7). This steroid caused a dwarfed condition in Lebistes and conse- quently the whole skeletal apparatus of the treated fish was reduced in size (Fig. 3). Adult females, given a single feeding of 5 milligrams of pregneninolone, all developed interhaemal bones much like the adult males (Fig. 4) within three weeks. Normal males show this skeletal dimorphism 50-60 days after birth. Fifteen fish were fed alpha estradiol from birth. All developed a typical female anal fin with the 10 separate interhaemal bones attached (Figs. 2, 6). One group of 9 fish received 3.25 milligrams of the hormone over a period of 50 days. The male controls had begun to show a difference in the anal fin skeleton but not the treated animals. Continuing treatment until 110 days in an- other series did not vary results. All of the estrogen-treated fish were stud- ied for the effect upon ossification of the interhaemal bones. It was found in every case that there was no increase in ossifica- tion and these bones had the exact appear- ance of the female untreated animals (Figs. 5, 6). On the other hand, all of the fish that were fed pregneninolone, without exception, showed an increase deposition of bone around the central cartilaginous structure of the interhaemal bones toward the condi- tion of the adult male (Fig. 7). The first three haemal spines showed an increase in size which approached that of a normal adult male. The supporting structures of the gonado- podium itself undergo profound changes in the course of sexual maturation in the male. These are induced precociously by pi’egne- ninolone, as described by Eversole (6). The vertebral column was studied care- fully for possible effects of hormone treat- ment. In all of the 67 fish treated with either hormone there was no increase in ossification or other histological changes in the vertebrae, except for the larger spines associated with the anal fin apparatus of the pregneninolone-treated fish. Discussion. The failure of estrogenic substances in the amounts used here to produce any ef- fect upon ossification is conflicting with the results obtained with this hormone in higher forms. Gardner and his co-workers have reported an increase in endosteal bone for- mation in birds (12, 14, 15) and mice (7, 8, 24 ) and some increase in periosteal bone in the pigeon (25). Sutro (21) reported that there was no change in periosteal bone in estrogen-treated mice, although increases in new bone formation of the medullary cav- ity were noted. The changes in bone forma- tion in rats seemed to be confined to the area round the epiphysis of the long bones, where an increase in the density of this area occurred (Day and Follis. 11). Silber- berg and Silberberg (19, 20) found that both estrogens and androgens intensified the ossification of cartilage and increased the deposition of bone in the guinea pig, but that the effects of the androgens were much less intense. Perhaps an explanation for the lack of increased ossification of the estrogen-treated fish is the absence of true endosteal bone and the presence primarily of membrane bone which has been described for certain parts by Moorkerjee (16). The liter- ature seems to reveal no information at present as to the effects of estrogens upon the membrane bones of higher vertebrates. Pregneninolone with its androgenic action in Lebistes, stimulates ossification in cer- tain of the bones of this fish. This is just the opposite of the effects produced by tes- tosterone in birds and most mammals. 1944] Scott: Effects of Steroids on Lebistes reticulatus 51 Gardner and Pfeiffer (9) and Turner and co-workers (23) found that testosterone propionate inhibited hypercalcification in the mouse and rat. In ducks, Landauer (15) reported that estrogens caused hyper-ossi- fication even in the presence of testosterone. In view of the difference in size of the two sexes of the guppy, the female being the larger, it would be logical to expect that if sex steroids had any effect on ossification, the androgens would stimulate the process thus stopping growth. However, the well es- tablished inhibition of growth in these fish by androgens cannot be attributed to a hyper-ossification, since the latter occurs only in the region of the anal fin. Finally, it should be pointed out that the regions of the skeleton affected by hormone treatment are exclusively those in which a normal sexual dimorphism exists- The changes induced were in character and ex- tent parallel to that expected from the pic- ture of normal sexual dimorphism. In other words, this fish seems to be refractory to overdosage effects and most parts of its skeleton are probably totally unresponsive to sex steroids, except in generalized size difference. The variability and specialization in tele- ost fish is such that experimental results of the type reported here may be applicable only to the species in which they are ob- served. Summary. 1. Most parts of the skeleton of Lebistes reticulatus are not responsive to either es- trogenic (alpha estradiol) or androgenic (pregneninolone) sex steroids, except in size. Pregneninolone caused a reduction in skeletal size, associated with a generalized dwarfing, while alpha estradiol produced no clearly significant change in size. 2. Skeletal constituents associated with the anal fin (the gonopod in the male) differ markedly in the two sexes. This anal fin of the immature fish of both sexes is sup- ported by a group of ten separate inter- haemal bones. These bones are entirely car- tilage in the young fish and in the adult female they are surrounded by a thin layer of membrane bone. In the adult male, these bones are made up of thicker portions of membrane bone than in the female and, in addition, the second, third, and fourth in- terhaemal bones are fused as one. The male- like condition can be fully developed by androgens in either sex and the female con- dition can be induced by estrogens. No changes in the anal fin region could be induced which were greater in extent than those normally seen — that is, there were no apparent overdosage phenomena under the conditions of these experiments. Bibliography. 1. Berkowitz, P. 1937. Effect of estrogenic substances in Lebistes reticulatus Proc. Soc. Exp. Biol, and Med., 36: 416-418. 1938. The effects of estrogenic sub- stances in Lebistes reticulatus. Anat. Rec. 71 : 161-175. 1941. The effects of estrogenic sub- stances in the fish (Lebistes reticu- latus). Jour. Exp. Zool., 87: 233- 244. 4. Bloom, M., W. Bloom, L. V. Domm, and F. C. McLean. 1940. Changes in avian bone due to in- jected estrogens and during the reproductive cycle. Anat. Rec. (Suppl.), 78: 143. 5. Eversole, W. J. 1939. The effects of androgens upon the fish, Lebistes reticulatus. Endo- crinology, 25: 328-330. 6 1941. The effects of pregneninolone and related steroids on sexual develop- ment in fish, Lebistes reticulatus. Endocrinology, 28: 603-610. 7. Gardner, W. U. 1940. Modification of bones of animals receiving sex hormones. Anat. Rec. (Suppl.), 76: 22. 8. Gardner, W. U. and C. A. Pfeiffer. 1938. Skeletal changes in mice receiving estrogens. Proc. Soc. Exp. Biol, and Med., 37: 678-679. 1938. Inhibition of estrogenic effects on skeleton by testosterone injections. Proc. Soc. Exp. Biol, and Med., 38: 599. 10. Hollister, G. 1935. Clearing and dyeing fish for bone study. Zoologica, 12: 89-101. 11. Day, H. G. and R. H. Follis, Jr. 1941. Skeletal changes in rats receiving- estradiol benzoate as indicated by histological studies and determina- tions of bone ash, serum calcium and phosphates. Endocrinology, 28: 83-93. 12. Kirschbaum, A., C. Pfeiffer, J. Van Hau- VERSWYN AND W. U. GARDNER. 1939. Studies on gonad-hypophyseal re- lationshiji and cyclic osseous changes in the English Sparrow, Passer domesticus. Anat. Rec., 75: 249-264. 13. Kyes, P. and T. S. Potter. 1934. Physiological marrow ossification in female pigeons. Anat. Rec., 60: 377. 52 Zoologica: New York Zoological Society 14. Landauer, W., C. A. Pfeiffer, W. U. Gard- ner, and E. B. Mann. 1939. Hypercalcification, -calcemia, and -lipemia in chickens following the administration of estrogens. Proc. Soc. Exp. Biol, and Med., 41 : 80-81. 15. Landauer, W., C. A. Pfeiffer, W. U. Gard- ner and J. C. Shaw. 1941. Blood serum and skeletal changes in two breeds of ducks receiving estrogens. Endocrinology , 28 : 458- 464. 16. Mookerjee, H. J., G. N. Mitra and S. R. Mazumdar. 1940. The development of the vertebral column of a viviparous teleost, Lebistes reticulatus. Jour, of Mor- . phology, 67 : 241-271. 17. Regan, C. T. 1911. The osteology and classification of the teleostean fishes of the order Microcyprini. Ann. Mag. Nat. Hist., 8th series, 7 : 320-327. 18. Ringeon, R. 1940. Seasonal hyperossification of the skeletal system in the female quail. Anat. Rec. (Suppl.), 78: 143. 19. Silberberg, M. and R. Silberberg. 1938. Effects of anterior pituitary im- plants and extracts on the epiphy- sis and joints of immature guinea pigs. Arch. Path., 26: 1209-122-5. 20 1941. Effects of hormones on the skeleton of mice, guinea pigs, and rats. Endocrinology, 29: 475-482. 21. Sutro, C. J. 1940. Effects of subcutaneous injections of estrogen upon skeleton in imma- ture mice. Proc. Soc. Exp. Biol, and Med., 45: 151. 22. Talbot, N. B. 1939. The effect of estrogen on the skele- tal age of rats. Endocrinology, 25: 325-327. 23. Turner, H. H., E. Lachmann, and A. A. Hellbaum. 1941. Effect of testosterone propionate on bone growth and skeletal ma- turation of normal and castrated male rats. Endocrinology, 29: 425-429. 24. Wentworth, J. H„ P. K. Smith, and W. U. Gardner. 1940. The composition of bones of mice receiving estrogens and andro- gens. Endocrinology, 26: 61-67. 25. Pfeiffer, C. A. and W. U. Gardner. 1938. Skeletal changes and blood serum calcium level in pigeons receiving estrogens. Endocrinology, 23 : 485- 491. EXPLANATION OF THE PLATE. Plate I. Figs. 1-4, inclusive, are camera lucida draw- ings made from whole mounts which had been stained with Ali- zarin dye. They are all of the same magnification. Fig. 1. Lateral view of the skeleton of an adult male, in the mid-body region, showing three enlarged haemal spines and the fusion of three of the inter- haemal bones. These characters are typical of the mature male. Fig. 2. Lateral view of the skeleton of an adult female, in the mid-body region, showing small normal haemal spines and ten separate interhaemal bones, typical of the adult female and the immature fish. Fig. 3. Lateral view of the skeleton, in the mid-body region, of a female guppy fed 5 milligrams of pregneninolone from birth for a period of 21 days showing the enlarged haemal spines of the first three caudal vertebrae and the fusion of 5 interhaemal bones. These characters are typical of the adult male. Fig. 4. Lateral view of the skeleton, in the mid-body region, of an adult female, treated with 5 milligrams of pregneni- nolone over a period of 21 days, show- ing slightly enlarged haemal spines of the first three caudal vertebrae and the fusion of the first 4 interhaemal bones. These are typical male char- acters. Fig. 5. A median sagittal section of the mid- body region of an adult female 110 days old, showing separate interhae- mal bones with a central cartilage por- tion which is surrounded by a thin sheath of membrane bone. X 100. Fig. 6. A median sagittal section of the mid- body region of a 110-day-old male treated with 5 milligrams of estradiol from birth. The structure of the sepa- rate interhaemal bones is similar to the female control (Fig. 5). X 100. Fig. 7. A median sagittal section of the mid- body region of a female animal treated from birth with 5 milligrams of pregneninolone over a period 21 days showing fused interhaemal bones and a thick layer of membrane bone surrounding the central portion of cartilage. X 100. Abbreviations. H S Haemal spine. F I S Fused interhaemal bone. ANT Anterior end of skeleton. R Radial bulbous portion of the anal fin to which are attached the fin rays. I S Interhaemal bone or spine. C Cartilage. B Membrane bone. SCOTT PLATE I 16 15 14 13 16 15 14 13 THE EFFECTS OF STEROIDS ON THE SKELETON OF LEBISTES RETICULATUS. Beebe Secondary Sexual Characters in Dynastidae 53 8. The Function of Secondary Sexual Characters in Two Species of Dynastidae ( Coleoptera ) / William Beebe. Director, Department of Tropical Research, New York Zoological Society. (Plates I-V). | This is a contribution from the Forty- third or Venezuelan Expedition of the De- partment of Tropical Research of the New York Zoological Society made under the direction of Dr. William Beebe. The expedi- tion was sponsored by grants from the Com- mittee for Inter-American Artistic and In- tellectual Relations and from four trustees of the Zoological Society, George C. Clark, Childs Frick, Laurance S. Rockefeller and Herbert L. Satterlee, and by invaluable as- sistance from the Standard Oil Companies of New Jersey and Venezuela. For maps and ecological data see Zoologica, Vol. XXVIII, No. 9, pp. 53-59, 1943.] These notes have to do with two species of dynastids, Megasoma elephas Fabr., 1775, and Strategics aloeus Linnaeus, 1758. The Venezuelan name for these beetles is Torna- dor, the borer. Another name is Congarocho, and in the Andes they are known as Bobute. These so-called elephant beetles were ob- served and collected at Caripito, north- eastern Venezuela, in the course of the Forty-third Expedition of the Department of Tropical Research of the New York Zoo- logical Society, during a period of seven months, from February to September, 1942. The photographs, both stills and motion- pictures, were taken by Miss Jocelyn Crane, Research Zoologist on the staff of the de- partment. Only once did I find the big Megasoma and the smaller Strategus under normal con- ditions. On March 18 a female of the latter was observed crawling up the trunk of a small jungle tree, and a few days later two male elephas were discovered resting be- neath an overhanging branch, on a half- rotten log, quite hidden from view. On March 20 I found a female Strategus in a spider web of unusual strength and size, the large rufous-bodied owner frantically wrapping up the struggling beetle. On the 1 Contribution No. 689, Department of Tropical Research, New York Zoological Society. same day, a half-mile away, a member of my staff came across an identical occurence, only here the victim was a male. All others of both species were taken when flying at night against the screened windows of the laboratory, or around the electric lights in the compound, or on the ground on their backs, in early morning, within the radius of the same lights. Their appearance, how- ever, was only during or after a rain. Fewer than fifty of each species were taken during our whole stay. In Megasoma the sexes seemed about equal, but Strategus females dominated almost six to one. From the begin- ning of the rains on April 27, both species became more abundant, several often being taken close together in the mornings beneath the lights of the refinery. None were seen after July 12. Thg great development of horns of vary- ing sizes and shapes on the head and thorax of male beetles of the family Dynastidae has long attracted attention and excited speculation. These specialized structures in connection with the great size of the beetles are reflected in many technical names : Dy- nastes, Megasoma, Megaceros, Goliathus, elephas, hercules, rhinoceros, atlas, etc. Charles Darwin in 1871, in the first edi- tion of “The Descent of Man,” (Vol. I, pp. 371-372) wrote as follows: “The extraordinary size of the horns, and their widely different structure in closely- allied forms, indicate that they have been formed for some important purpose; but their excessive variability in the males of the same species leads to the inference that this purpose cannot be of a definite nature. The horns do not show marks of friction, as if used for ordinary work. Some authors suppose that as the males wander much more than the females, they require horns as a defence against their enemies; but in many cases the horns do not seem well adapted for defence, as they are not sharp. The most obvious conjecture is that they are used by the males for fighting together ; 54 Zoologica: New York Zoological Society [XXIX: 8 but they have never been observed to fight; nor could Mr. Bates, after a careful examin- ation of numerous species, find any sufficient evidence in their mutilated or broken con- dition of their having been thus used . . . The conclusion, which best agrees with the fact of the horns having been so immensely yet not fixedly developed, — as shown by their extreme variability in the same species and by their extreme diversity in closely-allied species — is that they have been acquired as ornaments. This view will at first appear extremely improbable; but we shall here- after find with many animals, standing much higher in the scale, namely fishes, am- phibians, reptiles and birds, that various kinds of crests, knobs and horns have been developed apparently for this sole purpose.” Eight years later, Alfred Russel Wallace in “Tropical Nature” (p. 372) writes of the “immense horns of some beetles of the families Copridae and Dynastidae, which Mr. Darwin admits are not used for fight- ing, and therefore concludes are ornaments, developed through selection of the large- horned males by the females. But it has been overlooked that these horns may be protective. The males probably fly about most, as is usually the case with male in- sects; and as they generally fly at dusk they are subject to the attacks of large- mouthed goatsuckers and podargi, as well as insect-eating owls. Now the long, pointed or forked horns, often divergent, or movable with the head, would render it very difficult for these birds to swallow such insects, and would therefore be an efficient protection, just as are the hooked spines of some sting- less ants and the excessively hard integu- ments of many beetles, against the smaller insectivorous birds.” Passing by some other comfortingly easy explanations we come to the antithesis, given for what it is worth at third hand in “A Year of Costa Rican Natural History” by A. S. and P. P. Calvert (1917). The authors write (p. 69), “Dr. Ohaus has also kindly informed us that a correspondent of his friend Herr Nagel has observed, in Vene- zuela, that the males of Dynast es Hercules fight very violent battles among each other for the females; that they seize and crush with the cephalic and prothoracic horns, the weaker male often having its thorax and elytra crushed, and that the victorious males take the females between the horns and carry them away.” This anonymous information coming to us through three subsequent channels requires, I think, con- siderable confirmation, especially as to the direct crushing power of the horns and the ultimate use of these structures in the Sabine-like kidnapping of the female. In these and numerous other contribu- tions we have a number of suggestions as to the use of these horns : ornaments, direct defence against attacks from enemies, and indirect in making ingestion by large- mouthed assailants difficult or impossible, transportation of the female, overdevelop- ment as sheer, useless impetus of secondary sexual characters, scraping and puncturing bark to induce the flow of edible sap, and fighting among themselves for the posses- sion of the female. Although I have often kept these giant beetles in vivaria in the tropics, I have never seen, either in free or captive specimens, any confirmation of any of these suggested uses of the horns until April 13, 1942, in the Zoological Society’s laboratory at Cari- pito, northeastern Venezuela. For a fortnight I had kept three males and a female Megasoma elephas in a glass battery jar. The quartet of big beetles had fed steadily on over-ripe banana, their method of feeding being to push and nuzzle into the soft pulp until they were quite plastered with the fruit. In a general cleaning of laboratory cages on this 13th of April I cari-ied the jar and its contents to an outside faucet and held one beetle after another beneath the swift stream of water. When picked up, as usual they turned and twisted, the tearing power of their sharp claws and the pinching strength of head armor against the edge of the thorax making it difficult and painful to hold them. During the process of washing they became quiet and did not move a leg again until returned to their cleansed cage. I returned the jar to its place, gave the beetles a supply of fresh banana and forgot them until a half-hour later when I heard a confused sound. I found the food neglected and the three males in an intricate pile, massed around the female. I removed her and from this moment on the battling of the males among themselves occupied consider- able of our attention. Food was completely ignored in the newly aroused heat of battle. When from exhaustion or other cause the combative instinct died down, I could always initiate new and violent encounters by either an application of the water treatment, or the introduction of a female. Two or three days of enforced drought would reduce the war to casual skirmishes, ineffectual feints, and a renewed interest in the mushy fruit. It seemed more than a coincidence that on April 27, only fourteen days after the trans- formation wrought by the artificial deluge of the water faucet, the actual rainy season began. In my precocious breaking of the dry season I had anticipated the effect of the annual rains in unlocking the repro- ductive reactions of these great insects. Concomitant with this suddenly aroused combativeness was nocturnal activity. From now on we were constantly disturbed by loud drummings and reverberations from the beetles, and found that from dark to 1944] Beebe Secondary Sexual Characters in Dynastidae 55 midnight both males and females sought to escape from their cages by flight. The loud banging was due to the impact of the spread elytra against glass and wire, driven by the powerful vibration of the flight wings. This activity ended about twelve or one o’clock and for the rest of the night the beetles rested quietly or fed. Mating of Megasoma elephas: After being drenched with water a pair of beetles would mate almost immediately, whether after long confinement together during which their chief interest was in the banana fodder, or whether both insects had been just caught and placed together. There was no hint of preliminary courtship, no oppor- tunity for appraisal of the horns as orna- ments by the female or sexual selection. If no rival was present, the male went straight for the relatively small female and when within reach hauled her toward him, and mating took place at once. When the male first mounts he wraps his fore legs around the female, sliding them into the lateral crevice between the armor of her head and thorax. In this grip he makes no use of tarsi and terminal claws, but only of the stout, spiny tibia. The second pair of legs hook underneath her body on to the bases of her legs, the ter- minal claws functioning in this case. When copulation is attained his position changes radically. He now rears almost straight up- right, his whole body becoming vertical, resting on his rear tarsi, with the front pair of legs and often the second dangling in the air above the back of the female. This identical procedure was followed in the first three matings and in a fifth. In the fourth, between a newly caught female and the same male in mating number two, the finale was slightly varied. While the male was in quite as vertical a position, he had shifted the hold of his fore legs, which now clung by the terminal claws of the tarsi to the basal joints of the forelegs of the female, while both second and third pairs of legs dangled free in mid-air. This resulted in a triangular support; his junc- ture with the female plus the backward pull with his forelegs. After the water treatment, even if a male was half-immersed in a banana, if a female was dropped into the cage, he instantly became aware of her and rushed in her direction. If a male was substituted the feeding male made no mistake, but, slimy banana and all, went for the new-comer and engaged him head on. In nervousness and quickness his reactions speeded up one hundred per cent. His ordinary activity is a heavy crawling, a slow, bungling creeping with the body dragging. When preparing to mate or fight the body is raised clear and the movements are quick and dynamic- ally directed. In the preliminaries of both activities there is often a series of rhythmic jerks. EGGS: On May 23 we noticed in the cage which contained five Megasoma females that the detritus was in the form of small, rounded balls of dark-colored material, al- though there was nothing in the enclosure but sections of over-ripe banana. In two cages containing males the excreta was nothing but formless masses of food. I dissolved several of the balls but found no trace of eggs. On May 25, however, Miss Crane discovered the first egg lying loose on the bottom, and after this several ap- peared every day, but with no relation to the rounded masses of material. Decayed logs put in with the females aroused no interest whatsoever. The eggs measured 4 by 4.7 mm., the surface being smooth, and ivory white. Fighting of Megasoma elephas : Although the battle between each individual pair of male beetles is, to a certain extent, slightly different from every other, yet there are several fundamental phases which seem in- variable. The opponents meet head on, and either warily wait for the other to attack, or one may rush headlong and begin the encounter. Usually both wait and spar at a little distance. The object first noticeable is an attempt with one or both fore tarsi and claws to trip and unbalance the op- ponent. This is evident in a long series of single photographs and in several complete kodachrome motion picture sequences. There are quick forward lunges and Teachings out with one or both legs, sometimes at the same moment by both insects. This may or may not succeed, but one will force the fighting and the result may be straight pushing and butting for a considerable period, exactly like two antlered deer. Now and then an effort will be noticed to lower the head and get the cephalic horn beneath the other insect. Again and again this is tried, and both may attempt it at the same moment. Then recur the rearing and trip- ping attempts. Periods of rest or waiting may intersperse the encounter and twice I have seen one beetle turn and rush after the female. In both cases the other was after him full speed and the battle began again. The female never remained, but went off as far as the confines of the cage, or in the case of the fight taking place in the open, as far as we would allow her to go, when we would recapture her for fear of losing her in the underbrush. The only certainty was that she showed not the remotest interest in the encounter or in either of her suitors. In all our experience with these beetles, which was invariably in the daytime, we never saw either sex take to wing. Unexpected phases often interrupt the regular succession of the happenings I have 56 Zoologica : Ne w York Zoological Society [XXIX: 8 mentioned. The pull of the sharp, curved claws often unbalances both beetles at once and they lose their footing and roll over and over, all twelve legs tangled together and entailing considerable awkward effort before they separate and face each other again. It is astonishing how loud the clash of horn against horn becomes when one’s ear is close to the fighters. The insects are usually horizontal when they begin pushing against each other, but attempts at tripping will cause both to rear up high on the second and third pairs of legs. Then, if at all, comes the final phase, the all out attempt to get the tip of the curved bifur- cated horn caught in the soft skin of the ventral joint between the thorax and ab- domen. Once secured, we realize this is evi- dently the chief object of the encounter. The successful one puts forth all the strength of which he is capable and lifts again and again with all his might. The higher the other is lifted the more helpless he becomes as his feet, one after the other, leave the ground, and with several super- beetle flings the victim of this grip is thrown over on to his back. Not once, but again and again this was the end result. Often the beetle simply rolled over and came back on his feet again and the whole engagement recommenced, but sometimes he landed on his back and if the surface was at all level and smooth, he spun help- lessly waving all six feet in mid-air. The winner began searching in all directions, evidently for the female. Yet if replaced by himself in a cage he soon settled down to immobility. The whole encounter was reminiscent of the broomstick or cane encounters of our childhood’s parties, where the hands of two boys were tied and a cane inserted behind knees and above elbows and the ensuing encounter was, by manipulation of the ends of the stick or otherwise, to roll the oppon- ent over on his back, when he became as helpless as an upset beetle. The Middleweight Elephant Beetles ( Strategics aloeus ) fought as readily as their larger relations and in almost exactly the same manner. An important difference between the two species is that Strategus has all three horns on the thorax, while Megasoma has the central curved horn on the head itself. Although thus denied the inter-mobility of the horns, the smaller beetles fought with equal fury and quite as satisfactory results. The general plan of battle was identical, to get the anterior horn beneath their opponent and lever him up and over. In one of the first fights watched there were several momentary lockings and once the attacker was himself pried into the air and almost on his back. When upside down this species seemed even more helpless than the larger, and I believe would die of starvation on a smooth surface if left to themselves. The general movement and ac- tivity was less evident owing to the lack of separate play of horn number one, but there was no lack of fierce effort. These accounts have been general ones. The following are notes which I took on two individual encounters : On April 13 two Megasoma males after showing desire for battle were placed by themselves in a large open space and they instantly began fighting. They were the two largest of the three which had been kept together for two previous weeks and during that time had, as I have said, shown no great interest in each other or the female, concentrating solely on resting and feeding. This fight was short. The two rushed to- gether and the horns met with a distinct click. For three minutes they pushed, brac- ing their feet with all possible power. Then separating they did it all over again. The larger one of the two seemed to have the advantage and after three attempts inserted his horn beneath the body of his adversary and actually tossed him clear of himself and the ground. In this case he was on top of the beetle before he could right him- self or attempt to do so, and in the ensuing maze of waving legs, the successful male was almost overcome. Again he upset the smaller and again he foolishly helped him to his feet by rushing upon him. At the third and last upset I distinctly saw the horn of the larger beetle push and tear the membrane of the ventral joint, and exam- ination confirmed this damage, slight though it was. A stalemate of pushing ensuing, I separated them for the night. Disparity in size both between individual males and also sexually is very marked in Megasoma. An average male weighs 32 grams and a female 24 grams. The males vary between weights of 28 grams and 36 grams, with corresponding total lengths of 85 mm. and 103 mm. When a number of males are compared they seem to fall into two general nodes to which we gave the names of Major and Minor. For several days a female and a minor male had been confined together and they had mated. I introduced a major male on May 21, and after righting himself he clambered awkwardly over the small male and toward the female. His antennae played over her back for a few seconds and then the lesser male blundered past him. Like a flash the major turned on the other beetle and the fiercest fight we had seen thus far was on. Both of course tried to get the curved horn under the other, both tried to trip the other off balance. Three times Minor was actually tossed into the air and landed on his back. He levered himself upright and 1944| Beebe Secondary Sexual Characters in Dynastidae 57 after the third event he seemed to become thoroughly aroused and fought twice as hard as before. His very smallness of size was a help in some ways and the locked horn gave the larger insect little advantage. Once the giant was turned over and fell on the female and instantly the minor rushed at both and while they were tangled, butted and drove against them and rolled them about. The larger finally got his opponent in a corner and hooked and twisted violently, securing some strange, secure lock and after a wrench we saw the right middle leg of the smaller beetle break off near its base and lie kicking by itself on the ground. Not for a moment did the injured beetle stop his efforts, but I now retrieved him. The same five-legged minor on the following day mated without trouble with two freshly caught females. After removing the wounded beetle I substituted another major and the combat went on. Both beetles met head on and for at least three minutes, pushing and twisting and jerking with all their might, neither could apparenly break the lock. It was like two deer whose antlers have sprung to- gether beyond all possibility of breaking apart. I separated the beetles and found that the apparent locking was due only to un- interrupted pushing, and that in reality the two beetles were quite free to move apart whenever one of them should relax his efforts. As I said earlier, I neither anticipated nor saw any attempt at carrying the female, and in fact we were unable to push the large-bodied female between the three horns even temporarily, so that such an improb- able feat, as reported in Dynastes Hercules, seems impossible in the present species. Even at the height of violent combative activity, I saw no possibility, either in their encounters or in my handling, of any suffi- cient direct force strong enough to crush or fracture the extremely hard body or elytral armor. The only injuries observed were the slight tearing of the intra thoracic and ab- dominal ventral membrane, and the snap- ping of a leg by oblique leverage. The succession of still photographs of live, unposed beetles taken in two-hundredths of a second presents fairly satisfactory visual realization of this phase of activity of these giant beetles. Proper appreciation of the quickness of movement, the un-scarab swift- ness and deftness of use of legs and horns can only be had from the kodachrome motion pictures of the entire conflicts. Summary: Male beetles of two species of Dynastidae, Megasoma elephas and Stra- tegics aloeus, use their cephalic and thoracic horns for fighting with each other. The initial stimulus is the beginning of the rainy season. Nocturnal and mating activity are consequent upon individual rains. In captivity both can be aroused to highest pitch and culmination by artificial applications of water. The phases or methods in fighting in both species are identical, first an attempt to unbalance the opponent by tripping, and then by ventral attack with the anterior horn to lift and throw him upon his back. 58 Zoologica : New York Zoological Society EXPLANATION OF THE PLATES. These fourteen photographs supplement the verbal portion of this paper, and take the place of the kodachrome motion picture se- quences which are limited to a screen. The individual beetles are the same throughout, but three photographs have been interpol- ated from second and third encounters be- tween the same contestants. These fourteen are selected from a total of sixty-nine. All were taken by by Miss Jocelyn Crane. The battle took place in the compound of the laboratory at Caripito, Venezuela. Pho- tographic data: camera, Leica; lens, Leitz 90-mm. Elmar; film, Eastman Super-XX. Figs. 1-7 were made by sunlight alone, exposure 1/60 sec., f : 12.5 ; figs. 8-14 by synchroflash, exposure 1/200 sec., f : 16. A male elephant beetle, Megasoma elephas approaches a female (Figs. 1 and 2/. While he prepares to mate, a second male ap- proaches (Figs. 3 and 4). The new arrival attacks ( Fig. 5 ) and separates the mating pair (Fig. 6). The beetles meet head on (Fig. 7) and in their efforts to trip each other, rise on their second and third pairs of legs (Fig. 8). As frequently happens, the tripping is mutually successful and both are upset (Fig. 9). They separate at once. In the course of the continued battle the first male succeeds in getting his cephalic horn beneath the head of his opponent (Fig. ■ 0), and exerting all his strength lifts again and again, raising the beetle clear off his feet (Figs. 11 and 12), and rolling or throwing him over (Fig. 13). With the in- terfering beetle helpless on his back, beetle number one begins his search for the fe- male ( Fig. 14). BEEBE. PLATE I. FIG. I. FIG. 3. FIG. 4. THE FUNCTION OF SECONDARY SEXUAL CHARACTERS IN TWO SPECIES OF DYNASTIDAE (COLEOPTERA). BEEBE. PLATE II. FIG. 6. FIG. 7. THE FUNCTION OF SECONDARY SEXUAL CHARACTERS IN TWO SPECIES OF DYNASTIDAE (COLEOPTERA). v FIG. 10. THE FUNCTION OF SECONDARY SEXUAL CHARACTERS IN TWO SPECIES OF DYNASTIDAE (COLEOPTERA). BEEBE. FIG. 9. FIG. 8. PLATE III. BEEBE. PLATE IV. FIG. 12. THE FUNCTION OF SECONDARY SEXUAL CHARACTERS IN TWO SPECIES OF DYNASTIDAE (COLEOPTERA). BEEBE. PLATE V. FIG. 13. FIG. 14. THE FUNCTION OF SECONDARY SEXUAL CHARACTERS IN TWO SPECIES OF DYNASTIDAE (COLEOPTERA). Beebe & Vander Pyl: Pacific Myctophidae 59 9. Eastern Pacific Expeditions of the New York Zoological Society. XXXIII. Pacific Myctophidae. (Fishes).1 William Beebe & Mary Vander Pyl. Department of Tropical Research of the New York Zoological Society. (Text-figures 1-25). (This is the Thirty-third of a series of papers dealing with collections of the Eastern Pacific Expeditions of the Department of Tropical Re- search of the New York Zoological Society, made under the direction of William Beebe. The present paper is concerned with specimens taken on the Arcturus Oceanographic (1925), the Templeton Crocker (1936), and the Eastern Pacific Zaca (1937-1938) Expeditions. For data on localities, dates, dredges, etc., of these expe- ditions, refer to Zoologica, Vol. VIII, No. 1, pp. 1 to 45 ( Arcturus ) ; Vol. XXII, No. 14. pp. 33 to 46 (Templeton Crocker), and Vol. XXIII, No. 14, pp. 287 to 298 (Eastern Pacific Zaca). Contents. PAGE Introduction 59 Myctophidae collected by the Eastern Pacific Expeditions 61 Myctophum affine (Liitken) 61 Myctophum aurolaternatum Garman 64 Myctophum calif orniense Eigenmann & Eigenmann . . 66 Myctophum coccoi (Cocco) 67 Myctophum evermanni Gilbert 73 Myctophum laternatum Garman 74 Mi/ctophum pterotum (Alcock) 75 Myctophum rarum (Liitken) 76 Myctophum reinhardti (Liitken) 77 Myctophum valdiviae Brauer 78 Lampanyctus elongatus (Costa) , 79 Lampanyctus idostigma Parr 81 L,ampanyctus longipes (Brauer) 81 / ompanyctus macdonaldi (Goode & Bean) 82 Lampanyctus macropterus (Brauer) 82 Lampanyctus mexicanus (Gilbert) 84 Lampanyctus omostigma Gilbert 85 Lampanyctus ritteri Gilbert 86 Lampanyctus tenui forme (Brauer) 87 Lampanyctus townsendi (Eigenmann & Eigenmann) .87 Diaphus dumerili (Bleeker) 88 Diaphus gemellari ( Cocco ) 89 Diaphus , pacificus Parr 90 Diaphus rafinesquii (Cocco) 90 Bibliography 91 Introduction. This paper is concerned with the Eastern Pacific Myctophidae taken off Mexico, Cen- tral America and the Galapagos on the Sev- enteenth, Thirty-fifth and Thirty-eighth expeditions of the Department of Tropical Research sent out by the New York Zoologi- cal Society. The Seventeenth or Arcturus expedition in 1925 was in the Pacific for 83 days, dur- ing which time 314 deep-sea hauls were 1 Contribution No. 690, Department of Tropical Research, New York Zoological Society. made. The area covered extended from the Gulf of Panama to Cocos Island and the Gal- apagos, this general area being traversed four times on two round trips. The Thirty- fifth or Templeton Crocker expedition in 1936 was of 59 days’ duration, and 55 hauls were made along the west coast of Cali- fornia, out to Clarion Island, and from Maz- atlan half way up the Gulf of California. The last, Zaca, or Thirty-eighth trip in 1937- 1938, lasted 145 days. Twenty-seven deep-sea hauls were made from Lower California south along the coast of Mexico and Central America to Gorgona Island. In the course of these three expeditions 10,575 fish of the family Myctophidae were captured, and these resolved into 24 species. In 1939 Dr. Rolf L. Bolin wrote “A Review of the Myctophid Fishes of the Pacific Coast of the United States and of Lower Califor- nia.” I quote from him, “The area treated in this report is a semicircular segment of the North Pacific Ocean approximately 1,000 miles in radius, having its center a little east of Point Conception, California, and extending from Cape Flattery, Washington, to Cape San Lucas, Lower California, in- cluding the Gulf of California.” Within this area Bolin records 20 species of myctophids, derived from various collec- tions and museums (Bolin, 1939). One of these, Scopelengys tristis, I hesitate on known evidence to admit within the family Myctophidae. Of the remaining 19, 12 spe- cies occur in our collections. All of the remaining seven which we did not capture, appear to be essentially northern forms, none having been taken south of 31° 33' N. Lat., or the boundary between California proper and Lower California. This is also the northern limit of our collecting. The seven northern species in Bolin’s list and not in ours are, Electrona crockeri, Tarletonbeania crenularis, Diogenichthys scofieldi, Lampanyctus leucopsarus, Lampan- yctus nannochir, Lampanyctus steinbecki and Lampanyctus regalis. The deep-sea family of lanternfishes or Myctophidae appears to be second in abun- 60 Zoologica : New York Zoological Society [XXIX: 9 dance among- all other abysmal fish. In my “Preliminary List of Bermuda Deep-sea Fish” (Beebe, 1937) I recorded the fact that the 57 species of Myctophidae com- posed about 25 per cent, of the entire list of deep-sea species, while the total count of specimens, amounting to about 12,000, was 10 per cent, of the entire catch. This family was exceeded only by that of the genus Cyclothone of the family Gonosto- midae, which, with only three species or 1 per cent, of the whole, in individuals amounted to nearly 94,000 or 82 per cent. In the Pacific, proportions of actual num- bers are of no comparative value, owing to the fewer number of hauls (396 compared with 1,500), much shorter time involved, and a widely scattered area compared with a concentration within an eight-mile circle. On the three Pacific expeditions I took 10,575 myctophids of 24 species, whereas cyclothones were obtained in relatively much fewer numbers than in the Atlantic. The relation between myctophids, cyclothones and other bathypelagic fish in the two oceans as represented in the catches on my expeditions are clearly shown in the accom- panying diagrams. (Text-fig. 1). Atlantic Pacific Myctophids 10,008 10,575* Cyclothones 95,189 15,500 Other fish 10,550 2,200 Totals: 115,747 28,275 As to specific abundance in the two oceans, in the Pacific Myctophum coccoi (4,466) is far ahead, with M. affine next (3,708), and M. laternatum third (1,171). In the Atlan- tic, M. laternatum is first (2,853), and M. benoiti second (1,294). M. coecoi and M. affine are well down the list. * Out of this total, 4,460 were coccoi, all taken on the Arcturus, as compared with 165 taken in Bermuda. On the Templeton Crocker expedition the southern limits were Clarion Island, Cape San Lucas and Mazatlan. In 55 hauls we took 346 myctophids of eight species, all of which were collected on one or both of the other trips. Two species (M. laternatum and L. mexicanus ) out of the seven, comprised 89 per cent, of the whole. On the Zaca trip which included the whole of Lower California and south to Gorgona Island, in the course of 27 hauls, 621 mycto- phids were captured, of 15 species. Here again, two species (M. laternatum and L. omostigma ) predominated, and formed 86 per cent, of the entire catch. Four species ( M . pterotum, L. idostigma, Diaphus du- merili and D. pacificus ) were not taken on the other trips. On the Arcturus trip which was concen- trated in the comparatively limited area between Panama and the Galapagos, 314 hauls yielded 9,608 individual myctophids, resolving into 19 species. For the third time, two species ( M . affine and M. coccoi ) were so abundant that they totalled 84 per cent. Nine species ( M . rarum, M. valdiviae, L. elongatus, L. longipes, L. macdonaldi, L. macropterus, L. tenuiforme, L. townsendi, and D. gemellari ) were confined to the hauls of this expedition. In regard to the general distribution of the 24 species taken by us, 14 are cosmopol- itan (M. affine, M. coccoi, M. laternatum, M. pterotum, M. reinhardti, M, valdiviae, L. elongatus L. longipes, L. macropterus, L. tenuiforme, L. townsendi, D. dumerili, D. gemellari and D. rafinesquii) . Three of these have a restricted distribution in the Pacific (L. longipes, and D. gemellari only in the Galapagos, and M. valdiviae only in Japan and the Galapagos). Eight more species have so far been found only in the Pacific. Five of these are limited to the Eastern Pacific ( M. aurolaternatum, PACIFIC ATLANTIC Text-fig. 1. Relative percentages of deep-sea myctophids, cyclothones and other bathy- pelagic fish taken in the Pacific and in the Atlantic. 1944] Beebe & Vander Pyl: Pacific Myctophidae 61 L. mexicanus, L. idostigma, L. ritteri, and D. pacificus). One reaches Hawaii (L. omo- stigma) , and two extend as far as the west- ern boundary of this ocean ( M . calif orni- ense and M. evermanni) . The remaining two myctophids have the rather curious distribution of Galapagos and the Atlantic (L. macdonaldi and M. i-arum ) . These facts must always be held suspect and tentative from the fragmentary know- ledge which is implied by the necessarily sporadic character of the results of deep-sea trawling. A few facts may be gleaned from the record of breeding females and the total number of eggs ready to be deposited. The height of the breeding season seems to be about June 1, with scattered instances from May 11 to 25 and more rarely as early as April 12. I have no notes between July and November. Graded on number of ripe eggs we have the following list of selected examples: Diam- Length Total ETER Female Eggs MM. mm. Species 120 .28 22 Myctophum valdiviae 400 .45 25 Myctophum laternatum 1200 .3 55.5 Myctophum affine 2400 .27 76 Myctophum aurolaternatum 2700 .28 59 Lampanyctus omostigmx 3300 .25 50 Myctophum coccoi Correlation of number of eggs with size of female seems the only outstanding fact, and to a less degree a relation between the total eggs and their relative diameter. One feature which I hope will be of ser- vice to future workers in the field of Pacific myctophids is the brief annotation of data from the various bibliographies, which has been painstakingly excerpted by the junior author. In the present study, the criterion of adulthood is based on the presence of lumin- ous glands combined with relative size and other superficial external characters. This method necessarily includes many slightly immature individuals which, if classified as in my previous deep-sea papers, would have been considered “transitional adolescents,” since in those the term “adult” was reserved for individuals which were found on dis- section to have mature gonads. This sim- plification of the usual growth stage analy- ses into merely “young” and “adult” was necessitated by present practical difficulties in examining in greater detail such large numbers of specimens. Numerous post-lar- val and other extremely immature mycto- phids lacking all photophores were not in- cluded in this study. By “length” is always understood stand- ard length, measured from the tip of the snout to the end of the vertebral column at the base of the caudal fin. Abbreviations of trawling apparatus are as follows: T=ordinary meter net; Pt=Pe- tersen trawl; Ot=otter trawl; L=surface light. William Beebe Table I. Numerical Tabulation of Species Col- lected on the Three Expeditions Treated in This Paper. Templeton Species Arcturus Crocker Zaca Total Myctophum affine 3,675 1 32 3,708 aurolatemiatum 113 3 9 125 calif orniense 5 8 13 coccoi 4,460 3 3 4,466 evermanni 105 3 108 laternatum 617 129 425 1,171 pterotum 6 6 rarum 1 1 reinhardti 131 24 i 156 valdiviae 35 35 Lampanyctus elongatus 3 3 idostigma 12 12 longipes i . . . 1 macdonaldi 3 3 macropterus 137 137 mexicanus 36 177 4 217 omostigma 246 94 340 7'itteri 19 1 20 tenuiforme 5 5 townsendi 1 1 Diaphus pacificus 19 19 dumerili 3 3 gemellari i 1 rafinesqun 19 4 i 24 Totals : 9,608 346 621 10,575 Myctophidae Collected by the Eastern Pacific Expeditions. Myctophum affine (Liitken, 1892). Specimens taken by Easterii Pacific Ex- peditions: 3,675 specimens; Arcturus Ocea- nographic Expedition; surface to 800 fath- oms; between 2° 00' S. and 6° 58' N. Lat., and 83° 34' and 91° 53' W. Long.; lengths 15 to 67.5 mm.; between March 31 and June 19, 1925. 1 specimen; Templeton Crocker Expedi- tion; at the surface; 25° 02' N. Lat., and 115° 52' W. Long.; length 22 mm.; May 19, 1936. 32 specimens; Eastern Pacific (Zaca) Ex- pedition; surface to 500 fathoms; between 4° 30' and 28° 42 N. Lat., and 78° 35' and 117° 50' W. Long.; lengths 6 to 66 mm.; between Nov. 8, 1937, and March 25, 1938. Sgecimeiis previously recorded: 296 spe- cimens; ? to 96 mm.; 0 to 1,914 fathoms; north and south Atlantic, Mediterranean, Indian Ocean, Japan, East Indies, Tasman Sea, Hawaiian Islands, central and eastern Pacific from lower California to Chile. 62 Zoological New York Zoological Society [XXIX: 9 Text-fig. 2. Distribution of Myctophum affine. Photophore count and variation: 42 males, 21 females and 421 young fish examined as to relative number of photophores on both sides of the body. The mean is very evident- ly 8-5 on both sides of the body, as 221 out of the 484, or 45.6 per cent., possess this combination. The extremes in numbers are 6-9 and 4-6. Parr in counts on Atlantic spe- cimens found as many as 10 in the anterior, and as few as 3 in the posterior AO. In addition to the dominant pattern of 8- 5, only seven other numbers are found which show svmmetrical lateral distribu- tion, 7-4, 7-5, 7-6, 8-4, 8-6, 9-4 and 9-5. The remainder fall irregularly into a total of 31 variations on the two sides. The three groups show no age or sexual differences in total numbers or in variations of photophores. Luminous glands: The youngest fish in which traces of male, supracaudal luminous glands are visible, measures 29 mm., and shows two glands. The youngest female with glands showing is 34.5 mm. in length, and also has two glands. The largest male, 66 mm., has seven glands, and the largest female of 67.5 mm. possesses four glands. This gradual increase in both sexes with age, is confirmed by five-millimetre succes- sive averages as follows : males, from 29 to 66 mm. 2, 3.2, 3.6, 3.8, 4, 5, 5.5 and 7 glands; females, from 34.5 to 67.5 mm. 2, 2.15, 2.25, 3 and 4 glands. Three specimens taken on the Arcturus expedition (Cat. Nos. 5281, 5281a and 6059; lengths 44, 54.5 and 59 mm. ) show both supra- and infracaudal luminous glands. The first shows an unusually small number of photophores, viz: 7-5, 7-5; and the third fish possesses almost the highest number, 9- 5, 9-6. No. 5281 shows two dorsal and two ventral glands, No. 5281a has two dorsals and three ventrals, and No. 6059 has two dorsal and five ventral glands. Sex proportions : Out of the total of 3,708 specimens collected, 63 are adult; 42 males and 21 females, a proportion of two to one. Breeding: Two egg counts are as follows: Female No. 5281c, length 51.5 mm.; taken April 12, off Galapagos; total egg count, both ovaries ca. 1,400, eggs measuring .34 mm. in diameter. Female No. 6058b, length 55.5 mm.; taken June 1, off Cocos Island; total eggs ca. 1,200; .3 mm. in diameter. Anatomy: Typical adult male, No. 6058a, taken with above female (No. 6058b) and two young at Station 74, surface haul T-35. Length 54.5 mm. ; air bladder large, 14.5 by 4.5 mm. ; stomach heavily pigmented with black, 7.5 by 2.2 mm. (for food see separate section). The pyloric opening surrounded, at a distance of 3 mm. along the gut, by 9 caeca, which measure from 3 by .8, to 4.5 by 1.1 mm. The intestine has but one convolution to break the straight line from the point at which the caeca enter it to the anal opening; total length 20 mm. The testes measure 13 by 1.5 mm. and lie along the ventral side of the air bladder. The kidney occupies the space between the slightly pigmented lining of the body cavity and the vertebral column, and thickens anteriorly to form a large lobe over the stomach. Pouch of gut from which the pyloric caeca open with two large nema- tode parasites. The accompanying female (No. 6058b) possesses 11 pyloric caeca. Seven parasitic worms coiled partly in caeca, in pouch and also in upper intestine. Development: Of the total catch of this species, 3,645 are young, showing no signs of luminous glands, and measuring from 6 to 27.5 mm. in length; 42 males measure 29 to 66 mm. ; and 21 females measure from 34.5 to 67.5 mm. Vertical distribution: Of the 3,708 spe- cimens, the vertical distribution is as fol- lows : Depth Fathoms No. of Fish 0 3,286 200 406 400 8 500 6 800 2 The surface fish were all taken after dark, with the exception of 21 individuals, comprising one-half of one per cent., all young, which were captured at the surface in daylight. This would indicate that Myctophum af- fine is essentially a nocturnal surface swim- mer ( 88 per cent, of the entire collection being captured at this level), descending to extremely dark depths (200 fathoms maxi- mum) during the hours of daylight. As the nets were non-closing it seems likely that the 16 fish taken at 400, 500 and 800 fath- oms, may actually have entered the nets on their way to the surface. Sociability : Seasonally, at least, it would seem certain that Myctophum affine lives in large schools. As will be seen from the ac- companying table, nine nets contained from 100 to i,100 fish each. These totalled 3,213, or 86 per cent, of all this species collected. Especially significant is the fact that eight 19441 Beebe & V ancle?- Pyl: Pacific Myctophidae 63 of these nine hauls were made between April 14 and 29. All the fish in these captured schools were young, except for 16 adults taken in the single non-April haul of June 13. Also all except one net was a surface haul, and that contained 406 fish from 200 fathoms. No. OF Depth Cat. Date Fish Fathoms No. April 12 124 Young- 0 5281 April 14 280 Young 0 5234 April 15 266 Young 0 5332 April 20 1,100 Young- 0 5373 April 21 406 Young- 200 5507 April 22 456 Young 0 5387 April 23 111 Young- 0 5424 April 29 306 Young 0 5507 June 13 164 ( 148 Young) 0 ( 14 Males) ( 2 Females) 6399 Aside from the above nets, 52 others averaged only eight fish to a haul, and in 17 there was only a single fish each. Food: The stomachs of 10 specimens con- tained only copepods, with a few small am- phipods and fish eggs. Luminescence and viability: Three notes on the lights of this species will show the more usual reactions after capture. A 19 mm. individual was taken at Station 30 T-l, Cat. No. 5133. In an aquarium it survived for ten minutes. For a considerable time it showed no light and then the entire fish was outlined by almost every one of the 80-odd organs. Then it would become completely dark and light up again, this occurring about every 15 seconds. Rarely several spark-like rays shone forth, penetrating, clear and brilliant. I happened to lift my wrist watch with its dully luminous dial close to the fish and it reacted at once, giving out two strong discharges of the caudal glands. Con- cealing the watch and later displaying the light of its face resulted in instant reac- tion. This happened eight times. I then flashed on my much stronger flash-light with no result. For five minutes I alternated the two artificial sources of illumination with identical results, the fish reacting vig- orously to the watch dial, but paying no attention to the electric torch. From the haul Station 63 T-l a 20 mm. myctophid was taken to the dark room, but although apparently dead, throughout ten minutes it showed only one source of light, that from the pair of terminal, sub-mandi- bular, branchial organs. Twenty other fish were lively. At first the presence of numer- ous Noctiluca and Sapphirina made accurate observation difficult, but when clear water was substituted, two general types of il- lumination were distinguishable, an indirect glowing, constant but impossible to restrict to any organ or group of organs. The entire fish glowed with a generalized light, the scale mirrors gleaming in a marvelous man- ner, which gave the appearance of coming from some outside source in the room, were it not that I was sitting in absolute darkness. The other lights were sudden, abrupt, short, sharp, distinct flashes as from a make-and- break apparatus. After an hour of watching, when rigor mortis seemed to have set in, and all movement had ceased, a short re- newal of the lights could be brought about by shaking the water or pinching the fish. Net T-2 of Station 88 on the Arcturus was a surface haul made at 9 P. M. on June 13, 1925, three miles north of the island of Albemarle, Galapagos. A number of full grown Myctophum affine were taken at once down to the dark room where they darted about a small aquarium. A slight tap on the glass was sufficient to increase the constant luminescence of the lateral photophores, but this light was at all times so steady that it was easy to distinguish every organ and hence to make certain of the species. The lumination was light yellowish-green, as near as I can give it a name. As the fish grew weaker this light did not die out but became dull and indistinct. The ventral batteries were stimulated ir- regularly and unexpectedly. It was difficult to say whether a tap on the glass, or colli- sion with another fish, or some unidentifiable emotion was the cause, or perhaps all three at various times. As with other species of this genus, I never saw the secondary sexual caudal lights function when the fish was alone; only when several were together and in a state of excitement or fear, as on the first introduction to an aquarium, or at a sudden disturbance due to the adding of more water or fish. Under normal conditions of environment this light may well have a sexual function, as in attracting the oppo- site sex, but in the narrow, artificial confines of an aquarium, its principal effect on human spectators was a momentary blinding, with immediate visual disappearance of the fish. The power of this glandular light is very much greater than that given off by all the other organs together. The entire dark room was momentarily illumined when the supra- or infracaudal glands flared out. There seemed no difference in intensity between the light of the males or females. In daylight the general appearance of the fish was deep purple, in sharp contrast to the brilliant silver scales of Myctophum coccoi. Study Material. A total of 3,706 specimens was taken, as follows: Arcturus Expedition, Station 30 T-l (1) ; 31 surfaced); 33 T-2(35); 38 T-l (6) ; 38 Pt-2 (1 ) ; 39 T-l(l) ; 39 T-2 ( 8 ) ; 39 T-3 (124) ; 39 T-5(3); 41 T-l (6); 41 T-3 (279 ) ; 45 T-l (266) ; 47 T-l (1100) ; 49 T-l (406); 49 T-2(l); 50 T-2 ( 1 ) ; 50 64 Zoologica: New York Zoological Society [XXIX: 9 T-5 (456 ) ; 52 T-l (38 ) ; 52 T-l ( 1 ) ; 52 T-l (2 ) ; 52 T-2(lll); 52 T-3(3); 53 T-3(4); 57 T-l (306); 62 T-l(l); 63 T-l ( 1 ) : 63 T-l (10); 65 Ot-1 and T-3 and T-4 ( 2 ) ; 65 surface (8); 74 T-l ( 1 ) ; 74 T-23(17); 74 T-26 (2 ) ; 74 T-29(2) ; 74 T-35 to T-40(4) ; 74 T-41 (8) ; 74 T-43 and T-44(4) ; 74 T-61 to T-66(3 ) ; 84 T-2(l); 84 T-5 and T-6 (5) ; 84 T-14 ( 1 ) ; 84 T-16H); 84 T-16 ( 13 ) ; 84 T-15 to T-17 (61 ) ; 85 T-l (30) ; 86 T-4 (2) ; 86 T-7(l) ; 86 T-8(l) ; 87 T-2 ( 4) ; 87 T-3 ( 1 ) ; 87 T-6 ( 1 ) ; 88 T-l and T-2 to T-4 ( 164) ; 88 T-l (3 ) ; 88 T-2 (2 ) ; 88 T-3 (20 ) ; 88T-4 ( 5 ) ; 90 T-l and T-2 ( 18 ) ; 91 T-l (28); 91 T-l and T-2 (12); 91 T-2 (29). Templeton Crocker Expedition, Station 170 L-l(l). Eastern Pacific ( Zaca ) Expedition, Station 177 L- 1(2); 225 T-l (1) ; 227 T-l (3); 228 T-l (1) ; 231 L-l(25). For detailed trawling data, refer to Zoologica, Vol. VIII, pp. 6-23 ( Arcturus ) ; Vol. XXII, pp. 37-46 (Templeton Crocker) ; Vol. XXIII, pp. 287-298 (Zaca). References and Synonymy. (Synonymy accepted according to Bolin, 1939, p. 110). Scopelus affinis: Liitken, 1892, p. 252, fig. 10. (Brief de- scription; key; ? to 37 specimens, including type; ? mm.; ? fathoms; mid-Atlantic, off south Africa, Indian Ocean). M. ( Myctophum ) affine: Brauer, 1903, p. 190, figs. 105-7. ( partim : Synonymy and description; ? to 76 speci- mens; ?-78- ? mm.; 380 to 1,914 fathoms; Gulf of Guinea, Cape Verde, Bay of Bengal, Indian Ocean, Seychelles Islands).. Zugmayer, 1911, p. 27. (1 specimen; 34 mm.; ? to 275 fathoms; Mediterranean). Pappenheim, 1914, p. 193. (4 specimens; 15 to 19 mm.; ? fathoms; north Atlantic, near Cape Verde, mid-south Atlantic). Bolin, 1939, pp. 92, 110, fig. 11. (Key and description; 9 specimens; 28.7 to 72 mm.; ? fathoms ; open Pacific, off Guadalupe Island, Atlantic Ocean). Myctophum opalinum: Goode & Bean, 1895, pp. 72, 511, fig. 81. (61+ specimens, including types; ? mm.; 104 to 1,536 fathoms; northwest Atlantic). Jordan & Evermann, 1896, p. 571. (Brief description) . Waite, 1904, pp. 153, 193. (1 specimen; 95 mm.; ? fathoms; Lord Howe Island, near Australia). Breder, 1927, p. 17. (? specimens; 25 to 50 mm.; ? fathoms; West Indies). Myctoph um nitidulum : Gar man, 1899, p. 266, pi. LVI, fig. 3. (Type description; ?' specimens; ? mm.; ? fathoms; northwest of Hawaiian Islands). Rhino scopelus oceanicus: Jordan & Evermann, 1903, p. 68, fig. 15. (General description; 2 specimens; type and co-type; 25 mm.; surface; near Hawaii). Myctophum margaritatum : Gilbert, 1905, p. 596. (General descrip- tion ; ? specimens, including type of 80 mm. ; surface; near Hawaii). Myctophum affine: Goode & Bean, 1895, p. 72. (Brief des- cription) . Jordan & Evermann, 1896, p. 570. ( Brief description; key). Lonnberg, 1905a, p. 63. (1 specimen; ? mm.; surface; South Atlantic, off Brazil). Gilbert, 1908, p. 21/. (Brief discussion of range and anal photophores; ? specimens; ? mm.; ? fathoms; mid-Pacific, south to Marquesas) . Gilbert, 1911, p. 14. (Name listed; 1 spe- cimen ; ? mm. ; ? fathoms ; west of Gala- pagos ) . Gilbert, 1913, pp. 75, 76. (Key and discus- sion of anal photophores; 17 young speci- mens; ? fathoms; Japan). Gilbert, 1915, p. 312. (Synonymy and dis- tribution; 1 young specimen; surface; Cat- alina Island, off coast California). Weber, 1913, p. 87. (Synonymy and dis- tribution; 2 specimens; 14 mm.; ? to 821 fathoms ; Banda Sea ) . Weber & Beaufort, 1913, p. 161. (Syno- nymy and description). Jordan & Jordan, 1922, p. 11.. (Name listed ) . Parr, 1928, pp. 65, 69-74, figs. 8, 9. (Key; synonymy; general and detailed description, including study of light and dark pigmen- tation; 32 specimens; ? mm.; surface; Ba- hamas). Parr, 1934, p. 45. (Synonymy). Taning, 1928, p. 53. (Key to species). Fowler, 1928, p. 69, fig. 13. ( Synonymy and description). Norman, 1929, p. 514. (Brief reference). Norman, 1930, p. 325. (8 specimens; 12 to 30 mm.; surface to 300 fathoms; south Atlantic) . Beebe, 1929, p. 15. (Preliminary listing; 4 specimens ; ? mm. ; surface to 800 fath- oms; Hudson Gorge, North Atlantic). Beebe, 1937, p. 204. (Preliminary listing; 37 specimens; 13 to 25 mm.; 400 to 1,000 fathoms; Bermuda). LeGendre, 1934, p. 336. (Key to species). hiyctophum aurolaiernatum Garman, 1899. Specimens taken by Eastern Pacific Ex- peditions: 113 specimens; Arcturus Oceano- graphic Expedition; surface to 800 fath- oms; between 2° S. and 6° 40' N. Lat., and 80° 48' and 91° 53' W. Long.; lengths 18 to 82.5 mm.; between April 12 and June 19, 1925. 1944] Beebe & Vander Pyl: Pacific Myctophidae 65 3 specimens; Templeton Crocker Expedi- tion; at the surface; between 18° 44' and 22° 30' N. Lat., and 110° 15' and 114° 20' W. Long.; lengths 31 to 80 mm.; between May 7 and 9, 1936. 9 specimens; Eastern Pacific ( Zaca ) Ex- pedition ; surface to 500 fathoms ; between 4° 30’ and 9° 15' N. Lat. and 78° 33' and 85° 10' W. Long.; lengths 21 to 37 mm.; be- tween Feb. 7 and March 26, 1938. Specimens previously recorded: 31 spe- cimens ; ? to 90 mm. ; surface ; mouth of the Gulf of California and off Panama. Our spec- imens extend the range south to Cocos Island and the Galapagos. Text-fig. 3. Distribution of Myctophum aurolaternatum. Photophore count and variation: Out of the total of 125 fish taken, 115 were avail- able for photophore counts, the remaining 10, from 18 to 23 mm., being too young to show any trace of these light organs. In the 115 fish there were 21 different combinations of the anterior and posterior AO photo- phores, 11 of which were asymmetrical, showing different combinations on the two sides. With 10 symmetrical variations, the balance between the two was much more even than in Myctophum affine. There were two very evident numerical nodes, 10-6, 10-6; and 11-6, 11-6, with 16 fish showing the first, and 29 the latter ar- rangement. Fifty per cent, of the fish pre- sented one or the other of these two combi- nations. The extremes in actual number of photophores were 9-6, 9-6 totalling 30, and 12-6, 12-6 with a total of 36. Far from showing any correlation with age, the small- est (18 and 20 mm.) and the largest (82.5 mm.) shared the largest number of photo- phores. Luminous glands: Out of a total of 125 fish, only 11 showed evidences of the supra- and infracaudal luminous glands, of which I six were males and five females. Unlike the condition in Myctophum affine there was no correlation of size with number of glands, the males showing 45 mm. (5 glands), 60 mm. (7), 66 mm. (6), 70 mm. (9), 75 mm. (8), and 80 mm. (7 glands). In a 68 mm. female the glands were barely discernable, while a 70 mm. male and female each had 9 well developed. Bolin records a fish of 60.5 mm. showing no trace of the glands. Sex proportions: As already stated, of the 11 specimens with sexually distinct lu- minous glands, six were males and five were females. Breeding: Three dissected fish showed the following conditions: Female No. 5539, 68 mm. in length, taken April 23 off the Galapagos, had the infracaudal glands barely discernable, and the ovaries with no enlarged eggs, showing no signs of immin- ent breeding. Female No. 25580, 80 mm. long, captured on May 8, 180 miles southwest of the tip of Lower California, had four small caudal plates, and the ovaries quiescent, with no enlarged eggs. Female No. 6059, 76 mm. long, taken June 1, 60 miles south of Cocos Island, showed three well developed infracaudal glands. The ovaries were in breeding condition, with ap- proximately 1,200 eggs in each ovary, large and about to be laid. Scattered among them were many very small eggs. Development : Out of 125 fish taken, 114 were glandless young, measuring from 18 to 41 mm. Eleven showed the glands of sexual differentiation, 45 to 82.5 mm. lengths, al- though these are not necessarily to be classed as full-grown adults (see under Breeding ) . Vertical distribution: All 125 specimens of Myctophum aurolaternatum were taken at the surface after dark, or at depths which negatived surface conditions of sunlight. In all, 109 (including the 11 large, sexually distinct individuals) or nearly 88 per cent, of the total catch, were taken at the surface, mostly in the evening from 6 to 10 P. M., but a few as late as 4 A. M. The remaining 16 fish were distributed so evenly between 200 and 800 fathoms that they may be accredited to casual captures, perhaps entering the nets at higher eleva- tions than the extreme depths of the hauls indicate. The actual data are; 200 fathoms (1 fish), 300 (1), 400 (2), 500 (6), 600 (3), 700 (2), and 800 fathoms (1 fish). The fact that none were taken in daylight surface hauls suggests that there is a decided mi- gration downward during daylight hours. Sociability: Very convincing evidence is shown of schooling, since 66 specimens, representing 53 per cent, of the total 125, were taken in four nets. The details are as follows: 25 nets (1 fish each), 9 (2), 2 (3), 2 (4), 1 (6), 2 (8), 1 (16), and in 1 net (30 fish ) . Food: Representatives of three Phyla were found in the stomachs of Myctophum aurolaternatum: Coelenterata, Crustacea and Mollusca. The contents of 10 stomachs are typical of the food in general. All these fish were taken at the surface at night: Cat. No 5324 66 Zoologicn : New York Zoological Society (XXIX: 9 (36 mm.) 1 hyperian amphipod, 1 pycno- gonid; 5539 (68 mm.) 3 amphipods; 6057 (41 mm.) 1 caridean; 6057 (37 mm.) 1 stalk-eyed shrimp, 3 amphipods, 3 copepods; 6059 (76 mm.) 4 copepods; 6061 (71 mm.) 4 isopods; 6061 (41 mm.) 1 isopod, 2 cope- pods; 6270 (38 mm.) 1 small squid, red eye with accessory light organs, 2 copepods; 6270 (36 mm.) 1 small squid, red eye, no light organs, 3 copepods; 6421 (75 mm.) 2 euphausids, 2 siphonophore segments, 3 zoea. Luminescence and viability: Even in the specimens taken at the surface, viability was so poor that only three fish survived transportation to the dark room from the lighted deck. In all of these the lateral AO photophores were glowing with a faint bluish radiance. From one fish, three short, very intense, pale flashes were recorded, so powerful that the entire fish, its jar and our hands were brilliantly illumined. Parasites: Four of the larger specimens, with well-developed glands, had large para- sitic copepods attached to the sides, all with a pair of dangling ovaries. Study Material. A total of 125 specimens was taken, as follows: .4 returns Expedition, Station 39 T-3 ( 1 ) ; 41 T-3 ( 1 ) ; 45 T-5(3) ; 47 T-l (2) ; 49 T-l C 1 > ; 50 T-l (1) ; 52 T-l ( 1 ) ; 52 T-2 (2 ) ; 57 T-l (2) ; 58 T-l(l) ; 59 T-8 (1 ) ; 59 T-9 ( 1 ) ; 61 Pt-l(l); 61 T-4(l); 61 T-5 ( 1 ) ; 62 T-l (2) ; 65 T-4(3) ; 65 Ot-1 to Ot-4 (8) ; 73 T-l (2); 74 T-26(l); 74 T-35 to T-40 ( 16) ; 74 T-40(l): 74 T-41 (2 ) ; 74 T-42 (30) ; 74 T-41 and T-42C6) ; 74 T-43 and T-44 ( 8 ) ; 74 T-69 to T-71(2>; 74 T-70 ( 1 ) ; 84 T-15 to T-17(4) ; 84 T-2K1) ; 86 T-4 ( 1 ) ; 87 Pt-1(2) ; 88 T-2(l) ; 90 T-l and T-2 ( 1 ) ; 91 T-2(l). Templeton Crocker Expedition, Station 160 L-l(l); 161 L-l(l) ; 162 L-l(l). Eastern Pacific (Zaca) Expedition, Station 210 T-3(l); 210 L-l (2) ; 210 T-8 ( 1 ) ; 228 T-l (1) ; 231 L-l (4 ). For detailed trawling data, refer to Zoologica, Vol. VIII, pp. 6-23 ( Arcturus ) ; Vol. XXII, pp. 37-46 (Templeton Crocker) ; Vol. XXIII, pp. 287-298 (Zaca). References and Synonymy. ( Svnonymy accepted according to Bolin, 1939,' p. 108)'. Myctophum aurolaternatum : Garman, 1899, p. 264, pi. IV, fig. 3. (4 large and 21 small specimens, including type; ? to 90 mm. ; surface; Gulf of Panama. Description and reference to luminosity ) . Parr, 1928, p. 65. (Key to species). Parr, 1934, p. 43, fig. 2. ( Examination and redescription of type specimens). Mycto'phum ( Myctophum ) aurolaternatum: Bolin, 1939, pp. 92, 105, 108-109, fig. 10. (6 specimens; 31 to 72.4 mm.; ? fathoms; Gulf of California and off Panama). Myctophum californiense E i g enmann & Eig- enmann, 1889. Specimens taken by Eastern Pacific Ex- peditions: 5 specimens; Templeton Crocker Expedition; surface to 500 fathoms; be- tween 23° 25' and 30° N. Lat., and 108° 31' and 116° 27 W. Long.; lengths 31 to 38 mm.; between April 29 and May 23, 1936. 8 specimens; Eastern Pacific (Zaca) Ex- pedition; surface; 31° 25' N. Lat., and 116° 58' W. Long.; lengths 30 to 44 mm.; taken on Nov. 7, 1937. Specimens previously recorded: 13 speci- mens; ? to 12/ mm.; ? to 534 fathoms; both shores of the Pacific, on the east from Puget Sound to Cape Lucas, and in the west, off Japan. Material in the present paper ex- tends the known range about 600 miles south, 48 miles east of Arena Point, Lower California. Text-fig. 4. Distribution of Myctophum. californiense. Photophore variation: There is little var- iation in the number of anal photophores in our 13 specimens. The commonest pattern is a symmetrical 7-9, shown by 8 specimens, or more than 60 per cent, of the fish. Eleven out of the 13 possess symmetrical patterns. Other data: The 13 fish were all young, showing no trace of luminous glands. The extremes of measurements were 30 and 44 mm. One fish of 42 mm. taken at night at the surface had eaten 2 copepods and 1 mysid. Of the 13 specimens, 11 were taken with dipnets as they came to our night lights, three on May 23, 1926, and eight on November 7, 1937. The remaining two, al- though taken in a 500-fathom net, were prob- ably much nearer the surface when they entered. Bolin says, “In life the luminous organs emit a pale greenish light.’’ Study Material. A total of 13 specimens was taken, as fol- lows : Templeton Crocker Expedition, Sta- tion 158 T-4 (2) ; 174 L-l (3). Eastern Paci- fic (Zaca) Expedition, Station 176 L-l (8). For detailed trawling data, refer to Zoolo- gica, Vol XXII, pp. 37-46 (Templeton Crocker); Vol. XXIII, pp. 287-298 (Zaca). 1944] Beebe & Vander Pyl: Pacific Myctophidae 67 References and Synonymy. (Synonymy accepted according to Bolin, 1939, p. 106). Myctophum calif orniense: Eigenmann & Eigenmann, 1889, p. 124. (Type specimen; ? mm.; 45 fathoms; Cortez Banks). Jordan & Evermann, 1896, p. 572. (Key and description). Gilbert, 1913, p. 78. (3 specimens; all 127 mm. ; ? fathoms ; off Santa Barbara Islands) . Townsend & Nichols, 1925, p. 10. (1 spe- cimen ; ? mm. ; 534 fathoms ; southwest of Santa Barbara Islands). Parr, 1928, p. 64. (Key and synonymy). Parr, 1929, p. 10, fig. 4. (Further descrip- tion of type). Bolin, 1939, pp. 92, 106, fig. 9. (8 speci- mens ; 24 to 103.3 mm. ; ? fathoms ; eastern Pacific off Central California, south to Mex- ico; key to species and description). Scopelus calif orniense: Liitken, 1892, p. 267. (Reference). Myctophum ( Myctophum ) humboldti: Brauer, 1906, p. 192. (Part of this des- cription confused with Myctophum calif or- niense) . Myctophum coccoi ( Cocco , 1829). Specimens taken by Eastern Pacific Ex- peditions: 4,460 specimens; Arcturus Ocea- nographic Expedition ; surface to 1,000 fa- thoms; between 6° 58' N. Lat., and 2° 00' S. Lat.; and between 80° 48' and 91° 47' W. Long.; lengths 13 to 51 mm.; between April 3 and June 19, 1925. 3 specimens; Templeton Crocker Expedi- tion; surface; at 18° 44' N. Lat., and 114° 20' W. Long.; and at 21° 20' N. Lat., and 115° 14 W. Long.; lengths 20 to 29 mm.; May 9 and 17, 1936. 3 specimens; Eastern Pacific ( Zaca ) Ex- pedition; surface; 4° 30' N. Lat., and 78° 33' W. Long.; lengths 13 to 20 mm.; March 26, 1938. Specimens previously recorded: 996+ specimens; ? to' 110 mm.; ? to 2,620 fath- oms; in all parts of the Atlantic Ocean, north, south, east and west, as well as mid- ocean, between 40° S. Lat. and 48° N. Lat. ; - Mediterranean Sea, Cape of Good Hope, south of Zanzibar, Bay of Bengal, Indian Ocean, east and west of Australia, Admir- alty Islands, Philippines, Japan, mid-Pacific, south-east of Hawaii, between Galapagos and Marquesas Islands, Gulf of Panama, coast of Central America, off coast of Chile. In the eastern Pacific Myctophum coccoi is very evidently more at home in southerly warm waters, as witness the 4,400 speci- mens taken on the Arcturus, between Pa- nama and the Galapagos, this number being almost half the total of all the members of this family taken on the whole of this expe- dition. On the other two expeditions only 6 were taken, all at the surface at night lights. One swam into a dipnet about 50 miles north of Gorgona off the Columbia coast, and a single specimen came from well off shore at a latitude only 100 miles south of Cape Lucas, forming the northernmost Pacific record for this species. Thus 20 degrees north is the limit in the Pacific, whereas in the Atlantic, coccoi is found as far north as 48 degrees. Forty degrees south seems the abrupt deadline for the species all around the world. Within these latitudinal limits the fish may be said to be almost cosmopolitan. Text-fig. 5. Distribution of Myctophum coccoi. Photophore count and variation: There is a great variation in the number of anal photophores, even in the same specimen, fish after fish having a different count on either side. This has nothing to do with sex or age. A count of the anal photophores on more than 200 specimens, reveals that the most common combination (76 specimens) is 6-11. The next most common is 6-10. To summarize, there are six combinations in 136 fish, 35 combinations in 73 fish, or a total of 41 separate combinations in 209 fish. Previous studies on about 1,500 freshly caught fish resulted into their division into dominant variants from the standard num- bers of anal photophores, there being eight distinct types in addition to the more nor- mal ones. Types of variations other than Numer- ical, in Anal Photophore Groupings. 1. No break between anterior and poster- ior groups, one side only. 2. No break between the posterior group and the preeaudal photophores. 3. Anterior group divided between second and third photophores. 4. Anterior group irregular in spacing because of the apparent insertion of an extra photophore. 5. One photophore in series out of line with the rest. 6. Break in either group or in the pre- caudals. 68 Zoologica: New York Zoological Society [XXIX: 9 7. Three precaudals instead of two, on one or on both sides. 8. Smaller photophores in the series. The extent of the eight variants in 1,104 individuals is unexpectedly great. Including males, females and older young, 57.8 per cent, possess the normal complement and arrangement of anal photophores, while those fish showing one, two or three of the variations comprise the remaining 42.2 per cent. Luminous glands: As in nearly related species, the presence of supracaudal glands indicate male fish more or less adult, while the female is associated with infracaudal glands. Ontological development is shown by the following 70 fish, 37 of which were males, and 33 females. Males of 34 to 40 mm. average 4.4 glands ; those of 40 to 45 mm. 4.4 glands; males 45 to 50 mm. 5.2 glands. Females of 34 to 40 mm. possess an average number of 3 glands; those 40 to 45 mm. 3.5 glands; those of 45 to 50 show 4 glands. This indicates a slow but certain increase in number of glands with increase in standard length. Sex 'proportions: Eleven hundred and four M. coccoi were taken at two Arcturus stations in the Pacific, Station 39, T-l ; and 84, T-15 T-16 and T-17 l. This number com- prised 24.7 per cent, of the entire Pacific catch of this species. Of these 1,104 fish, 619 were young, 253 males and 232 females. In a few other small catches the females actually equalled the numbers of the males, so that occasionally the balance is quite even be- tween the sexes. The total count of all 4,466 Pacific M. coccoi resolves into 45.7 per cent, young, 30.6 per cent, males, and 23.7 per cent, females. Breeding: A full grown female of 50 mm. length, taken at the surface off the Galapagos on April 12, had ovaries 9 mm. in length. The egg diameters were .25 mm. and the total count of the eggs in one ovary was 1,648, or about 3,300 eggs altogether. Vertical distribution and migration: The number of M. coccoi taken below the surface is negligible, except in the case of a single net drawn at a depth of 200 fathoms, in the evening, from 7.40 to 8.35 P. M. and which contained 42 fish. Deeper hauls averaged 300 fathoms (4 fish), 400 (2 fish), 500 to 1,000 fathoms (1 fish each). The total of sub-surface fish is 59 as compared with 4,407 from the surface. Doubtless the great majority, if not all, of the 17 individuals from below 200 fathoms were adventitious, probably entering at or near the surface during the brief passage of the net at the upper levels. The daily migration is evident from the following table, giving the number of fish drawn at successive half hours. A. M. hauls were actually consecutive. Hour No. of Fish A. M. 12.00-12.30 39 1.00- 1.30 26 2.00- 2.30 25 3.00- 3.30 10 4.00- 4.30 10 4.30- 5.00 13 6.00- 6.30 2 fish taken between 6.30 A.M. and 6.30 P.M. P.M. 6.30- 7.00 5 7.00- 7.30 63 7.30- 8.00 113 8.00- 8.30 186 9.00- 9.30 145 9.30-10.00 88 This sequence was confirmed by several Station series well removed from one an- other. At Station 74 I made half-hour hauls around the clock, and again fifteen-minute hauls at 5 to 7 o’clock both morning and evening, to confirm the regularity of ap- pearance and disappearance. This slender- tailed lantern fish comes up from rather shallow depths just after dark on the equa- tor and rapidly reaches its greatest abund- ance, or perhaps at this time, 8-9.30 P. M., keeps in very dense, compact schools. Throughout the night its numbers in the nets decrease, probably from a general scat- tering in search of food, and the last indi- vidual dives into darker levels at 6.30 A. M. at the very latest. Judging from the contents of our nets, the younger fish appear at the surface first, and seem to vanish sooner than the older ones, but this last statement is less certain ibecause of the suddenness with which all descend. Rough water seems to deter them but little in their vertical movements, but a much more potent factor is the presence or absence of moonlight. On a cloudy night a half hour tow brought in 40 fish, while on the following night, in clear, nearly full moonlight, a similar net at a corresponding hour captured only two. Fifteen additional instances all provided identical data. Only rarely have I observed this lantern fish in its native element. At the foot of the companionway with an electric light cluster swung low over the water, on three occa- sions I have observed this species, and twice captured it. I can account for this unexpec- ted positive phototropism only by a reversal of reaction due to some factor in the greater intensity and concentration of the illumina- tion. As my nightly tows show, coccoi is pres- ent in large numbers at the surface of a wide area of the Pacific. Whenever many individuals have been taken in one net, the variation in age has been great, very young and fully developed adults being mingled together. In general shape and in the re- markably nondeciduous scales coccoi differs greatly from its generic relations. 1944] Beebe & Vander Pyl: Pacific Myctophidae 69 Sociability: While the summation of catches presents a uniform and logical pic- ture of vertical distribution, migration and sociability, irregularities are now and then apparent. For example, two nets at Station 39, T-l and T-2 drawn simultaneously, par- allel, but from different sides of the stern, in a half hour’s time caught respectively 4 and 72 fish, illustrating the concentration of schooling in certain places and at definite times. A third net attached ten minutes after the other two, captured 600 M. coccoi in thirty minutes. The record for numbers in a short period of time was at Station 88, on the equator in the Galapagos, when four small nets, one- half metres, were pulled from 8.10 to 9.00 P. M.; 9.00 to 9.30; 9.35 to 9.40 and 9.45 to 9.50. The count of M. coccoi was 293, 315, 116 and 549. The last two hauls of five minutes each thus captured more than 66 fish a minute, and this in nets only 18 inches in diameter. The 1,273 coccoi here taken in a little more than an hour and a half, re- present almost one-third of the entire catch on the Arcturus in the Pacific. If we assume that 40 or more coccoi taken in a single net constitute evidence of school- ing, then we find that out of 52 nets pulled on the surface, 25 come within our assump- tion. These include captures of 40 to 600 individuals, in all 4,147 fish, or about 92 per cent, of the entire catch, an average of 165 coccoi to each net. Food: The slender-tailed lanternfish is altogether a plankton feeder. In the begin- ning of our deep sea work, my assistants several times brought me individual fish which had partly ingested the entire head of a fellow M. coccoi. This astonishing sight was subsequently observed a number of times, and proved to be wholly accidental. The newly caught fish, gasping and gaping widely, wriggling about at full speed in a small aquarium, sooner or later ran against and partly engulfed the head of a smaller fish. A haul at the surface on June fifth, near the Galapagos, at Station 78, T-l, yielded 40 Myctophum coccoi. These were all exam- ined at once for the contents of their stom- achs with the following results: 1. 3 small calanids, 5 hyperids, 3 Cory- caeus, 1 euphausid. 2. 12 calanids, 3 hyperids, 1 Halobatis, 1 small Atlanta. 3. 1 calanid, 5 hyperids, 1 Atlanta, 1 Pontella. 4. 21 calanids, 1 Candace, 1 Monops, 1 Pontella, 1 Sapphirina, 3 hyperids, 1 euphausid, 1 Atlanta, 1 Limacina. 5. 15 calanids, 1 Corycaeus, 5 hyperids, 1 Monops, 1 Limacina. 6. 18 calanids, 3 hyperids, 1 mysid, 1 fish egg, 1 Globigerina. 7. 22 calanids, 4 hyperids, 3 Atlanta, 3 Globigerina, 1 Limacina. 8. 2 Monops, 1 schizopod, 1 fish egg, 13 calanids, 1 Sagitta. 9. 9 hyperids, 21 calanids, 25 Limacina, 1 Atlanta, 1 Monops. 10. 8 calanids, 3 hyperids. 11. 9 calanids, 6 hyperids, 2 Limacina, 3 Sagitta, 3 Globigerina. 12. 14 calanids, 4 hyperids, 2 Sagitta. 13. 7 calanids. 14. 8 calanids, other material. 15. 2 hyperids, 5 Sagitta, 2 Corycaeus. 16. 1 hyperid. 17. 3 Sagitta, 4 calanids, 1 hyperid, 5 fish eggs. 18. 1 hyperid, 1 euphausid, 8 calanids, 4 Limacina, 1 ostracod. 19. 7 hyperids, 1 Monops, 2 fish eggs, 1 Globigerina. 20. 8 hyperids, 1 Ianthina, 2 fish eggs, 1 Limacina. 21. 2 hyperids, 1 euphausid, 3 calanids. 22. 2 Candace, 3 fish eggs, 6 hyperids. 23. 3 small mollusks, 2 hyperids, 2 small worms. 24. 14 Candace, 2 hyperids. 25 to 40. Fifteen stomachs empty or with unrecognizable material. The general aspect of this food of M. coccoi resolves into the following: Annelids No. Times Eaten 2 No. Individuals 2 Atlanta 5 7 Calanids 17 183 Corycaeus 2 5 Euphausids 4 4 Fish eggs 7 15 Gastropods 1 3 Globigerina 5 9 Halobatis 1 1 Hyperids 21 80 Ianthina 1 1 Limacina 6 34 Monops 4 5 Mysids 2 2 Ostracods 1 1 Poyitella 3 3 Sagitta 5 14 Sapphirina 1 1 Schizopods 1 1 89 371 Carrying this finally to groups, we find that fish eggs have entered 7 times into the diet; insects once; mollusks 13 times, crus- taceans 55 times ; annelids 77 times and protozoans 5 times. Viability: Roughness of water seems to make very little difference to this fish, and on nights when the sea was rough and the 70 Zoologica: New York Zoological Society [XXIX: 9 swell heavy, I have still taken them in abundance. At such times other species of myctophids have remained below, perhaps in part due to the delicacy of attachment of their scales. Coccoi can stand an aston- ishing amount of buffeting and handling without the loss of a single scale. Viability is another quality which sets coccoi apart. A net full of mixed surface fish will always have living coccoi after all the others have died. Not only this, but there is marked individuality within the species. One fish which we named “Methu- selah” lived on and on after I had subjected him to all sorts of experiments, dropped him on the floor twice, and placed him for a moment by mistake in a bowl of formalin. The moment a slender-tailed lanternfish is removed from the tow net and placed in a container of clear salt water, he begins swimming violently and ceases activity only with death. Other myctophids sometimes rest or relax somewhat their violent efforts at escape, but coccoi never. The contour of the body, slender and compressed anteriorly and drawn out into an elongated narrow posterior, compels a wholly different method of progression from that of its stouter, more rounded relatives. There is no steady or slightly undulatory movement, but a violent wagging of the entire body and tail, such as is seen in small elvers and murae- noids. In a still living lanternfish the tail can be bent around more than 360 degrees from the anterior portion of the body and tail, forming more than a complete circle. Much of this entire flexibility is apparently utilized at each progressive effort. As in the case of many other deep-sea fish, the first instinct on being exposed to light, is to descend. This is the case both in shallow dishes and in deep aquaria. The fish makes one straight dive to the bottom, and from then on gyrates head down- ward, twisting and pushing, endeavoring to achieve the only known method of escape. From this and other facts, it seems certain that danger threatens these small beings almost altogether from above. Once when a short haul had provided some unusually virile, active lanternfish, by placing several immediately in an aquarium of running water, I managed to have one remain at the surface. It swam obliquely with the head at the very surface, and around and around rapidly, wriggling violently and covering every inch of the water. Furthermore, when the lights were turned out in the aquarium room, a large individual which had been bumping its head off against the bottom, suddenly veered upward, and wiggled to the surface, to dash about for a minute like its fellows. A sudden glare of light sent them both down to a few minutes more of gyration and then death. Unexpectedly and with no cessation of the frantic wriggling movements I have seen fish at several different times, turn, seize and swallow small forms of plankton life which swam close to them, this in dif- fused room light plus the ventral sheet of luminescence. Luminescence: The abundance and the unusual viability of Myctophum coccoi made possible the observation of these fish in the dark room of the Arcturus. Imagine a minnow which is iridescent copper above and silvery white below, not over two inches in length, with large eyes and moderate fins. A full-grown fish weighs a gram, which means that it would take about 450 to make a pound. It feeds on copepods, amphipods, floating snails and other minute plankton fry, and from this food it generates suffi- cient energy to swim, to make daily migra- tions up and down, to illumine one hundred lights and to deposit upwards of two to three thousand eggs. Scattered over the body are many, small, round luminous organs, which we may di- vide into three general sets. First, 32 ventral lights on each side of the body, extending from the tip of the lower jaw to the base of the tail ; second, about 12 lateral lights arranged irregularly along the head and body, and third, a series of three to six median light plates or scales, either above or beiow the base of the tail. The lower battery, when going full blast, casts a solid sheet of light downward, so strong that the individual organs could not be detected. Five separate times when I got fish in a large, darkened aquarium, I saw good-sized copepods and other organ- isms come close, within range of the ventral light, then turn and swim still closer to the fish, whereupon the myctophid twisted around and seized several of the small beings. Once it turned completely on its back. I could never have seen this except that the glass sides reflected sufficient light. Whether this is the chief object of the ventral lighting I do not know, but it was at least effective on the several occasions when I was watching. Perhaps the best distinction between var- ious species of lanternfish is the arrange- ment of the lateral light organs, and in the darkroom in absolute darkness I could tell at a glance what and how many of each spe- cies were represented in a new catch, solely from their luminous hieroglyphics. When several fish were swimming about, these side port-holes were almost always alight, and it seems reasonable that they may serve as recognition signs, enabling members of a school to keep together, and to show stray individuals the way to safety. The light scales of the tail are apparently of considerable importance. Ordinarily when the whole fish is glowing with the pale greenish light of luminescence, these caudal 1944] Beebe & Vander Pyl: Pacific Myctophidae 71 lights are seldom seen. A clue to their use is found in the fact that they show a sexual difference, the males having them on the upper side of the peduncle and the females on the lower side. In my brief and sporadic observations, when no fish lived longer than thirty-six hours, there was no chance to watch signs of courtship or any such use which these lights might subserve. But when a fish exerted itself unduly to get out of the way of another, either of its own or another species, these lights would flash and die in quick succession. Three separate times in unusually strong, vigorous fish when the body luminescence was very dim, these scale searchlights flashed like heliographs, being much stronger than the combined, steadier glow of all the others. This luminescence was of a much deeper green than that of the ventral lights. If continuously alight, a single fish would have illumined the whole darkroom, and enabled one easily to read fine print. In the dark it was thus possible to dis- tinguish species of lanternfish by the lateral hieroglyphics and the sexes by the upward or downward direction of the tail lights. I have never seen the latter illumination given out by a fish swimming alone in an aquar- ium. Although it is very evident that the caudal flashes have some sexual significance, yet another very important function seems that of obliteration. It certainly was to my eyes and I have no reason to think that a natatory enemy might not also be frus- trated. When the ventral lights die out they do so gradually, so that the eye holds the image of the fish for a time after their disappearance, but the eye is so blinded by the sudden flare of the tail lights that when they are as instantly quenched, there fol- low several seconds when our retina can make no use of the faint, diffused, remain- ing light, but becomes quite blinded. A better method of defense and escape would be difficult to imagine. Specimens of Myctophum coccoi taken around midnight seemed to show less bril- liant light and more intense color, a deeper turquoise green than those captured earlier. It seems possible that the color varies inversely in strength with increase of lu- minescence (at least in its effect on the human eye), that is, the more powerful the actual luminescence, the more it ob- scures the specific color. In lighted sur- roundings the luminous activity is mini- mized, and the color at maximum, while as the luminescence increases in intensity the color shows up less. To repeat, specimens taken under the above conditions show the color very well, whereas the luminescence was not as intense as that in fish caught at the surface soon after dark. This was ob- served independently by several of us again and again and seems to indicate a certain relative independence of light intensity and color functioning. Enemies: While the number of the ene- mies of myctophids must be legion there is one which seems to consider these small fish as an almost exclusive item of diet. Not long after the first of the coccoi come to the surface, a somewhat larger, black fish is to be found in the same nets. In the glare of the laboratory lights this was not a very unusual appearing fish, although it had a short, dependent chin tentacle and a mouth with exceedingly wide gape. Its name is Astronesthes. Almost every specimen taken of this fish contained a recently swallowed coccoi, usually measuring about two-thirds the length of its enemy. In the dark, Astron- esthes itself is a gorgeous sight, the skin covered with a host of minute luminous specks, while the fins fairly glowed with paie green light. Strangely enough, the stem only of the tentacle was luminescent, while the slightly enlarged tip was dark. Parasites : Myctophum coccoi appears to be more susceptible to parasitic infection than is the case with other species of mycto- phids. For example, in a female 35 mm. in length, (Cat. No. 6397 ) , three parasites were found in the body cavity; a larval cestode of the genus Rhyncobothrium, lying free in the left side between an ovary and the intestine. It measured 7 by 2.2 by .62 mm. and was transparent enough to show the proboscides. On the right side of the body cavity a 7 mm. larval acanthocephalid was situated slightly above and partly inside the pancreas. In front of the liver, between it and the pig- mented peritoneal lining and just below the oesophagus, partly coiled around it, was an immature nematode, a 6 mm. Anisakis. Study Material. A total of 4,446 specimens was taken, as follows : Arcturus Expedition, Station 33 T-l(l); 33 Ptl(l); 33 T-2 ( 58) ; 38 Pt-2 ( 2 ) ; 38 T-l (8) ; 39 T-l (4) ; 39 T-2 ( 71 ) ; 39 T-2(l) ; 39 T-3 (70) ; 39 T-3 (529); 39 T-3(l); 39 T-5 ( 87 ) ; 40 T-l (2); 41 T-l (2); 41 T-3 (135) ; 45 T-l (29) ; 45 T-l (1 ) ; 47 T-l (44) ; 47 T-l (27); 47 T-l (12) ; 47 T-l (2); 47 T-l (61); 49 T-l (42); 50 T-l ( 1) ; 50 T-5 (105) ; 52 T-l (6) ; 52 T-2 (24) ; 53 T-3 (2 ) ; 57 T-l (424); 58 T-l (5); 62 T-l (2); 63 T-l (3) ; 65 T-l (2) ; 65 T-3 ( 2 ) ; 65 T-4 ( 10 ) ; 65 surface) 10 ) ; 67 T-l(2);68 T-l (4); 68 T-5 (1 ) ; 70 T-l (3) ; 73 T-l (5) ; 74 T-10(6) ; 74 T-ll (20) ; 74 T-23(94); 74 T-26(39) ; 74 T-29 (43 ) ; 74 T-35 ( 2 ) ; 74 T-35 to T-40 ( 78 ) ; 74 T-41(24); 74 T-43 and T-44 ( 18 ) ; 74T-46 ( 13 ) ; 74T-61 to T-66 (50 ) ; 74 T-70(l); 77 T-l (17); 84 T-2 (2); 84 T-5 (41) ; 84 T-6(2) ; 84 T-8(3) ; 84 T-14(l) ; 84 T-15 to T-17 ( 223 ) ; 84 T-16(282); 84 T-20 (2 ) ; 85 T-l (103); 86 T-5 ( 1 ) ; 86 72 Zoologica: New York Zoological Society [XXIX: 9 T-8 ( 1 ) ; 88 T-l and T-2 to T-4 (49 ) ; 88 T-l (274) ; 88 T-2 (296); 88 T-3(11G); 88 T-4 (512) ; 90 T-l and T-2 (163); 91 T-l (174); 91 T-l (2); 91 T-2 (16). Templeton Crocker Expedition, Station 162 L-l ( 2 ) ; 167 L-l(l). Eastern Pacific ( Zaca ) Expedition, Station 231 L-l (3). For detailed trawling data, refer to Zoologica , Vol. VIII, pp. 6-23 ( Arcturus ) ; Vol. XXII, pp. 37-46 (Temple- ton Crocker); Vol. XXIII, pp. 287-289 ( Zaca ) . References and Synonymy. (Synonymy accepted according to Parr, 1928, p. 61 ). Scopelus coccoi: Cocco, 1829, p. 143. (Reference not seen). Cocco, 1838, p. ?, pi. II, fig. 6. (Reference not seen ) . Bonaparte, 1832-41, p. 138. (Description). Cuvier & Valenciennes, 1828-49, vol. 22, p. 440. (Brief description; 1 specimen; no details ) . Gunther, 1864, vol. 5, p. 413. ( Descrip- tion; ? specimens; ? mm.; ? fathoms; Gulf of Guinea, Mediterranean and Atlantic). Gunther, 1889, vol. 31, p. 30. (Descrip- tion; 80 specimens; ? mm.; ? fathoms; south of Cape Verde, near West Indies, south Atlantic, south Pacific, mid-Pacific, near Admiralty Islands and near Japan I. Canestrini, 1870, p. 125. (Reference not seen ) . Doderlein, 1878-79, p. 54. (Name listed only) . Raffaele, 1889, p. 182, pi. 7, fig. 5. (De- scription ) . Moreau, 1891, p. 90. (Description). Collet, 1896, p. 116. (Description; syn- onymy; 42 specimens ? mm.; surface; Azores and south of Flores). Gilchrist, 1908, p. 165. (150 specimens; ? mm.; surface; Cape Point, South Africa). Thompson, 1916, p. 83. (Name listed, with synonymy). Scopelus gracilis: Lutken, 1892, p. 255, fig. 13. (Descrip- tion ; 5 specimens ; ? to 50 mm. ; ? fathoms ; open Atlantic near Madeira, mid-north At- lantic, off Cape of Good Hope, south of Zanzibar, off coast of Brazil ) . Scopelus jagorii: Peters, 1859, p. 411. (Reference not seen). Scopelus langerhansi : Johnson, 1890, p. 454. (1 specimen; 110 mm.; ? fathoms; Madeira). Lutken, 1892, p. 267. (Reference only). Scopelus ( Rhinoscopelus ) coccoi: Lutken, 1892a, p. 243. (? specimens; ? to 50 mm. ; ? fathoms ; large number of stations, both east and west, north and south Atlantic, east and west of Australia, south Africa, Zanzibar. All recorded on dis- tribution map). Cams, 1889-1893, p. 564. (Description). Rhinoscopelus coccoi: Goode & Bean, 1895, p. 90. (Brief de- scription; ? specimens; ? mm. ;_ surface to 2,620 fathoms; northwest Atlantic from Newfoundland to Gulf of Mexico, Bermuda, Madeira ) . Jordan & Evermann, 1896, p. 568. (De- scription ) . Fowler, 1928, p. 70, fig. 14. (References; synonymy; brief description). Rhinoscopelus tenuiculus : Gilbert, 1908, p. 222. (Brief description; 1 specimen; 50 mm.; ? fathoms; southeast of Hawaii). Gilbert, 1911, p. 15. (Brief description; 12 specimens; 35 to 50 mm.; ? fathoms; midway between Galapagos and Marquesas Islands) . Jordan & Jordan, 1925, p. 11. (Name listed only). Alysia loricata: Lowe, 1839, p. 87. (Brief description). Lowe, 1849, p. 14. (Brief description). Myctophum coruscans: Richardson, 1844-48, vol. 2, p. 40, pi. 27, figs. 1-5. (Description; 1 specimen; 57 mm.; ? fathoms; south Atlantic, Australia). Scopelus coruscans: Hutton, 1873, p. 270. (Brief description). Myctophum hians: Richardson, 1844-48, vol. 2, p. 41, pi. 27, figs. 19-21. (Description; 1 specimen; 50 mm.; ? fathoms; ? locality). M. (Myctophum) hians: Brauer, 1906, pp. 162, 194, fig. 12. (Key to species; description; synonymy; 1 spe- cimen; 22 mm.; 1,903 fathoms; south At- lantic) . Barnard, 1925, p. 244. (Brief description). M yctophum tenuiculum : Garman, 1899, p. 262, pi. J, fig. 5. (De- scription; 1 specimen; 25 mm.; surface; Gulf of Panama ) . Myctophum gracile: Goode & Bean, 1895, p. 74. (Brief descrip- tion). Jordan & Evermann, 1896, p. 572. (Brief description ) . Myctophum ( Rhinoscopelus ) coccoi: Roule & Angel, 1933, p. 28. (1 specimen; 53 mm. ; surface ; Azores ) . M. ( Myctophum ) coccoi: Brauer, 1906, pp. 163, 199, figs. 116-120. (Key to species; description; synonymy; 380 specimens ; ? to 52 mm. ; between Azores and Newfoundland, north of Bermuda, West Indies, north of Bahamas, Gulf of Mexico, 1944] Beebe & Vander Pyl: Pacific Myctophidae 73 south Atlantic, coast of Brazil, Madeira and Canary Islands, Cape Verde, Gulf of Guinea, Indian Ocean, Zanzibar, Admiralty Islands, East Indies). Pappenheim, 1914, p. 194. (93 specimens; 20 to 46 mm.; ? to 1,642 fathoms; south Atlantic, Cape Verde). Barnard 1925, p. 244. (Brief description). Myctophum coccoi: Weber & Beaufort, 1913, pp. 153, 158. (Key to species description; synonymy). Regan, 1916, p. 134. (Description of larval and postlarval fishes of New Zealand and 3 Kings Island; 1 specimen; 20 mm.; 2 fath- oms). Sanzo, 1918a, p. 135. (Description of young). Taning, 1918, pp. 67-70, figs. 23, 24. (De- tailed description of adult and postlarval fish ; distribution map ; 64 specimens ; 5 to 34 mm. ; surface to ? fathoms ; eastern Mediterranean) . Taning, 1928, p. 55. (Key to species). Taning, 1932, p. 113, figs. 1, 2. (Descrip- tion of adults and postlarval; synonymy). Barnard, 1925, p. 236. (Key to species). Parr, 1928, p. 61. (Key to species and synonymy). Parr, 1934, p. 45. (Brief discussion). LeGendre, 1934, p. 337. (Key to species). Beebe, 1937, p. 204. (Preliminary list; 165 specimens; 7 to 26 mm.; 300 to 1,000 fathoms; Bermuda). Myctophum evermanni Gilbert, 1905. Specimens taken by Eastern Pacific Ex- peditions: 105 specimens; Arcturus Oceano- graphic Expedition; surface to 1,000 fath- oms; between 6° 40' N. Lat., and 0° 03' S. Lat., and 80° 49' and 89° 50' W. Long.; lengths 18 to 73 mm.; between Feb. 23 and July 29, 1925. 3 specimens; Eastern Pacific ( Zaca ) Ex- pedition; 1 at 300 and 2 at 500 fathoms; between 9° 09' and 8° 03' N. Lat., and 83° 12' and 85° 04' W. Long.; lengths, 16, 16 and 18 mm.; Feb. 8 and March 10, 1938. Of these myctophids 80 have been avail- able for study. Specimens previously recorded: 41 speci- Text-fig. 6. Distribution of Myctophum evermarmi. mens; surface to 1,100 fathoms; Hawaii, Marquesas, East Indies and Japan. Our specimens are the first recorded from the eastern Pacific. Photophore variation: Unlike other spe- cies of this genus the lateral photophores show no bilateral asymmetry. The pattern of 8-5, 8-5 is the most common, comprising 61 per cent, of the whole. The extremes are 7-5 and 8-6. Luminous gland.s: Of 80 specimens, 44, from 16 to 31 mm. show no signs of caudal plates. Seventeen males, 40 to 69 mm. pre- sent 6 to 8 supracaudal plates. There is a decided increase in number with size, i.e., 6 fish with 6 plates average 48 mm.; 10 with 7 plates average 57 mm. and 1 fish with 8 plates is 56 mm. in length. Nineteen are females and starting with a single plate faintly visible in a 43 mm. spe- cimen, the series ends with four well-devel- oped plates and a length of 73 mm. The corresponding series of lengths in the as- cending plate scale is 45, 52, 53, 61, 61 and 67 mm. Breeding: Specimen No. 6059, 68 mm. in length, taken June 1, is a female in full breeding condition with ovaries measuring 28 by 5.5 mm. Vertical distribution: Eighty - five per cent, of the total catch was taken at the surface, all after dark. The 12 specimens brought up from open nets descending to 300 and 600 fathoms are so evenly dis- tributed that they may well have entered at considerably more shallow depths. All 36 recognizable males and females were taken at the surface. Sociability : Schooling, especially of older fish, is definitely shown from the fact that 43 of the 80 (including all the sexually recognizable specimens) comprising 54 per cent, of the whole, were taken in three hauls at the surface at night, while the re- mainder averaged one and one-half fish to the net. In one haul of 10 young fish, all were of exactly the same age, 22 mm. in length. Food: In ten stomachs there were remains of three phyla; Crustacea, Mollusca and Pisces. The following groups were present in this order of abundance; copepods, 44 individuals of at least six species; 29 am- rhipods ; 13 pteropods of three species ; 13 ruphausids; 4 schizopods; 3 larval fish; 2 shrimp. Study Material. A total of 84 specimens was taken, as follows: Arcturus Expedition, Station 29 surfaced); 31 surfaced); 47 T-l(10); 59 T-9 (3) ; 65 T-3(l); 65 T-4(2); 65 T-4(l) ; 66 T-2 (3) ; 66 T-2 (1 ) ; 67 T-l(l) ; 68 T-5 (1) ; 73 T-l(4); 74 T-4(l) ; 74 Pt-2 (1) ; 74 T-10(l) ; 74 T-10(l); 74 T-ll(l); 74 T-15(l) ; 74 T-24(3) ; 74 74 Zoologica: New York Zoological Society [XXIX: 9 T-26 (2) ; 74 T-29(l); 74 T-35 ( 0 ) ; 74 T-42 ( 10) ; 74 T-43(23); 74 T-61(3); 74 T-74 (1 ) ; 90 T-l(l); 91 T-2(2). Eastern Pacific ( Zaca ) Expedition, Station 210 T-7 ( 2 ) ; 219 T-2(l). For detailed trawling data, refer to Zoologica, Vol. VIII, pp. 6-23 ( Arcturus ) ; Vol. XXIII, pp. 287-298 ( Zaca ) . References and Synonymy. Myctophum evermanni: Gilbert, 1905, p. 597, pi. 70, fig. 2. (Holo- type; 29 mm.; surface; south of Oahu, Hawaii ) . Gilbert, 1908, p. 218. (11 specimens; ? mm.; ? fathoms; 1,300 to 1,500 miles south- east of Hawaii and to Marquesas). Gilbert, 1913, p. 80. (15 specimens; ? mm.; surface; off southeast Japan). Brauer, 1906, p. 162. (Key to species). Weber, 1913, pp. 86, 87. (4 specimens; 20 to 48 mm.; ? — 550 to 1,100 fathoms; East Indies). Weber & Beaufort, 1913, p. 162, fig. 63. (Description and figure). Jordan & Jordan, 1922, p. 11 (Name listed ) . Fowler, 1928, p. 70. (10 specimens; ? mm.; ? fathoms; Hawaii). Parr, 1928, p. 65. (Key to species). Parr, 1929, p. 10. (Reexamination of type specimen, and considers it adequately described and figured by Gilbert). Norman, 1929, p. 514. (1 specimen in British Museum, identified by Gunther as M. caninianum, reexamined and referred to M. evermanni; ? mm.). Myctophum laternatum Garman, 1899. Specimens taken by Eastern Pacific Ex- peditions: 617 specimens; Arcturus Ocea- nographic Expedition; surface to 1,200 fa- thoms; between 2° 00' S. and 6° 27' N. Lat., and 81° 08' and 91° 53' W. Long.; lengths 8 to 28 mm.; between April 13 and June 13, 1925. 129 specimens; Templeton Crocker Ex- pedition; 300 to 550 fathoms; between 20° 36' and 26° 48' N. Lat. and 108° 31' and 113° 25' W. Long. ; lengths 5 to 27 mm. ; between March 28 and April 29, 1936. 425 specimens; Eastern Pacific {Zaca) Expedition; 300 to 500 fathoms; between 5° 10' and 17° 45' N. Lat., and 78° 42' and 103° 05' W. Long.; lengths 5 to 28 mm.; be- tween Nov. 23, 1937 and March 26, 1938. Specimens previously recorded: 2,973 spe- cimens; surface to ca. 2,734 fathoms; north and south Atlantic Oceans, Indian Ocean, Japan and East Indies, eastern Pacific from San Diego to Galapagos. Photophore variation: This species shows very general bilateral symmetry in the dis- tribution of anal photophores, only four out of 75 fish possessing different numbers on Text-fig. 7. Distribution of Myctophum laternatum. the two sides of the body. The dominant pattern is 6-3 (41 fish or 54.5 per cent.), and a second node at 7-3 (16 fish or 21 per cent. ) . The extremes are 5-3 and 8-3. As these last are exhibited by full-sized fish (26 mm.) and by young ones (14 mm.) respectively, it will be apparent that age has nothing to do with relative numbers of photophores. Luminous glands: One hundred and seventy fish selected at random from the entire catch were examined for luminous glands. This lot of fish resolved into 87 males and 35 females, the remaining 48 being juvenile, glandless fish. The antorbital photophore is a reliable sex differentiation character, the older males having it large and well-developed, while in the females it is somewhat smaller. In the juvenile group, from 8 to 14 mm. no trace of glands was visible. In the males, from 17 to 20 mm. the gland is of medium extent, (1.5 to 1.7 mm. in a 17 mm. fish). In large males of 20 to 28 mm., the gland is elongate (averaging 2.5 mm. in a 25 mm. fish). Sex proportions: If the above mentioned proportion of the sexes holds good among larger numbers, it means the females stand in the proportion of only two-fifths of the number of males. Breeding: Four breeding females, Nos. 5642, 5671, 5718 and 5719, taken from May 11 to 25 off Cocos Island, eastern Pacific, average 25.5 mm. in length. The diameter of the eggs is .45 mm. and the average length of the ovaries is 6.5 mm. The number of eggs in each ovary vary from 181 to 210, approximately 400 eggs ready to be deposited in each female. Vertical distribution: M. laternatum is peculiar in our hauls in that there is no sharp demarcation of vertical limits, no special emphasis on certain depths. The average fish per net is as follows : surface (14 fish per net), 300 fathoms (13), 400 (9), 500 (16), 600 (10), 800 (26), 1,000 (18), and 1,200 fathoms (15 fish per net). Sociability : Twenty nets contained only a single fish each, but in 14 there were from 1944] Beebe & Vander Pyl: Pacific Myctophidae 75 21 to 99. The average of these was 43 fish per net. The 602 fish in these 14 nets re- present more than 50 per cent, of the whole, so the evidence for schooling is reasonably positive. The depths at which these larger numbers were taken are as follows : 300 fathoms (2 nets), 450 (1), 500 (8), and 600 fathoms (3 nets). Study Material. A total of 1,171 specimens was taken, as follows: Arcturus Expedition, Station 26 surfaced); 26 T-3(l); 33 Pt-l(l); 37 surface (4); 38 Pt-2 ( 1 ) ; 39 T-4 (12); 50 T-2C3) ; 50 T-3(13); 50 T-4(42); 59 T-3 (10 ) ; 59 T-4 ( 6 ) ; 59 T-5(5) ; 59 T-8(8) ; 59 T-9 ( 2 ) ; 59 T-10(10); 59 Pt-l(l); 61 Pt-l(l) : ; 61 T ’-3(11) ; 61 T-4 ( 9) ; 63 T-l (7) ; 66 T-2 0 24); 66 T-3 (12 ) ; 67 T-l (14) ; 68 T-3 (45) ; 68 T-5 (25) ; 68 Pt-1 (28) ; 74 T-l (7) ; 74 T- •3(2) ; 74 r r-4 < 8 0 ; 74 T-5 (9); 74 T-7 (3); ' 74 T-9 (8) ; 74 T-18 (5) ; ; 74 T-20 (6) ; 74 T-21 (10) ; 74 T-71(6) ; 74 T-74 (26 ); 74 : Pt-1 (3); 74 Pt-3 ( 15) ; ; 74 V-2 (14) ; 84 T-2 (11) : ; 84 T-3(l) ; 84 T-6 (5) ; 84 T-7 (3) ; 84 T-10 (7) ; 84 T-19 ( 8 ) ; 84 T-20 (19) ; 84 T-21 (3) ; 84 PL-1 (3) ; 84 Pt-4(1) ; 86 T-2 (50) ; 86 T-5 (15) ; 86 T-7 (2) ; 86 T-8 (7) ; 86 T-9 (4) ; 86 T-10 ( 2 ) ; 86 T-ll(10) ; 86 Pt-1 (5 ) ; 87 T-3 (24) ; 87 T-5 (17) ; 87 Pt-l(l). Templeton Crocker Expedition, Station 130 T-l (11) ; 134 T-3 ( 1 ) ; 148 T-2 ( 1) ; 148 T-3 ( 1 ) ; 148 T-6 (3) ; 158 T-2 (14) ; 158 T-3 ( 12 ) ; 158 T-4 (3) ; 159 T-l ( 8 ) ; 159 T-2 (2 ) ; 159 T-3 (68) ; 165 T-3 (5) ;. Eastern Pacific (Zaca) Expedi- tion, Station 185 T-l (7) ; 185 T-2 ( 4) ; 185 T-3 Cl); 210 T-l (99) ; 210 T-2 ( 16 ) ; 210 T-3 (7) ; 210 T-6 ( 62 ) ; 210 T-7 (79 ) ; 210 T-8 ( 12) ; 210 T-10(22) ; 219 T-l (51) ; 219 T-2 (17 ) ; 227 T-l(21); 230 T-l(27). For detailed trawling data, refer to Zoologica, Vol. VIII, pp. 6-23 {Arcturus); Vol. XXII,. pp. 37-46 (Templeton Crocker) ; Vol. XXIII, pp. 287-298 {Zaca). References and Synonymy. (Synonymv accepted according to Parr, 1928, p. 61). Myctophum laternatum : Garman, 1899, p. 267, pi. LVI, fig. 1. (7 specimens, including type; ? mm.; 200 to 1,168 fathoms; Gulf of California, 800 miles west of Costa-Rica, Gulf of Panama). Gilbert, 1913, p. 77. (4 specimens; ? to 20 mm.; 300 fathoms; Japan; partim Bolin’s Diogenichthys scofieldi). Weber, 1913, p. 86. (2 specimens; 11 mm.; ? to 821 fathoms; Banda Sea north of Australia ) Weber & Beaufort, 1913, p. 156. (De- scription). Regan, 1916, pp. 135, 139. (1 specimen; 8 mm. ; surface ; south Atlantic) . Taning, 1918, p. 150, fig. 46. (Descrip- tion of postlarval fish; 5 specimens; 5 to 20.3 mm.; ? to 875 fathoms; west of Tan- giers). Parr, 1928, pp. 61, 67. (Key to species; synonymy; discussion re M. laternatum at- lanticum; 2 small specimens; 5,000 feet wire; near Bermuda). Parr, 1931, p. 23, fig. 8. (2 specimens; ? mm.; 300 to 625 fathoms; cable; off west coast Mexico). Parr, 1934, p. 42, fig. 1. (Examination and description of type specimens). Norman, 1930, p. 324. (15 specimens; 11 to 27 mm.; ? to 1,368 fathoms; south At- lantic). Roule & Angel, 1930, p. 48. (1 specimen; 16 mm.; surfaace to 2,734 fathoms; be- tween Canary and Azores Islands). Beebe, 1929, p. 15. (1 specimen; ? mm.; ? fathoms; Hudson Gorge, northwest At- lantic) . Beebe, 1937, p. 205. (Preliminary list of 2,853 specimens; 9 to 22 mm.; 100 to 1,000 fathoms; Bermuda). M. { Myctophum ) laternatum: Brauer, 1904, p. 388 (Key to species). Brauer, 1906, p. 178, figs. 90-91. (63 specimens; ? to 22 mm.; ? fathoms; west coast Africa at Cape Verde, Sierra Leone, Gulf of Guinea, Indian Ocean, Sumatra, Bay of Bengal, Ceylon, Seychelles and Zanzibar, Gulf of Aden ) . Zugmayer, 1911, p. 23. (1 specimen; 20 mm.; surface to 984 fathoms; Atlantic Ocean near Gibraltar). Pappenheim, 1914, p. 193. (18 specimens; 11 to 19 mm.; 500 to 3,000 m. wire). M. laternatum atlanticum: Taning, 1928, p. 56. (Key to new sub- species ; 1 specimen ; ? mm. ; ? fathoms ; Gibraltar ) . Taning, 1932, p. 118, figs. 1-2. (Descrip- tion and synonymy). LeGendre, 1934, p. 337. (Key to species). Parr, 1934, p. 57. (Description). Diogenichthys laternatus : Bolin, 1939, pp. 92, 119, figs. 14, 15. (Key to species; detailed description; syn- nonymy; 15 specimens; 11.9 to 25.1 mm.; lower California and Japan). Myctophum pterotum (Alcock, 1890). Specimens taken by Eastern Pacific Ex- peditions: 6 specimens; Eastern Pacific {Zaca) Expedition; surface to 500 fathoms; between 7° 08' and 16° 05' N. Lat., and 81° 57' and 98° 17' W. Long.; lengths 20 to 43 mm.; taken between Nov. 30, 1937 and March 20, 1938. Specimens previously recorded: 180+ spe- cimens; ? to 64 mm.; ? to 3,000 metres wire; mid- and south Atlantic and Indian Oceans, Borneo, Philippines, Japan, eastern 76 Zoologica : New York Zoological Society [XXIX: 9 Pacific from tip of lower California to Panama. Text-fig. 8. Distribution of Myctophum pterotum. Photophore variation: Both Bolin’s and Parr’s keys give 6-4 as the number of anal photophores of this species. Two of our six fish present exceptions to this combination, one of 38 mm. with 7-4, and another 43 mm. in length 6-4, 5-4. Luminous glands: No luminous sex glands are visible on any of the specimens. Bolin (1939, p. 118) examined 28 specimens from the western Pacific measuring 30.8 to 53.1 mm. and states that “in most specimens, even the largest ones, no caudal luminous patches are developed.” Vertical distribution: Four specimens (34 to 43 mm.) taken at the surface after dark; two specimens (20 to 24 mm.) taken in a 500-fathom haul. Study Material. A total of 6 specimens was taken, as fol- lows: Eastern Pacific ( Zaca ) Expedition, Station 193 L-l (3) ; 210 L-l(l); 225 T-l(2). For detailed trawling data, refer to Zoologica , Vol. XXIII, pp. 287-298. References and Synonymy. (Synonymy accepted according to Bolin, 1939, p. 116). Scopelus ( Myctophum ) pterotus: Alcock, 1890, p. 217. (60 specimens, in- cluding type; ? mm.; 98 to 102 fathoms; Bay of Bengal ) . Alcock, 1896, p. 333. (Listing of pre- vious reference). Myctophum pterotus: Goode & Bean, 1895, p. 5. (Reference to 60 specimens of Alcock’s, listed above). Myctophum pterotum: Garman, 1899, p. 401. (Listing of Alcock’s material) . Fowler, 1928, p. 70. (Description; syn- onymy) . Norman, 1929, pp. 512, 513, fig. 2. (Syn- onymy ; reexamination of 8 specimens, in- cluding some of the types). Parr, 1928, pp. 60, 67. (Key to species; synonymy ; discussion of synonymy : M. pter- otum vs. M. fibulatum) . Parr, 1929, p. 7. (Reexamination of type specimens; synonymy). Parr, 1831, p. 22. (2 specimens; ? mm.; 525 to 1,800 fathoms; off west coast of Mexico). Weber, 1913, pp. 86, 670. (5 specimens; 20 to 38 mm. ; 6 fathoms ; Macassar Straights near Borneo). Weber & Beaufort, 1913, p. 157. (Descrip- tion and synonymy). Myctophum gilberti: Evermann & Seale, 1907, p. 55, fig. 1. (4 specimens, including type; 54 to 64 mm.; ? fathoms; Philippines). Gilbert, 1913, p. 81. (65 specimens; ? to 70 mm.; collected in market at Kagoshima, Japan) . Myctophum ( Myctophum ) pterotum: Brauer, 1906, p. 182, figs. 93, 94. (6 speci- mens; ? to 61 mm.; 821 to 1,368 fathoms; Indian Ocean). Barnard, 1925, p. 242. (Description). Pappenheim, 1914, p. 193. (10 specimens; 13 to 18 mm.; ? to 3,000 metres wire; mid- north Atlantic). Myctophum pterotum panamense : Tailing, 1932, p. 129, fig. 3. (? specimens; 28 to 36 mm.; ? fathoms; Gulf of Panama). Myctophum ( Benthosema ) pterotum: Bolin, 1939, pp. 92, 116, fig. 13. (Key to species ; synonymy ; description ; 28 speci- mens ; 30.8 to 53.1 mm.; ? fathoms; Japan, Philippines) . Myctophum rarum (Liitken, 1892). Specimens taken by the Eastern Pacific Expeditions: 1 specimen; Arcturus Oceano- graphic Expedition; 500 fathoms; 0° 17' S. Lat., 91° 34' W. Long.; length 28 mm.; June 9, 1925. Specimens previously recorded: 16 speci- mens; 21 to 40 mm.; ? to 1,640 fathoms; north and south Atlantic. Our specimen is the first to be recorded from the Pacific. Text-fig. 9. Distribution of Myctophum rarum. 1944] Beebe & Vander Pyl: Pacific Myctophidae 77 Photophore variation: The single speci- men corresponds in all details with the de- scription of the species, having only two SAG and with a greater space between the first and second anterior anal photophores than between the others. Luminous glands: This specimen, a male, has a very long and definite supracaudal gland, five plates running together with only faintly distinct divisions. Study Material. One specimen was taken, as follows: Arcturus Expedition, Station 84 Pt-3(1). For detailed trawling data, refer to Zoo- logica, Vol. VIII, pp. 6-23. References and Synonymy. (Synonymy accepted according to Parr, 1928,' p. 59). Scopelus (Rhinos pop elus) rarus: Liitken, 1892a, p. 246. (Key to species; description of type specimens ; 4 specimens ; 38 to 40 mm.; ? fathoms; north and south Atlantic) . ? Rhinoscopelus rarus: Goode & Bean, 1895, p. 91. (Description). Jordan & Evermann, 1896, p. 569. (De- scription). Myctophum varum forma integer: Brauer, 1904, p. 390. (Key to species; 1 specimen; 23 mm.; 1,640 fathoms; Gulf of Guinea) . Brauer, 1906, p. 204. (Further descrip- tion). Pappenheim, 1914, p. 197. (1 specimen; 31 mm.; 1,640 fathoms; south Atlantic). LeGendre, 1934, p. 337. (Key to species). ? Rhinoscopelus (Loweina) rarus: Fowler, 1925, No. 162, p. 2. (Listing only as new sub-genus “Loweina”). Myctophum rarum: (Not Brauer 1906, pp. 163, 404; not Bar- nard 1925, p. 245). Parr, 1928, p. 59. (Key to species and synonymy) . Taning, 1928, p. 55. (Key to species). Taning, 1932, p. 155. (Description and figure) . LeGendre, 1934, p. 337. (Key to species). Beebe, 1937, p. 205. (Preliminary list; 10 specimens; 21 to 26 mm.; 500 to 1,000 fathoms; Bermuda). Myctophum reinhardti (Liitken, 1892). Specimens taken by Eastern Pacific Ex- peditions: 131 specimens; Arcturus Oceano- graphic Expedition ; surface to 1,000 fath- oms; between 2° S. and 2° 12' N. Lat., and 88° 11' and 91° 53' W. Long.; lengths 10 to 17 mm. ; between April 14 and June 13, 1925. 24 specimens ; Templeton Crocker Expedi- tion ; surface to 400 fathoms ; between 18° 44' and 25° 25' N. Lat., and 108° 49' and 110° 33' W. Long.; lengths 13 to 51.5 mm.; between April 18 and May 7, 1936. 1 specimen; Eastern Pacific ( Zaca ) Ex- pedition; surface; 28° 42' N. Lat., and 117° 50' W. Long.; length 51 mm.; Nov. 8, 1937. Specimens previously recorded: 259 speci- mens; ? to 57 mm.; surface to 2,187 fath- oms ; north west Atlantic, Azores, west and east coasts of Africa, Indian Ocean, Japan, Australia, mid-Pacific and eastern Pacific from lower California south to mid-Chile. Text-fig. 10. Distribution of Myctophum reinhardti. Photophore variation: Anal photophore counts were made on 121 fish, the others being two rubbed or otherwise damaged. Photophores are distinct on fish as small as 8 mm. in length. The count reveals two dominant, numerical nodes, 5-6 and 6-7, comprising 52 fish and 43 per cent., and 36 fish or 30 per cent, of the whole, respec- tively. Extremes in numbers are 5-6 and 6-9, with only two fish showing the bilateral numerical asymmetry of 6-6, 6-5 and 6-6, 7-5. Luminous glands and sex proportions : All of the 131 specimens taken on the Arcturus Expedition are young. Of the Templeton Crocker Expedition catch of 24, only four showed developed, sexual, luminous glands; three males of 46, 51 and 51.5 mm. with supracaudal glands well developed, and one female of 40 mm. with two small, separate, infracaudal glands. Vertical distribution: The vertical distri- bution of the 156 specimens is as follows: surface (58 fish), 300 fathoms (11 fish), 400 (28), 500 (35), 600 (4), 800 (9) and 1,000 fathoms (11 fish). This proportion of surface captures of 58 fish or only 37 per cent, of the whole is un- usual, and would seem to indicate that this species lives at greater depths than other myctophids. Sociability : Schooling is indicated by the fact that 62 fish, or 40 per cent, of the total were taken in four nets, whereas the remain- ing 94 were distributed among 34 nets, of which a single fish occurred in each of 15 nets. 78 Zoologica : New York Zoological Society | XXIX : 9 Food: Our only notes on food is the record of copepods in eight fish, amphipods in three, and euphausids in three other fish. Study Material. A total of 156 specimens was taken, as follows: Arcturus Expedition, Station 41 T-l (7) ; 49 T-2 (1) ; 50 T-5(6) ; 52 T-l(2) ; 57 T-l (17) ; 78 T-l (4) ; 84 T-2(l); 84 T-8 (3) ; 84 T-10(3); 84 T-10(6) ; 84 T-14 (3) ; 84 Pt-3 (1) ; 84 T-19(l) ; 84 T-19 (1) ; 84 T-20 ( 1 ) ; 84 T-21(2) ; 86 T-2 (20) ; 86 T-2 (2); 86 T-4 (6); 86 T-5 (8); 86 T-7 ( 1 ) ; 86 T-9 ( 1 ) ; 86 T-10(l); 86 Pt-1 (2) ; 86 T-ll (3) • 87 T-2(7) ; 87 T-3 (12) ; 87 T-5(l). Templeton Crocker Expedition, Station 149 L-l (1); 158 T-2(l); 159 T-l (11 ; 159 T-2 (2); 138 L-l (13); 160 L-l (1); 162 L-l (5). Eastern Pacific ( Zaca ) Expedition, Station 177 L-l ( 1 1 . For detailed trawling data, refer to Zoologica, Vol. VIII, pp. 6-23 (Arcturus); Vol. XXII, pp. 37-46 (Templeton Crocker) ; Vol. XXIII, pp. 287- 298 {Zaca) . References and Synonymy. (Synonymy accepted according to Bolin, 1939,* p. 113). Scopelus reinhardti: Liitkin, 1892, p. 257, fig. 16. (Partim: 2 specimens, including type; ? to 43 mm.; ? fathoms, off Madeira and Liberia). M yctophum atratum : Garman, 1899, p. 268. (? specimens; ? mm.; 1,218 fathoms; southern part of Gulf of California). Myctophum b raueri : Gilbert, 1905, p. 598. (Description; 17 specimens including type of 57 mm.; no other lengths given ; surface ; Hawaiian Islands). Myctophum lutkeni: Gilbert, 1905, pi. 70, fig. 1. (lapsus calami pro Myctophum braueri) . Myctophum reinhardti: Goode & Bean, 1895, p. 74. (Description). Waite, 1904, pp. 154, 194. (2 specimens; 35 and 45 mm.; ? fathoms; Australia). Gilbert, 1908, p. 219. (6 specimens; young; ? fathoms; 300 miles east of Mar- quesas ) . Gilbert, 1913, p. 80. (3 specimens; young; surface; off southwest coast of Japan). Jordan & Jordan, 1925, p. 11. (References and synonymy). Taning, 1928, p. 57. (Key to species). Parr, 1928, p. 66. (Key to species, syno- nymy). Fowler, 1928, p. 70. (Synonymy). Beebe, 1929, p. 16. (4 specimens; ? mm.; surface to 800 fathoms; Hudson Gorge, northwest Atlantic). Beebe, 1937, p. 205. (Preliminary list; 72 specimens; 11 to 35 mm.; 100 to 1,000 fath- oms; Bermuda). Myctophum, benoiti reinhardti: Brauer, 1904, p. 388. (Key to species). Brauer, 1906, p. 185. (100 specimens; ? to 25 mm.; 670 to 2,187 fathoms; Gulf of Guinea, southwest Africa, Bay of Bengal, Seychelles Islands, Gulf of Aden, Ceylon, Indian Ocean, west coast of Chile). Weber, 1913, p. 86. (5 specimens; 9 to 16 mm.; surface to 825 fathoms; Banda Sea, Almahera Sea, East Indies). Weber & Beaufort, 1913, pp. 152, 155. (Key to species ; synonymy and description) . Pappenheim, 1914, p. 193. (16 specimens; 15 to 32 mm.; ? to 1,642 fathoms; south Atlantic) . Barnard, 1925, p. 242. (References and brief description). Myctophum ( Hygophum) reinhardti: Bolin, 1939, pp. 93, 113-116, fig. 12. (Key; synonymy; description; 31 specimens, of which 7 were paratypes of M. braueri, Gil- bert ; 17.7 to 31.2 mm. ; ? fathoms ; Hawaiian Islands and Philippines). Myctophum valdiviae Brauer, 1904. Specimens taken by Eastern Pacific Ex- peditions: 35 specimens; Arcturus Oceano- graphic Expedition; 300 to 700 fathoms; between 6° 27' N. Lat., and 0° 17' S. Lat., and 91° 53' and 84° 35' W. Long.; lengths 8 to 23 mm.; between May 1 and June 13, 1925. Specimens previously recorded: 924 speci- mens; ? to 23 mm.; surface to 2,260 fath- oms; north and south Atlantic Ocean, west and east coasts of Africa, Indian Ocean, Japan and eastern Pacific off Cocos and Galapagos Islands. Text-fig. 11. Distribution of Myctophum valdiviae. Photophore variation: All 35 specimens were in rather bad condition, with nearly all the anal photophores rubbed or torn away. The species is easily identified with- out them, it being the only Myctophum which has the four upper photophores above the lateral line. Because of the generally poor condition of the anal photophores, 1944] Beebe & Vander Pyl: Pacific Myctophidae 79 nothing of interest could be determined in regard to their variation. Luminous glands: The 35 fish measured from 8 to 23 mm. Of these 20 showed no signs of supracaudal luminous glands, and (with the exception of one fish of 16 mm.) were all 12 mm. or under in length. The dis- tinction between male and female glands is not satisfactory enough to class the remain- ing 16 fish as anything but nearly or not quite adult. Breeding: Fish No. 5569, a female of 22 mm., was taken 100 miles south of Cocos Island, on May 1. The ovaries were 3.5 mm. in length and the diameter of the eggs .28 mm. The total number of eggs about to be laid was 120. Vertical distribution : All 35 fish were taken at depths from 300 to 700 fathoms; none at the surface. The relative numbers were 300 fathoms (4 fish) ; 400 (11) ; 500 (14); 600 (5); and 700 fathoms (1 fish). Sociability : With only 35 specimens at hand it is difficult to hazard any opinion as to schooling, especially as 8 fish were taken in one net, and 8 others each in a single net. Food: Copepods and amphipods were found in several stomachs. Study Material. A total of 35 specimens was taken, as follows: Arcturus Expedition, Station 59 T-4(l) ; 59 T-5 ( 2 ) ; 59 T-9(l) ; 59 Pt-2(3) ; 61 T-3 (3) ; 68 T-3(2) ; 74 T-8(l) ; 74 T-71 (2) ; 74 T-74 (2 ) ; 74 T-78(l) ; 84 T-l(l) ; 84 T-8 ( 1 ) ; 84 T-20 ( 1 ) ; 86 T-2 ( 5 ) ; 86 T-8(l); 87 T-2 (8). For detailed trawling data, refer to Zoologica, Vol. VIII, pp. 6-23. References and Synonymy. (References accepted according to Parr, 1928, p. 57). Myctophum valdiviae: Brauer, 1904, p. 398, fig. 6. (Preliminary listing of .88 specimens, including type; de- tails in following reference). Brauer, 1906, p. 206, fig. 127. (68 speci- mens; 8 to 23 mm.; 274 to 2,260 fathoms; Atlantic Ocean, south of Canary Islands, Gulf of Guinea, 500 miles west of southwest Africa, Indian Ocean, between Aden and Madagascar, mid-ocean between Ceylon and Zanzibar, Cocos Island, Bay of Bengal). Gilbert, 1913, p. 84. (1 badly injured specimen ; ? mm. ; ? to 300 fathoms ; Japan ) . Taning, 1918, p. 151, fig. 47. (Description of post-larvae; 4 specimens; 1 adult and 3 post-larvae; 4.7 to 15 mm.; ? fathoms; mid- ocean northwest and southwest of Ber- muda) . Taning, 1928, p. 57. (Key to species). Taning, 1932, p. 121, figs. 1, 2. (Descrip- tion). Parr, 1928, p. 57. (Key to species; syn- onymy). LeGendre, 1934, p. 335. (Key to species). Beebe, 1929, p. 15. (4 specimens; ? mm.; ? fathoms; Hudson Gorge, northwest Atlan- tic Ocean). Beebe, 1937, p. 205. (Preliminary list; 846 specimens; 9 to 23 mm.; 300 to 1,000 fathoms; Bermuda). Myctophum ( Myctophum ) valdiviae: Pappenheim, 1914, p. 194. (16 specimens; 9 to 24 mm. ; 800 to 3,000 metres ; north and south Atlantic) . Lampanyctus elongarus (Costa, 1844). Specimens taken by Eastern Pacific Ex- peditions: 3 specimens; Arcturus Oceano- graphic Expedition; surface to 700 fath- oms; between 2° 00' S. Lat. and 4° 50' N. Lat., and between 87° 00' and 89° 48' W. Long., lengths, 25, 72.5 and 83 mm. ; taken April 3 and 23, May 31, 1925. Specimens previously recorded: 706 speci- mens ; ? to 153 mm.; surface to 2,600 fath- oms; western Atlantic from Greenland south _ to Cape Horn, eastern Atlantic from mid- Norway and Mediterranean south to Cape of Good Hope, Australia, Japan, eastern Pacific from San Diego to Galapagos. Our specimens extend the range to the Galapagos and Cocos Islands. A widely distributed but rare species. Text-fig. 12. Distribution of Lampanyctus elongatus. General data: Our three specimens show the following details: Cat. No. 5214a, sex uncertain, 25 mm. in length, anal photo- phores 9-6, 9-5. Cat. No. 5429, a female, 83 mm., anal photophores 8-?, 8-?, taken April 23, 1925, ovaries injured but many eggs visible; not countable but almost ready to be deposited. Cat. No. 6053, female, 72.5 mm. ; anal photophores 8-6, 9-6. Two small, infracaudal luminous glands just appearing. Eggs undeveloped. Anal photophore counts of other authors are 7-9, 6-7 ; 7-9, 5-7 ; and 9-6, 6-7. General color: Cornea over pupil reflect- ing iridescent green in all illuminations; iris dark brown except for the infero- posterior third which is brilliant coppery gilt; body in general brownish-black; glints of blue and green on opercles and body, but 80 Zoologica: New York Zoological Society [XXIX: 9 scales are colorless. All large, round photo- phores are pink, smaller ventral ones sil- very. All light given off is decidely pink. Food: The 83 mm. female had eaten two amphipods and one euphausid. Study Material. A total of 3 specimens was taken, as fol- lows: Arcturus Expedition, Station 33 Pt-1 (1); 52 T-l(l); 74 T-35 ( 1 ) . For detailed trawling data, refer to Zoologica, Vol. VIII, pp. 6-23. References and Synonymy. ( Synonvmy accepted according to Bolin, 1939, p. 151). Scopelus elongatus: Costa, 1844, p. 2, pi. 38. (Original descrip- tion; reference not seen; Naples, fide; Jor- dan & Evermann, 1896). Canestrini, 1870, p. ? (Reference not seen ) . Steindachner, 1881, p. 397. (Description and discussion of synonymy of ? specimens in Vienna Museum; 135 to 150 mm.; ? depth, ? locality) . Vinciguerra, 1885, p. 462. (Description and synonymy). Raffaele, 1889, p. 184, fig. 11. (Discussion of synonymy) . Lilljeborg, (1884-91), p. ? ( Reference not seen) . Liitken, 1892a, p. 264. ( Description and key to species; 4 specimens; about 80 to 84 mm. ; ? fathoms ; Greenland, Gulf of Guinea and south of Cape of Good Hope). Koehler, 1896, p. 506. (1 specimen; 62 mm.; 760 fathoms; off Lorient, France). Holt & Byrne, 1905, p. 23. (1 specimen; ? mm.; 350 fathoms; northeast coast of Ireland) . Holt & Byrne, 1911, p. 24, fig. 7. (Descrip- tion and synonymy). Scopelus crocodilus : Cuvier & Valenciennes, 1849, vol. 20, p. 332. (Reference). Cuvier & Valenciennes, 1849, vol. 22, p. 447. (Description). Scopelus pseudocrocodilus : Moreau, 1891, p. 84, fig. 227. (Descrip- tion). Brauer, 1906, p. 233. (Synonymized with M. ( Lampanyctus ) elongatum) . Scopelus kvoyeri: Malm, 1863, p. 100. (Reference not seen). Brauer, 1906, p. 232, figs. 152, 153. (Syn- onymizes this with M. ( Lampamjctus ) elon- gatum) . Winther, 1879-80, p. 42. (Brief refer- ence). Collet, 1880, p. 162. (1 specimen, found in stomach of gadus morrha) . Scopelus resplendens : Gunther, 1864, vol. 22, p. 415. (Descrip- tion and snyonymy). Scopelus ( Notoscopelus ) elongatus: Gunther, 1864, vol. 22, p. 415. (Descrip- tion and synonymy). Notoscopellus brachychier : (error in type). Eigenmann & Eigenmann, 1889, p. 126. (3 specimens; ? mm.; 45 fathoms; Cortez Bay, San Diego). Notoscopelus brachychier : Eigenmann & Eigenmann, 1890, p. 23. (Reference). Notoscopelus resplendens: Goode & Bean, 1895, p. 83, fig. 94. (De- scription and key). Notoscopelus elongatus : Goode & Bean, 1895, p. 83. (Description and key). Bolin, 1939, pp. 93, 151, 152, fig. 29. (Key; synonymy and description; 2 speci- mens; 110 and 113.5 mm.; ? fathoms; Misaki Sea, Japan). Notoscopelus quercinus : Goode & Bean, 1895, p. 83, fig. 97. (3 specimens; ? to 125 mm.; ? to 781 fathoms; north Atlantic, off Cape Cod, Massachusetts, and coast of Virginia). Notoscopelus ejectus: Waite, 1904, p. 150, pi. 18, fig. 2. (1 specimen; 40.5 mm.; ? fathoms; Australia). Catablemmella brachychir : Eigenmann & Eigenmann, 1890, p. 23. (Reference ) . Scopelus ( Lampanyctus ) elongatus : Liitken, 1892, p. 233. (Reference to speci- mens described under Scopelus elongatus) . Carus, 1889-1893, p. 565. (Description). Macrostoma angustidens: Jordan & Evermann, 1896, p. 555. (De- scription ) . Macrostoma quercinum: Jordan & Evermann, 1896, p. 554. (De- scription). Macrostoma brachychir: Jordan & Evermann, 1898, p. 2826. (Name listed) . Macrostoma quercinum japonicum: Tanaka, 1908, p. 5, pi. 1, fig. 3. (Descrip- tion). Macrostoma japonicum: Tanaka, 1911, p. 59, pi. 15, fig. 51. (Re- description of type specimen reported in previous reference). 1944J Beebe & Vander Pyl: Pacific Myctophidae 81 Myctophum elongatum: Collet, 1884, p. 104. (Description; discus- sion of synonymy; 5 specimens; 128 to 145 mm.; ? fathoms; coast of Norway, Trond- heim and vicinity). Smitt, 1895, p. 937, fig. 234. (Detailed description) . Sanzo, 1918a, p. 4. (Description of young). Myctophum ( Lampanyctus ) elongatum: Brauer, 1906, p. 232, figs. 152, 153. (Key; description; synonymy; 7 specimens; 15 to 125 mm.; 380 to 2,187 fathoms; Gulf of Guinea, west of Cape Colony). Fage, 1910, p. 15, fig. 13. (Description). Zugmayer, 1911, p. 37. (1 specimen; 57 mm.; 2,600 fathoms; off west coast of Por- tugal). Pappenheim, 1914, p. 196. (6 specimens; 22 to 29 mm.; ? to 1,642 fathoms; south Atlantic) . Alaejos, 1919. p. ? (Reference not seen). Barnard, 1925, vol. 21, p. 238. (Descrip- tion). Myctophum ( Lampanyctus ) quercinum: Brauer, 1906, p. 166. (Key to species). Lampanyctus japonicus : Gilbert, 1913, p. 99. (Brief discussion). Lampanyctus resplendens: Richardson, 1844, p. 42, pi. 27, figs. 16-18. (1 specimen; 112 mm.; ? fathoms; ? local- ity). Matsubara, 1938, p. 53, fig. 10. (Descrip- tion; 3 specimens; 123 to 153 mm.; about 150 fathoms; Japan). Lampanyctus elongatus : Gilbert, 1913, p. 99. (Reference; discus- sion of L. japonicus) . Taning, 1918, pp. 103-106, figs. 40, 41. (Detailed description of both young and adults; distribution map; propagation etc.; 651 specimens; ? to 60 mm. surface to 670 fathoms; western Meditteranean and At- lantic Ocean). Tanning, 1928, p. 64. (Key to species). Taning, 1932, p. 124, figs. 1, 2. (Descrip- tion) . Parr, 1928, p. 80. (Key to species; syn- onymy). Parr, 1929, p. 14. (Additional synonymy). Norman, 1930, p. 329. (4 specimens; 30 to 53 mm.; surface to 547 fathoms; mid and south Atlantic). LeGendre, 1934, p. 343. (Key to species). Beebe, 1937, p. 204. (Preliminary list; 11 specimens; 13 to 30 mm.; 400 to 1,000 fath- oms; Bermuda). Lampanyctus idostigma Parr, 1931. Specimens taken by Eastern Pacific Ex- peditions: 12 specimens; Eastern Pacific ( Zaca ) Expedition; 300 to 500 fathoms; be- tween 8° 03' and 9° 12' N. Lat., and 83° 12 and 85° 10' W. Long.; lengths 16 to 37 mm.; between Feb. 7 and March 10, 1938. Specimens previously recorded: 27 speci- mens; 50 to 76 mm.; 100 to 200 fathoms; Eastern Pacific from tip of Lower California to Nicaragua. Our collection extends the range to southern Costa Rica. Text-fig. 13. Distribution of Lampanyctus idostigma Anal photophores: Our 12 specimens, ranging from 16 to 37 mm.; are all too im- mature to be sexed. The photophore arrange- ment is as follows: (5 specimens) 6-6; (4) 5- 6; (2) 5-6, 6-6; and (1 specimen) 6- 6, 6-7. General data: Three fish were taken at 300 fathoms and nine at 500 fathoms. There is no hint of schooling as the 12 fish were scattered among eight nets. All the evidence points to this species living in deep water. Our fish were all taken at two localities off Costa Rica, 150 miles apart. Study Material. A total of 12 specimens was taken, as fol- lows: Eastern Pacific {Zaca) Expedition, Station 210 T-l(2); 210 T-3 ( 1 ) ; 210 T-6 (1) ; 210 T-7 ( 1 ) ; 210 T-8 (3) ; 210 T-10(l) ; 219 T-l(l) ; 219 T-2(2). For detailed trawl- ing data, refer to Zoologica, Vol. XXIII, pp. 287-298. References and Synonymy. Lampanyctus idostigma: Parr, 1931, p. 32, fig. 13. (27 specimens, including type; 50 to 76 mm.; 300 to 625 fathoms cable; Pacific coast of Nicargua, off southern Mexico, Gulf of California). Bolin, 1939, pp. 93, 138, fig. 22. (Key to species ; redescription of 2 of Parr’s speci- mens). Lampanyctus longipes (Brauer, 1906). Specimens taken by Eastern Pacific Ex- peditions: 1 specimen ; Arcturus Oceano- graphic Expedition; 500 fathoms; at 0° 40' N. Lat., and 91° 47' W. Long., length 21 mm. ; taken June 12, 1925. 82 Zoologica : New York Zoological Society [XXIX: 9 Specimen^ previously recorded: 18 speci- mens; surface to 1,230 fathoms; 11 to 32 mm.; rare but widely distributed. Bermuda, west coast of Africa, Indian Ocean, Bay of Bengal. Our record is the first for the Pa- cific, and was off the Galapagos. Text-fig. 14. Distribution of Lampanyctus longipes. Our fish, 21 mm., was too young for sexing; its anal photophore count was 5-3, 5-3. Study Material. One specimen was taken, as follows: Arcturus Expedition, Station 86 T-2(l). For detailed trawling data, refer to Zoo- logica, Vol. VIII, pp. 6-23. References and Synonymy. (Synonymy accepted according to Parr, 1928, p. 82). M. ( Lampanyctus ) longipes: Brauer, 1906, p. 236, fig. 155. (Original description; 14 specimens, including types; 11 to 32 mm.; surface to 1,230 fathoms; Madeira, Gulf of Guinea, Seychelles Is- lands, Bay of Bengal, south of Ceylon, Indian Ocean). Lampanyctus pyrsobolus longipes: Tailing, 1928, p. 65. (Key to species). LeGendre, 1934, p. 344. (Key to species). Lampanyctus longipes: Parr, 1928, p. 82. (Key to species; syn- onymy ) . Beebe, 1937, p. 204. (Preliminary list; 4 specimens; 12 to 20 mm.; 800 to 1,000 fathoms; Bermuda). Lampanyctus macdonaldi ( Goode & Bean, 1895 ) . Specimens taken by Eastern Pacific Ex- peditions: 3 specimens; Arcturus Oceano- graphic Expedition; 400 fathoms; between 0° 00' equator and 91° 53' W. Long.; lengths 20 to 29 mm.; June 13, 1925. Specimens previously recorded: ? speci- mens; lengths ? to 154 mm.; ? to 1,467 fathoms; northwestern Atlantic. Our speci- mens are the first recorded from the Pacific Ocean. Text-fig. 15. Distribution of Lampanyctus macdonaldi. Photophore variation: Our three speci- mens, (Cat. No. 6365a), taken 21 miles north of Narborough, Galapagos, show the following anal photophore count: 1 speci- men 20 mm. 6-6-4 on both sides ; 1 specimen 27 mm. 6-6-4 and 6-7-4; 1 specimen 29 mm. 6-6-4 on both sides. Study Material. A total of 3 specimens was taken, as fol- lows: Arcturus Expedition, Station 87 T-2 (3). For detailed trawling data, refer to Zoologica, Vol. VIII, pp. 6-23. References and Synonymy. ( Synonymy accepted according to Parr, 1929, p. 20). N annobrachium macdonaldi: Goode & Bean, 1895, p. 94, fig. 110, pi. 29. (Original description; ? specimens, includ- ing type; ? to 154 mm.; ? to 1,467 fathoms; northwestern Atlantic). Jordan & Evermann, 1896, vol. 1, p. 563. (Description ) . Taning, 1928, p. 69. (Reference to syn- onymy ) . Lampanyctus macdonaldi : Parr, 1929, p. 20, fig. 9. (Redescription of type specimen and synonymy). Lampanyctus macropterus (Brauer, 1904). Specimens taken by Eastern Pacific Ex- peditions: 150 specimens; Arcturus Oceano- graphic Expedition ; surface to 1,000 fath- oms; between 2° 00' S. and 4° 50' N. Lat., and 87° 00' and 91° 53' W. Long.; lengths 16 to 70 mm.; taken between April 22 and June 13, 1925. Specimens previously recorded: 29 speci- mens; ? to 105 mm.; surface to 1,642 fath- oms; West Indies, south Atlantic, north Indian Ocean, Bay of Bengal, Japan. Our specimens, all from the Galapagos, are the first recorded from the eastern Pacific. Photophore variation: Photophore count of 137 fish shows them in general to be sym- metrical, 30 per cent, or 42 specimens hav- 1944] Beebe & Vander Pyl: Pacific Myctopliidae 83 Text-fig. 16. Distribution of Lampanyctus macropterus. ing 5-9 anal photophores; 27 per cent, or 37 specimens having 5-8 anal photophores, and the balance are divided into the follow- ing combinations: 4-8, 4-10, 4-11, 5-7, 5-10, 5-11, 6-8, 6-9, and the asymmetrical combinations of 5-7, 4-8; 5-8, 5-7; and 5-9, 5-7. Development: All of the 150 specimens were measured with this result: 84 fish or 56 per cent, were 16 to 36 mm. and 66 fish or 44 per cent, from 37 to 70 mm. in length. Eggs: Three females (Cat. No. 6057), measuring 53, 55 and 59 mm. had eggs pres- ent, while another female, measuring 50 mm. (same Cat. No.) was without developed eggs. These were taken at Station 74, T-41, 42, 60 miles south of Cocos Island on June 1, 1925. Vertical distribution: Of the entire catch, 46 fish or 33 per cent, were taken at the surface. Of the remainder, 70 per cent, were brought up from 400 and 500 fathoms. An unusual fact is that 11 of the 46 surface fish were taken in full sunlight, in a haul made from 9.30 to 11.00 A.M. These 11 were all young (18 to 24 mm.), while the re- maining surface fish taken after dark were much larger (35 to 58 mm.) more nearly adult. Sociability: Schooling is definitely indi- cated. Seventy-five fish or 50 per cent, were taken in five nets, an average of 15 fish to a net, while the rest were distributed among 24 nets, averaging two and a half fish to each. Food and Anatomy: On May 24, 1925, sixty miles south of Cocos Island, at 8.00 to 8.30 P.M. from a rainy and a rough sea, I drew a surface net containing among other fish 13 Lampanyctus macropterus. These were large (50 to 70 mm.) and very vigorous and active. A male of 65 mm. ap- peared to have gorged itself, and I dissected it for food. The swim bladder was fully in- flated, silvery white and 20 mm. long. The anterior end was rounded and 5.5 mm. in diameter. This thickness was sustained pos- teriorly for 12 mm. when the organ tapered rapidly for the posterior 8 mm. to end in a fine point. The lining of the body cavity was slightly pigmened, while the stomach, unlike any of the other organs, was jet black. The fish had already fed well although it was so early in the evening. The contents comprised 9 fish eggs, 1 mollusk egg-mass, 1 small sagitta, 5 young euphausids, 3 hy- perid amphipods, 6 small calanids, and 1 Eucalanus elongatus. Luminescence and viability: The follow- ing notes are from my Journal made on May 25, 1925, upon fish from a surface haul made sixty miles south of Cocos Island, m the darkroom of the Arcturus. There were a number of Noctiluca and Sapphirina in the water but we had little trouble in differentiating between their light and that of the myctophids. We tested the fish in artificial sea water but if anything their light was exhausted sooner than in the warmer, normal salt water. There was no reaction to three drops of ammonia, nor to a slight shock of electricity sent through the water. When sent through the fish they died immediately. One observation made again and again during the evening was of an obliterative flashing, repeated so identically that there could be no question as to its repeated oc- currence, whatever may be the precise in- terpretation. The fish was lighted up dully and in the indirect manner for which as yet I have no explanation. It swam slowly about and when it encountered another fish o'- my finger, it turned with a quick flick of the tail, emitting a very brilliant flash from its two precaudal photophores, and then vanished completely. This quenching was achieved in one of two diverse ways. If the fish was swimming upside down, as it often does, it turned over after the caudal flash and exposed the unlighted dorsal area, or it sometimes actually turned out every light, from head to tail. The actions of various individuals dif- fered widely. One which was brought in some time after our first specimens had be- gun to die down, swam continually on its back and glowed dully from every photo- phore. Rarely it blackened, and twice I saw sets of four and of six photophores respec- tively, lit as individual units. The brightest light on the entire body was the most dorsal of the pre-pectoral set. When first caught, even when dipped gently in a net from the companionway and dropped at once into a bucket of water, the immediate instinct of the fish is to descend. It swims in small circles, making constant nose dives and bumping its head against the bottom. Two other Lampanyctus lighted up as a whole, but the general effect was always of separate lights, the photophores not being 84 Zoologica : New Y ork Zoological Society [XXIX: 9 so near together as in M. coccoi and the silver scales being absent. It is curious that although coccoi is so intensely active, its scales are far less deciduous than those of other species. Holding L. macropterus by the head resulted in intense general illumin- ation, and even in the air the light lasted a considerable time before being douched. When it went out, the head and tail lights were sustained alone. When a male was held by the head, while the entire under surface was aglow, the dorsal aspect was entirely black except for the median, dorsal, caudal plates. A few days later several large specimens of this species were watched. In the dark- room they glowed dully, never as brilliantly in general as M. coccoi. Their lunate light organs seemed to diminish the glow. They swam upside down, and suddenly one made a rush, seized the other by the jaw and be- gan pushing him about. Both seemed to have the same grip, and the pushing on the part of both kept up until death. The whole affair was of course accidental. Sometime after apparent death, the jaw lights glowed faint- ly. During life these six lights were the dominant ones, with the exception of the succeeding pair on the isthmus, which oc- casionally shone with a very powerful glare. When the entire fish was aglow, a hand lens showed every ventral light shining full strength, but to the unaided eye the illum- ination was indefinite, indirect. In daylight the photophores are bright purple. Study Material. A total of 150 specimens was taken, as follows: Arcturus Expedition, Station 50 T-5 (2 ) ; 74 T-l(24); 74 Pt-1 (3) ; 74 Pt-2 (1) ; 74 T-7 ( 1 ) ; 74 T-10 (4) ; 74 T-ll (3) ; 74 T-15 (2 ) ; 74 T-16 (1) ; 74 T-21 (1) ; 74 T-22 ( 14 ) ; 74 Ot-2(l) ; 74 Ot-3 ( 1 ) ; 74 T-27 (1) ; 74 T-35 (1 ) ; 74 T-35(l) ; 74 T-41(3) ; 74 T-43 (7) ; 84 T-2(24); 84 T-10 (1) ; 84 T-10 (2) ; 84 T-10 (5) ; 84 T-14C10) ; 84 Pt-3 (2) ; 84 Pt-3 (14) ; 86 T-2(2); 86 T-2 ( 1 ) ; 86 T-4 (2 ) ; 86 T-5 (3) ; 86 T-ll (3); 87 T-2 (6) ; 87 T-6 ( 4 ) . For detailed trawling data, refer to Zoologica, Vol. VIII, pp. 6-23. References and Synonymy. (Synonymy accepted according to Parr, 1931, p. 28). Myctophum ( Lampanyctus ) macropterum: Brauer, 1904, pp. 381, 404, fig. 5. (Key and description of type specimen). Brauer, 1906, p. 249, figs. 166, 167. (22 specimens; ? to 65 mm.; 547 to 1,368 fath- oms; Indian Ocean, Bay of Bengal, Ceylon, Cocos, Seychelles Islands, Gulf of Aden, North Africa, ( partim ) and fig. 166, nec. fig. 167). Pappenheim, 1914, p. 196. (1 specimen; 63 mm. ; ? to 1,642 fathoms; south Atlantic) . Lampanyctus macropterus : Gilbert, 1913, p. 106. (1 specimen; 105 mm.; 703 (?) fathoms; Kagoshima, Japan). Regan, 1916, p. 140. (1 specimen; 10 mm. ; 2 fathoms; New Zealand). Parr, 1928, pp. 88, 110, fig. 20. (Key to species; synonymy; description; 4 speci- mens; ? mm.; 7-8,000 feet wire; West In- dies). Parr, 1931, pp. 25, 28, fig. 10. (nec. Parr, 1928, p. 110 ; key to species ; discussion ; syn- onymy). Lampanyctus mexicanus (Gilbert, 1890). Specimens taken by Eastern Pacific Ex- peditions: 36 specimens; Arcturus Oceano- graphic Expedition; 255 to 833 fathoms; be- tween 0° 17' S. and 6° 24' N. Lat., and be- tween 81° 18' and 91° 34' W. Long.; lengths 13.5 to 64 mm.; March 30 to June 9, 1925. 177 specimens: Templeton Crocker Ex- pedition; surface to 500 fathoms; between 23° 25 and 30° 00' N. Lat., and 108° 31' and 116° 27' W. Long.; lengths 19 to 63 mm.; March 28 to May 23, 1936. 4 specimens; Eastern Pacific ( Zaca ) Ex- pedition; 500 fathoms; between 5° 10' and 9° 12' N. Lat., and 78° 42' and 85° 08' 30" W. Long. ; lengths 18 to 52 mm. ; February 7 to March 26, 1938. Specimens previously recorded: 143+ specimens; ? to 64.7 mm.; surface to 1,218 fathoms; eastern Pacific, from California at Point Conception south to Ecuador. Text-fig. 17. Distribution of Lampanyctus mexicanus. Photophore variation: Ten of the 36 fish from the Arcturus had asymmetrical counts; 15 were symmetrical, 5-9; and the remaining 11 were about equally divided among the symmetrical combinations of 4-8, 4-9, 4-10, 5-8, 5-10. The count on 25 fish from the Templeton Crocker group showed that 13 specimens, or 50 per cent., had the 4-9 pattern, with the others varying from 4-8 to 5-9. Three individuals have asymmetrical patterns on the two sides of the body. Vertical distribution: Lampanyctus mexi- canus is a deep water species as the vertical distribution of our catch well shows. 106 1944] Beebe & Vander Pyl: Pacific Myctophidae 85 fish (59 per cent.) were taken at 300 fath- oms; 53 (30 per cent.) at 400; and 21 fish (11 per cent.) from 500 fathoms. Only a single individual was captured in a surface haul. Sociability : Schooling was more evident than in any other species encountered. About 60 per cent, of the entire catch, or 105 fish, was taken in a single net, but this is only part of the truth. The same school evidently extended throughout 200 vertical fathoms, for on April 17, three nets on the same wire, drawn from 3.45 to 5.00 P.M., at 300, 400 and 500 fathoms contained 105, 25 and 19 fish respectively, or four-fifths of all of this species taken on all three expeditions. The remaining fish were distributed as fol- lows: one net (13 fish), one net (TO fish), and nine nets (1 fish each). Study Material. A total of 217 specimens was taken, as follows : Arcturus Expedition, Station 26 Pt-l(l); 39 T-4(l); 59 Pt-2(1); 66 Pt-1 (1) ; 74 Pt-l(l) ; 74 Pt-3(3) ; 74 Ot-3(l) ; 74 Ot-4 ( 1 ) ; 84 T-2 ( 5 ) ; 84 Pt-1 (7) ; 84 T-8 (4); 84 T-14(8) ; 84 Pt-3 ( 2) . Templeton Crocker Expedition, Station 130 T-l ( 13) ; 139 T-2 and T-3 and T-4(10) ; 148 T-2 (105) ; 148 T-3 (25) ; 148 T-4(19) ; 148 T-8 (1) ; 158 T-3(l) ; 158 T-4(l) ; 159 T-l(l) ; 174 L-l(l). Eastern Pacific ( Zaca ) Expedi- tion, Station 210 T-6(l) ; 210 T-10(l) ; 227 T-l (1) ; 230 T-l(l). For detailed trawling data, refer to Zoologica, Vol. VIII, pp. 6-23 ( Arcturus ) ; Vol. XXII, pp. 37-46 (Temple- ton Crocker); Vol. XXIII, pp. 287-298 (Zaca) . References and Synonymy. (Synonymy accepted according to Bolin, 1939, p. 135). Myctophum mexicanum: Gilbert, 1890, p. 51. (6 specimens, includ- ing type; ? to 50 mm.; 306 to 857 fathoms; Gulf of California). Scopelus mexicanus: Lutken, 1892, p. 266. (Name listed). N annobrachium mexicanum : Goode & Bean, 1895, p. 512. (Name listed) . Myctophum oculeum: Garman, 1899, p. 260, pi. 56, fig. 2. (? specimens; ? mm.; surface to 1,218 fath- oms; off Pacific coast lower California, south ! to Ecuador) . Myctophum ( Lampanyctus ) mexicanum: Brauer, 1904, p. 396. (Key to species). Myctophum . ( Lampanyctus ) oculeum: Brauer, 1906, p. 167. (Key to species). Lampanyctus oculeus : Parr, 1928, p. 85. (Key and synonymy). Parr, 1934, p. 46. (Synonymizes with L, mexicanus) . Lampanyctus mexicanus: Parr, 1928, p. 84. (Key and synonymy). Parr, 1929, p. 15, fig. 6. (Key and further description of type). Parr, 1931, p. 30, fig. 12. (23 specimens; ? to 60 mm. ; ? to 286 fathoms ; eastern Pacific coast of Mexico). Bolin, 1939, pp. 93, 135, fig. 21. (Key; synonymy; description; 114 specimens; 15.5 to 64.7 mm.; ? fathoms; from Point Concep- tion, California south to Gulf of California. These include 2 co-types and 4 of Garman’s M. oculeum). Lampanyctus omostigma Gilbert, 1908. Specimens taken by Eastern Pacific Ex- peditions: 246 specimens; Arcturus Oceano- graphic Expedition: surface to 1,200 fath- oms; between 2° 00' S. and 6° 27' N. Lat., and 83° 33' and 91° 53' W. Long.; lengths 12 to 68 mm.; between April 22 and June 13, 1925. 94 specimens; Eastern Pacific (Zaca) Ex- pedition: 300 to 500 fathoms; between 5° 10' and 17° 45' N. Lat., and 78° 42'71" and 103° 05' W. Long. ; lengths 13 to 96 mm. ; between Nov. 23, 1937 and March 26, 1938. Specimens previously recorded: 30 speci- mens ; ? to 62 to ? mm. ; surface to 200 fathoms; mid-Pacific near Marquesas Is- lands and eastern Pacific from Mexico to Galapagos. Our collection establishes the species as occurring nearly 800 miles south of the earlier known range. After careful examination of our speci- mens we fail to find sufficient consistent var- iation to differentiate these eastern Pacific fish as Lampanyctus omostigma parvicauda (Parr, 1931, p. 26, fig. 9). Text-fig. 18. Distribution of Lampanyctus omostigma. Photophore variation: The range in num- ber of anal photophores is 4-8 and 6-10, with the dominant nodes at 5-8 and 5-9, appear- ing on 34 specimens each, totaling 68 spe- cimens or more than 41 per cent, of the 163 specimens examined. Forty-four specimens or 27 per cent, were asymmetrical in the bilateral count. 86 Zoologica: New York Zoological Society [XXIX: 9 Breeding: One female (Cat. No. 6057) 59 mm. in length was in full breeding con- dition. This fish was taken at the surface on June 1; the ovaries were 13 mm. in length, and the total number of enlarged eggs about 2,700. Vertical distribution: This is one of the deeper-living species of the family. About one-third of the entire catch came from a depth of 500 fathoms, with numbers grading sharply down above and below this depth. The data is as follows: surface (96 fish), 100 fathoms (4 fish), 300 fathoms (18 fish), 400 fathoms (12 fish), 500 fathoms (130 fish), 600 fathoms (62 fish), 700 fathoms (11 fish), 800 fathoms and deeper (7 fish). Sociability : The only real hint of school- ing is in one surface net which captured 90 fish within an hour, 8.30 to 9.30 P. M. Of the remaining 54 nets, 28 contained only one or two fish each, while six nets took numbers ranging from 10 to 28 individuals. Luminescence and viability: Three large fish were taken into the darkroom as soon as caught. They showed at first no general luminescence but a multitude of bright sparks. These increased upon agitation of the water and seemed to come from every photophore on the body. Only twice there- after, within a space of five minutes, did I s°e very briefly the dull, indirect glow so characteristic of a species such as Mycto- phum affine. The light was clear white and brightly illumined the entire dish and all the surrounding plankton. Usually the fish lay on their sides, and there was little light visible from beneath their bodies. It van- ished abruptly with death. Study Material. A total of 340 specimens was taken, as follows : Arcturus Expedition, Station 50 Pt-l(l) ; 50 Pt-l(l) ; 53 T-l(3) ; 56 T-l(6) ; 59 T-2 (4) ; 59 Pt-1(7); 59 Pt-1 (7) ; 61 T-4 ( 2) ; 65 Ot-1 and T-3 and T-4(90); 66 T-2 (24); 66 Pt-1 (20); 66 Pt-l(l); 66 T-2 (1); 66 T-2 (6); 66 T-3 (16) ; 67 T-l(l); 68 Pt-1 (2); 68 Pt-1 (5); 74 Pt-3 (3) ; 74 T-22(3) ; 74 T-26(l); 74 T-42 (1 ) ; 74 T-69 (2 ) ; 84 T-10 (2 ) ; 84 T-14 ( 1 ) ; 84 Pt-1 (10); 84 Pt-3(1); 84 Pt-4 (2 ) ; 84 T-19 (1 ) ; 84 T-20(3); 86 T-4 (2) ; 86 T-7(l) ; 86 T-8(3) ; 86 Pt-1 (3) ; 87 T-3 (4) 87 Pt-1 (6). Eastern Pacific ( Zaca ) Epedition, Station 185 T-l(l); 210 T-l (6) ; 210 T-2 (2); 210 T-3 (5); 210 T-6 (3) ; 210 T-6 (28); 210 T-7(l); 210 T-7 (16); 210 T-8 ( 2 ) ; 210 T-10(8); 210 T-10 (1); 219 T-l (9); 219 T-2(2); 219 T-2 (1) ; 225 T-l (2); 225 T-l(l); 227 T-l (3); 228T-1 (2) ; 230 T-l(l). For de- tailed trawling data, refer to Zoologica, Vol. VIII. pp. 6-23 ( Arcturus <) ; Vol. XXIII, pp. 287-298 {Zaca). References and Synonymy. Lampanyctus omostigma: Gilbert, 1908, p. 232, pi. 5. (Type spe- cimen ; 62 mm. ; also co-type same length ; surface; 1,000 miles north of Marquesas). Jordan & Jordan, 1922, p. 12. (Name listed) . Fowler, 1928, p. 68. (Description, syn- onymy). Parr, 1928, p. 88. (Key to species). Parr, 1929, p. 22. (Reexamination of type ) . Parr, 1931, pp. 25, 28, fig. 9. (Key to species) . Lampanyctus omostigma parvicauda: Parr, 1931, p. 25, fig. 9. (28 specimens, including type and co-type; ? mm.; 100 to 200? fathoms; off western coast Mexico and Nicaragua) . Lampanyctus ritteri Gilbert, 1915. Specimens taken by Eastern Pacific Ex- peditions: 19 specimens; Arcturus Oceano- graphic Expedition; 400 to 800 fathoms; between 0° 17' S. and 6° 27' N. Lat., and 85° 00' and 91° 47' W. Long.; lengths 15 to 68 mm.; between May 13 and June 12, 1925. 1 specimen; Eastern Pacific {Zaca) Ex- pedition; 500 fathoms; 8° 03' N. Lat., and 83° 12' W. Long., length 45 mm.; March 10, 1938. Specimens previously recorded: 27 speci- mens; ? to 142 mm.; 350 (?) to 599 fath- oms off coast of southern California. The present paper extends the range of this species about 1,920 miles south to the Galapagos. Text-fig. 19. Distribution of Lampanyctus ritteri. Photophore count and variation: There is considerable variation in the count of anal photophores given by other authors in earlier keys. Gilbert, who described this species, gives 7-8; Bolin 6-8 and 8-9; Parr 8 and 7-8. Our single (Zaca) specimen shows 5-7 on left side, 5-8 on right. The majority of our Arcturus fish are slightly mutilated, but the extremes of the countable series are 5-8 and 9-10. Food: The stomach of one fish contained 1944] Beebe & Vander Pyl: Pacific Myctophidae 87 two euphausids and one copepod, while an- other had indulged in two copepods and one euphausid. Study Material. A total of 20 specimens was taken, as follows: Arcturus Expedition, Station 66 Pt-1 (2 ) ; 68 T-5 (3) ; 68 Pt-1 (8) ; 74 Pt-1 (1) ; 74 Ot-2 ( 1 ) ; 84 Pt-1 (1 ) ; 84 Pt-3 (6) ; 84 T-8(l) ; 86 T-10(l). Eastern Pacific (Zaca) Expedition Station 219 T-2 ( 1 ) . For detailed trawling data, refer to Zoologica, Vol. VIII, pp. 6-23 ( Arcturus ) ; Vol. XXIII, pp. 287- 298 (Zaca). References. Lympanayctus ritteri: Gilbert, 1915, p. 318, pi. 15, fig. 3. (5 specimens, including type; ? to 142 mm. total length ; 350 to 599* fathoms ; Monterey Bay, California). Parr, 1928, p. 89, fig. 24. (Key to species). Parr, 1929, p. 23. (Further description). Bolin, 1939, p. 141, fig. 24. (24 specimens; including 2 co-types; 19.9 to 100.0 mm.; ? fathoms; vicinity San Clemente Island to Point Reyes, California). Lampanyctus tenuiforme (Brauer, 1906). Specimens taken by Eastern Pacific Ex- peditions: 5 specimens; Arcturus Oceano- graphic Expedition; 500 to 700 fathoms; be- tween 0° 17' S. Lat. and 0° 00' equator, and 91° 34' and 91° 53' W. Long.; lengths 24 to 25.5 mm.; June 9 to 13, 1925. Specimens previously recorded: 4 spe- cimens; 16 to 24 mm.; (?) 1,093 to 1,642 fathoms; Hudson Gorge, northwest Atlantic, south Atlantic, Indian Ocean. Our speci- mens are the first recorded from the east- ern Pacific. Text-fig. 20. Distribution of Lampanyctus tenuiforme. Photophore variation: Analysis of our five specimens shows the following arrange- ments: 6-9, 6-9, 6-10, 7-8, and one with the asymmetrical combination of 6-9, 6-10. No sex glands are present in our indi- viduals. Study Material. A total of 5 specimens was taken as fol- lows: Arcturus Expedition, Station 84 Pt-3 (3 ) ; 84 T-14(l) ; 87 Pt-l(l). For de- tailed trawling data, refer to Zoologica, Vol. VIII, pp. 6-23. References and Synonymy. (Synonymy accepted according to Parr, 1928, p. 83). Myctoplmm (Lampanyctus) tenuiforme: Brauer, 1906, p. 243, fig. 160. (Original description; 1 specimen; 21 mm.; 1,093 fathoms; Indian Ocean). Pappenheim, 1914, p. 197. (1 specimen; 16 mm.; ? to 1,642 fathoms; south Atlantic). Lampanyctus tenuiforme : Parr, 1928, p. 83. (Key to species; syn- onymy ) . Parr, 1931, p. 28. (Reference). Beebe, 1929, p. 16. (2 specimens; 18 and 24 mm.; 600 fathoms; Hudson Gorge, north- west Atlantic). Lampanyctus townsendi (Eigenmann & Eigenmann, 1889). Specimens taken by Eastern Pacific Ex- peditions: 1 specimen; Arcturus Oceano- graphic Expedition; 600 fathoms; 6° 24' N. Lat., and 85° 00' W. Long.; length 35 mm.; June 9, 1925. Specimens previously recorded: 43 spe- cimens, ? to 78.5 mm.; surface to 1,367 fathoms ; north and south Atlantic Ocean, Indian Ocean, Australia, Japan, mid- and eastern Pacific from southern California to Panama Bay. Our specimen extends the range southward about 2,000 miles. Text-fig. 21. Distribution of Lampanyctus townsendi. General data: Our specimen is badly dam- aged and torn but all the characters are distinct. It is Cat. No. 5687b, and was taken at Station 66 Pt-1, 130 miles northeast of Cocos Island. Study Material. One specimen was taken, as follows: Arc- turus Expedition, Station 66 Pt-l(l). For detailed trawling data, refer to Zoologica, Vol. VIII, pp. 6-23. 88 Zoologica: New York Zoological Society [XXIX: 9 References and Synonymy. (Synonymy accepted according to Bolin, 1939, p. 147). Myctophum townsendi: Eigenmann & Eigenmann, 1889, p. 125. (Original description; type specimen; ? mm.; 45 fathoms; Cortez Bay, San Diego, California) . Scopelus ( Nyctophus ) warmingii: Liitken, 1892, p. 259, fig. 19. (Original description; type specimen; 63 mm.; mid- Atlantic). Scopelus townsendi: Liitken, 1892, p. 237. (This refers to M. calif orniense) . Lampanyctus warmingii : Goode & Bean, 1895, p. 80. (Description). Lampanyctus townsendi: Jordan & Evermann, 1896, p. 558. (Des- cription). Gilbert, 1908, p. 230, pi. IV. (1 specimen; 26 mm.; ? to 300 fathoms; Marquesas). Gilbert, 1913, p. 98. (Key to species; de- scription; 5 specimens; ? to 60 mm.; surface to 850 fathoms; southern Japan). McCulloch, 1923, p. 115. (1 specimen; 65 mm.; on beach, Lord Howe Island, Aus- tralia). Fowler, 1928, p. 68. (References). Parr, 1928, p. 79. (Key to species; syn- onymy). Parr, 1929, p. 13. (Considers M. ivarm- ingii after examination of type). Norman, 1930, p. 327. (14 specimens; 17 to 67 mm.; surface to 1,094 fathoms; north and south Atlantic). Matsubara, 1938, p. 56, fig. 11. (1 speci- men; 78.5 mm.; ? fathoms; Japan). Bolin, 1939, pp. 93, 147, fig. 28. (Key to species; description; 7 specimens; 17.4 to 57.1 mm.; ? fathoms; coast of California, Japan ) . Myctophum ( Lampanyctus ) warmingii: Brauer, 1906, p. 229. (Key to species; description; 7 specimens; ? to 21 mm.; surface to 1,367 fathoms; Gulf of Guinea, Indian Ocean, Bay of Bengal, south of Japan) . Pappenheim, 1914, p. 195. (5 specimens; 19 to 36 mm.; ? to 3,000 metres wire; mid- and south Atlantic). Myctophum ( Lampanyctus ) townsendi: Brauer, 1906, p. 167. (Key to species). Diaphus dumeril! (Bleeker, 1856). Specimens taken by Eastern Pacific Ex- peditions: 3 specimens; Eastern Pacific (Zaca) Expedition; surface; at 16° 30' N. Lat., and 99° 13' W. Long.; lengths 53, 54 and 64 mm.; Nov. 29, 1937. Specimens previously recorded: 65 spe- cimens; ? to 95 mm.; surface to (?) 800 fathoms; northwest Atlantic, Bermuda, Ba- hamas, West Indies, Gulf of Mexico, south- west of Ireland, south Atlantic, Fiji Islands, Hawaii and Celebes. Text-fig. 22. Distribution of Diaphus dumerili. Photophore variation: Two fish showed the combination of 7-5, while the third pre- sented the asymmetrical formula of 7-6, 6-6. Study Material. A total of 3 specimens was taken, as fol- lows: Eastern Pacific (Zaca) Expedition, Station 190 L(3). For detailed trawling data, refer to Zoologica, Vol. XXIII, pp. 287-298. References and Synonymy. ( Synonymy accepted according to Parr, 1928, ‘p. 126). Scopelus dumerilii: Bleeker, 1856, vol. 1, p. 66. (Description and type locality, Manado, Celebes). Gunther, 1864, vol. 22, p. 410. (Brief de- scription). Gunther, 1887, p. 198. (Description; 1 specimen; 76 mm.; 315 fathoms; Fiji Is- lands) . Scopelus schmitzi: Johnson, 1890, p. ? (Reference not seen; synonymized by Taning, 1928, p. 58). Myctophum nocturnum: Poey, 1860, p. 416. (Brief description; 1 specimen; 95 mm.; ? fathoms; ? locality). Myctophum ( Nyctophus ) lacerta: Brauer, 1904, p. 392. -(Key to species). Collettia nocturna: Jordan & Evermann, 1896, p. 567. (De- scription). Diaphus nocturnus: Gilbert, 1906, p. 255. (Detailed descrip- tion of types and co-types). Myctophum dumerili : Weber & Beaufort, 1913, p. 670. (Brief reference ) . 1944] Beebe & Vander Pyl: Pacific Myctophidae 89 Lampanyctus lacerta: Goode & Bean, 1895, p. 81, fig, 89. (Type description; 3 specimens; ? to 57 mm.; sur- face to 671 fathoms; northwest Atlantic, Gulf of Mexico). Jordan & Evermann, 1896, p. 560. (Brief description) . Breder, 1927, p. 17. (Brief description; ? specimens; 50 mm.; ? fathoms; West Indies) . Diaphus dumerili nocturnus: Taning, 1928, p. 58. (Key to species). Diaphus dumerili: Fowler, 1928, p. 68. (Brief description; synonymy; 1 specimen; ? mm.; ? fathoms; Hawaii) . Parr, 1928, pp. 115, 126-130, fig. 23. (Key to species; synonymy; detailed description and discussion; 47 specimens; ? to 73 mm.; surface to 800 fathoms; San Salvador and Acklin Islands, in the Bahamas). Parr, 1929, pp. 29, 31. (Key to species; synonymy ) . Parr, 1934, p. 48. (Synonymy). Norman, 1929, p. 511. (1 specimen; 53 mm. ; ? fathoms ; off southwest Ireland ) . Norman, 1930, p. 331. (1 specimen; 47 mm.; 54 to 108 fathoms; south Atlantic). Borodin, 1931, p. 76. (? specimens; ? mm.; ? fathoms; near Bermuda). Fraser - Brunner, 1935, p. 319. (Name listed) . Beebe, 1929, p. 15. (2 specimens; ? mm.; ? fathoms; Hudson Gorge, northwest At- lantic) . Beebe, 1937, p. 204. (Preliminary list; 7 specimens; 54 to 66 mm.; surface; Ber- muda). Diaphus gemellari (Cocco, 1838). Specimens taken by Eastern Pacific Ex- peditions: 1 specimen; Arcturus Oceano- graphic Expedition; 700 fathoms; between 0° 40' N. Lat., and 88° 51' W. Long.; length 33 mm.; April 3, 1925. Specimens previously recorded: 61 speci- mens; ? to 86 mm.; surface to 1,000 fathoms; northwest Atlantic, Bermuda, Ba- hamas, West Indies, Canary Islands, Med- iterranean, Indian Ocean. Text-fig. 23. Distribution of Diaphus gemellari. Photophore count: Our specimen shows 5-6, 5-6. Study Material. One specimen was taken as follows : Arc- turus Expedition, Station 33 Pt-l(l). For detailed trawling data, refer to Zoologica, Vol. VIII, pp. 6-23. References and Synonymy. (Synonymy accepted according to Parr, 1928, p. 123). Nyctophus gemellarii: Cocco, 1838, p. ? (Reference not seen). Scopelus gemellarii: Cuvier & Valenciennes, 1828-49, p. 445. (Brief description). Gunther, 1864, V. 22, p. 415. (Descrip- tion). Canestrini, 1870, p. ? (Reference not S66I1 ) Doderlein, 1878-79, p. 54. (Name listed). Giglioli, 1880, p. ? (Reference not seen). Vaillant, 1888, p. 117. (Description; 6 specimens; ? mm.; 550 to 1,200 fathoms; Canary Islands, Messina, coast of Morocco). Raffaele, 1889, p. 183, fig. 8. (Brief de- scription). Carus, 1889-93, p. 564. (Description). Scopelus uraeoclampus : Facciola, 1882-83, p. 50. (Description of type; Messina). Scopelus ( Nyctophus ) gemellarii: Lutken, 1892b, p. 260. (Description). Myctophum gemellarii: Bonaparte, 1832-41, V. 3, Fasc. 27, fig. 2. (Detailed description). Sanzo, 1918a, p. 9. (Description; 8 spe- cimens; 3 to 17 mm.; ? fathoms; Mediter- ranean ) . Myctophum ( Nyctophus ) gemellarii: Brauer, 1904, p. 393. (Key to species). Lampanyctus gemellarii: Goode & Bean, 1895, p. 80. (Description and synonymy). Myctophum ( Diaphus ) gemellari: Brauer, 1906, pp. 164, 212, figs. 130, 131. (Description; key to species; synonymy; 5 specimens; 12 to 18 mm.; ? fathoms; Mes- sina) . Fage, 1910, p. 13, pi. 1, fig. 7. (Descrip- tion; 1 specimen; 32 mm.; 500 fathoms; Mediterranean) . Zugmayer, 1911b, p. 29. (Description; synonymy; 1 specimen; 30 mm.; ? fathoms; Mediterranean, Atlantic and Indian Ocean ) . Diaphus gemellari: Taning, 1918, p. 73. (Description; devel- opment; 78 specimens; 5 to 24 mm.; sur- face to 55 fathoms; Mediterranean). 90 Zoologica : New York Zoological Society [XXIX: 9 Taning, 1928, p. 58. (Key to species). Parr, 1928, pp. 115, 123.-125, figs. 21, 22. (Key to species; synonymy; detailed de- scription; 5 specimens; ? to 45 mm.; 5 to 8,000 feet wire; Bahamas, Bermuda). Parr, 1929, p. 29. (Key to species). Parr, 1934, p. 64. ( Description ; 2 spe- cimens; 37 and 67 mm.; 600 fathoms; north Atlantic) . LeGendre, 1934, p. 340. ( Key to species ; 2 specimens 1 male, 1 female; 83 and 86 mm.; ? fathoms; Gulf of Gascony). Beebe, 1929, p. 15. (3 specimens; ? mm.; ? fathoms; Hudson Gorge, north Atlantic). Beebe, 1937, p. 204. (Preliminary list; 27 specimens; 10 to 28 mm.; 300 to 1,000 fathoms; Bermuda I . Diaphus pacificus (Parr, 1931). Specimens taken by Eastern Pacific Ex- peditions: 19 specimens; Eastern Pacific ( Zaca ) Expedition: 300 to 400 fathoms; at 17° 45' N. Lat., and 103° 05' W. Long.; lengths 24 to 30 mm.; November 23, 1937. Specimens previously recorded: 1 speci- men; 28 mm.; 625 fathoms cable; eastern Pacific coast of Mexico. Text-fig. 24. Distribution of Diaphus pacificus. Photophore count and variation: In gen- eral 5-4 on both left and right side; several were 4-4 on both sides, and one was 5-3. Asymmetrical combinations were 5-4 and 4-4, 4-5 and 5-4 and a third showed 5-3, and 4-5. Study Material. A total of 19 specimens was taken, as fol- lows: Eastern Pacific (Zaca) Expedition, Station 185 T-l (13), 185 T-2 (6). For de- tailed trawling data, refer to Zoologica, Vol. XXIII, pp. 287-298. References and Synonymy. Diaphus pacificus: Parr, 1931, p. 34, fig. 14. (Description of type specimen; 28 mm.; 625 fathoms cable; eastern Pacific, off coast of Mexico. Diaphus rahnesquii (Cocco, 1838). Specimens taken by Eastern Pacific Expe- ditions: 19 specimens; Arcturus Oceano- graphic Expedition; 300 to 1,000 fathoms; between 0° 17' S. and 4° 50' N. Lat., and 86° 43' and 91° 53' W. Long.; lengths 11 to 39 mm.; between May 1 and June 13, 1925. 4 specimens; Templeton Crocker Expedi- tion; 400 to 500 fathoms; at 23° 25' N. Lat., and 108° 31' W. Long. ; lengths 15 to 27 mm. ; April 29, 1936. 1 specimen; Eastern Pacific (Zaca) Expe- dition; 500 fathoms; at 17° 45' N. Lat., and 103° 05' W. Long.; Nov. 23, 1937. Specimens previously recorded: 1,032 specimens; ? to 90 mm.; surface to 1,000 fathoms; north and south Atlantic; Mediter- ranean, Japan, mid-Pacific, eastern Pacific from Sitka, Alaska down to San Diego, Cali- fornia. Text-fig. 25. Distribution of Diaphus rafinesquii. Photophore variation: There were five symmetrical combinations in 22 specimens; one was 4-4; eleven were 5-4; two were 5-3; three were 6-3 ; and five were 6-4. Two fish had the asymmetrical combinations of 4-5, 5-4, and 5-4, 4-4. One of these asymmetrical specimens, (Cat. No. 5571) had the charac- terestics of D. holti on one side. Vertical distribution: If anything of value can be deduced from the vertical distribu- tion of 24 fish, it is that almost half were taken at 500 fathoms. The exact data is : 300 fathoms (2 fish), 400 (7), 500 (11), 600 (1 ), 700 (2) and 1,000 fathoms (1 fish). Sociability : Nothing definite. Eight nets contained one fish each, five nets two, and one net six fish. Study Material. A total of 24 specimens was taken, as fol- lows: Arcturus Expedition, Station 59 T-3 (2) ; 74 T-78 ( 1 ) ; 84 T-20(2); 86 T-l (1) ; 86 T-2 ( 6 ) ; 86 T-7 ( 1 ) ; 86 T-9 (1 ) ; 86 T-ll(l) ; 87 T-2 ( 1 ) ; 87 T-3 ( 2 ) ; 87 T-6(l). Templeton Crocker Expedition, Station 158 T-3 (2) ; 158 T-4(2 ). Eastern Pacific (Zaca) Expedition, Station 185 T-3(l). For detailed trawling data, refer to Zoologica, Vol. VIII, pp. 6-23 (Arcturus)', Vol. XXII, pp. 37-46 (Templeton Crocker) ; Vol. XXIII. pp. 287- 298 (Zaca). 1944] Beebe & Vander Pyl: Pacific Myctophidae 91 References and Synonymy. (Synonymy accepted according to Bolin, 1939, p. 125). Nyctophus rafinesquii: Cocco, 1838, p. 20, pi. 3, fig. 7. (Original description; reference not seen). Myctophum rafinesquei : Bonaparte, 1832-1841, V. 3, Fasc. 27, fig. 2. (Description). Sanzo, 1918a, p. 5. (Discussion). Scopelus rafinesquii : Cuvier & Valenciennes, 1828-1849, vol. 22, pp. 330, 444. (Brief description). Gunther, 1864, vol. 22, p. 410. ( Descrip- tion of Mediterranean fish). Giglioli, 1880, p. ? (Reference not seen). Doderlein, 1878-79, p. 54. (Name listed). Canestrini, 1870, p. 125. (Reference not seen) . Leydig, 1881, p. ? (Reference not seen). Raffaele, 1889, p. 183. (Brief description; 13 specimens; ? mm.; ? fathoms; Messina). Cams, 1889-1893, p. 564. (Description). Moreau, 1891, p. 98. (Detailed description and synonymy). Holt & Byrne, 1910, p. 22, fig. 6. (Descrip- tion; 1 specimen; 75 mm.; 730 fathoms; southwest coast of Ireland). Scopelus rafinesque: Weill, 1926, p. 472. (Brief discussion of luminosity) . Scopelus ( Nyctophus ) rafinesquii: Liitken, 1892, p. 258, fig. 17. (Descrip- tion; 3 specimens; 43, 64 and 90 mm.; ? fathoms; ? locality). Scopelus protoculus : Liitken, 1892, p. 266. (Reference). Myctophum protoculus : Gilbert, 1890, p. 52. (Description of type; 3 specimens; ? mm.; 584 fathoms; off Cape Johnson, Washington). Parr, 1929, p. 32. (Discussion of syn- onymy) . Myctophum ( Nyctophus ) rafinesquii: Brauer, 1904, p. 393. (Key to species; synonymy) . Myctophum ( Nyctophus ) theta: Brauer, 1904, p. 393. (Key to species). Myctophum ( Diaplius ) theta: Brauer, 1906, p. 165. (Key to species). Diaphus theta: Eigenmann & Eigenmann, 1890, p. 4. (11 specimens from the mouths of rock cod, Point Loma, California). Evermann & Goldsborough, 1906, p. 271. (1 specimen; 57 mm.; 922 fathoms; Sitka, Alaska) . Diaphus nanus: Gilbert, 1908, p. 224, pi. 1. (Type de- scription ; 1 specimen ; 17 mm. ; surface to 300 fathoms; Marquesas). Gilbert, 1913, p. 87. (Description; ? speci- mens; ? 45 mm.; ? 300 fathoms; Japan). Diaphus holti: Taning, 1918, pp. 88-92, figs. 35, 36. (De- tailed description; distribution map; 638 specimens; 275 to 550 fathoms; near Mes- sina, Mediterranean and Bay of Cadiz). Diaphus mollis: Taning, 1928, p. 60. (Preliminary type description; key to species). Collettia rafinesquei: Goode & Bean, 1896, p. 88, pi. 26, fig. 100. (Description; synonymy; 3 specimens; ? to 77 mm.; northwest Atlantic). Jordan & Evermann, 1896, p. 567. (Des- cription ) . Myctophum ( Diaphus ) rafinesquei: Brauer, 1906, pp. 165, 223, figs. 144, 145. (Key to species; description; synonymy; 25 specimens; ? to 66 mm.; ? fathoms; Mes- sina) . Zugmayer, 1911, p. 32. (Synonymy; 2 specimens; ? to 30 mm.; ? fathoms; Medi- terranean ) . Diaphus rafinesquii: Taning, 1918, pp. 83-87, figs. 33, 34. (Detailed description; distribution map; 123 specimens; surface to 1,093 fathoms; Mediterranean) . Taning, 1928, p. 60. (Key to species). Parr, 1928, pp. 119, 131-135, figs. 25, 26. (Key to species; synonymy ; detailed discus- sion; 4 specimens; 38 to 46 mm.; 7 to 10,000 feet wire; Bahamas, Bermuda). Parr, 1929, p. 32, fig. 16. (Key to spe- cies; synonymy; detailed discussion and description ) . Parr, 1934, pp. 51, 65. (Reference to type in poor condition; also 1 specimen; ? mm.; 600 fathoms; mid-north Atlantic). Norman, 1930, p. 334. (References; brief discussion; 27 specimens; 16 to 70 mm.; ? fathoms ; south Atlantic ) . LeGendre, 1934, p. 341. (Key to species). Beebe, 1937, p. 203. (Preliminary list; 134 specimens; 9 to 70 mm.; 400 to 1,000 fathoms ; Bermuda ) . Bolin, 1939, pp. 93, 125, fig. 18. (Key to species; synonymy; detailed description; 23 specimens, including 4 co-types 11 to 65.6 mm.; ? fathoms; from mid-Washington to southern California, Japan, Messina). Chapman, 1940, p. 33. (Description; 22 specimens; 45 to 73 mm.; ? to 850 metres wire; eastern Pacific between Alaska and British Columbia). 92 Zoologica : New York Zoological Society [XXIX: 9 Bibliography. Alaejos, Sanz. L. 1919. Datos para la fauna ictiologica de las costas de Santander. Bol. Pescas, Ma- drid, 1919., Alcock, A. W. 1890. Natural history notes from H. M. In- dian Marine Survey Steamer “Investi- gator,” No. 18. On the bathybial fishes of the Arabian Sea obtained during the season 1889-1890. Ann. Mag. Nat. Hist., Ser. 6, Vol. 6, 1890. 1896. Natural history notes from H. M. In- dian Marine Survey Steamer “Investi- gator,” Ser. II, No. 23. A supplemen- tary list of the marine fishes of India, with descriptions of 2 new genera and 8 new species. Ibid., Vol. 65, Pt. 3, 1896. Barnard, H. K. 1925. A monograph of the marine fishes of South Africa, Pt. 1. Ann. South Afri- can Mas., Vol. 21, 1925. Beebe, William. 1929. Deep-sea fish of the Hudson Gorge. Zoologica, Vol. XII, No. 1, 1929. 1937. Preliminary list of Bermuda deep-sea fish. Based on collections from fifteen hundred metre-net hauls, made in an eight-mile circle south of Nonsuch Island, Bermuda. Zoologica, Vol. XXII, No. 27. Bleeker, P. 1856. Beschrijvingen van nieuwe of wenig bekende vischsoorten van Menado en Makassar. Act. Soc. Sci. Indo-Neerl., Vol. 1, 1856. 1862-77. Atlas ichthyologique des Indes Ori- entales Neerlandaises. Amsterdam, 1862-77. Bolin, Rolf L. 1939. A Review of the Myctophid Fishes of the Pacific Coast of the United States and of Lower California. Stanford Ich- thyological Bulletin, Vol. 1, No. 4. Bonaparte, C. L. 1832-41. Iconografia della fauna italica, Tomo. III. Pesci. Roma, 1832-41. Borodin, Nicholas. 1931. Atlantic Deep-Sea Fishes. Bull. Mu- seum Comp. Zool., Harvard College, Vol. LXXII, No. 3. Cambridge, August, 1931. Brauer, A. 1904. Diagnosen von neuen Tiefseefischen, welche von der Valdivia Expedition gesammelt sind. Die Gattung Mycto- phum. Zool. Anz., Vol. 28, 1904. Brauer, A. 1906. Die Tiefsee-Fische, I, Systematischer Teil (II, Anatomischer Teil, 1908). Wiss. Ergebnisse Deutsch. Tiefsee- Exp. Valdivia, Vol. 15, Lief. 1. Jena, 1906. Breder, C. M. 1927. Fishes. Scientific results of the first oceanographic expedition of the “Paw- nee,” 1925. Bull. Bingham Oceano- graphic Coll., Vol. 1, Art. 1, 1927. Canestrini, G. 1870. Pesci d’ltalia. Cornalia, E.: Fauna d’ltalia, Pt. Ill, Milano, 1870-1874. Carus, J. V. 1889-1893. Prodromus faunae Mediterra- neae, Vol. 2, Vertebrata, 1 class, Pisces. Stuttgart, 1889-1893. Chapman, W. McL. 1940. Oceanic Fishes from the Northeast Pacific Ocean. Occasional Papers Brit- Columbia Provincial Museum, No. 2, May 11, 1940. Cocco, A. 1829. Su di alcuni nuovi pesci del mare di Messina. Giorn. Sci. Lett. Arti. Sicil., Palermo, 1829. 1838. Su di alcuni Salmonidi del mare di Messina; lettera al Ch. D. Carlo Lu- ciano Bonaparte. Nuovi. Ann. Sci. Nat. Bologna, Vol. 2, 1838. Collet, R. 1880. Den norske Nordhavs-expedition 1876- 1878. Zoologi. Fiske., Christiania, 1880. 1880a. Meddelelser om Norges fiske i aarene 1875-1878. Vidensk. Selsk. Forhandl., 1879. Christiania, 1880. 1884. Meddelelser om Norges fiske i aarene 1879-1883. Nyt. Mag. for Naturv., Vol. 29. Christiania, 1884. 1896. Poissons provenant des campagnes du yacht “L’Hirondelle” (1885-88). Re- sult. Camp. Sci. Monaco, Vol. 10, 1896. Costa, O. G. 1844. Pesci della fauna napolitana con il- lustrazione di specie nuovi, Vol. 1, Na- poli, 1844. Cuvier, G. L. C. F. D., & Valenciennes, A. 1828-49. Histoire naturelle des poissons. Paris, 1828-49. Doderlein, P. 1878-79. Prospeto metodico delle varie specie di Pesci . . . della Sicilia. Atti dell’ Acad. Sci. Lett. Arti. Palermo, n. ser., Vol. 6, 1878-79. Eigenmann, C. H., & Eigenmann, R. S. 1889. Notes from the San Diego Biological Laboratory. The fishes of Cortez Banks. West Amer. Scientist, Vol. 6, No. 48. San Diego, 1889. 1890. Additions to the fauna of San Diego. Proc. Calif. Acad. Sci., Ser. II, Vol. 3, Pt. 1. San Francisco, 1891. Evermann, B. W., & Goldsborough, E. L. 1906. The Fishes of Alaska. Bull. U. S. Bu- reau of Fisheries, Vol. 26. 1944] Beebe & Vander Pyl : Pacific Myctophidae 93 Evermann, B. W. & Seale, A. 1907. Fishes of the Philippine Islands. Bull. Bur. Fisheries, Dept. Comm. Labor, Washington, Vol. 26, 1906 (1907). Facciola. 1882-83. Pesci nuovi o poco noti dello Stretto di Messina. (Various articles, see Zool. Record for 1883). FI. Natural. Sicil., I-III., Palermo, 1882-1883. Fage, L. 1910. Recherches sur les stades pelagiques de quelques Teleosteens. Ann. L’Inst. Oceanogr., Vol. 1, No. 7. Monaco, 1910. Fowler, H. W. 1925. New taxonomic names of West Afri- can marine fishes. Amer. Mus. Novi- tates, No. 162. New York, 1925. 1928. The fishes of Oceania. Mem. Bernice P. Bishop Mus., 10, 1928. Fraser-Bruner, A. 1935. New or Rare Fishes from the Irish Atlantic Slope. Proc. Roy. Irish Acad., B, 42 (9). Carman, S. 1899. The fishes. Report on an exploration by U. S. Fish Comm. Steamer “Alba- tross,” 1891. Mem. Mus. Comp. Zool., Harvard College, Cambridge, Mass., 1899. Gigli oli, E. H. 1880. Elenco dei mammiferi . . . appartenenti alia fauna italiana, e catalogo degli anfibi e dei pesci italiani. Firenze, 1880. Gilbert, C. H. 1890. Preliminary report on the fish collected by the steamer “Albatross” on the Pacific coast of North America dur- ing the year 1889. Proc. U. S. Nat. Mus., Vol. 13, 1890 (1891). 1905. The deep-sea fishes of the Hawaiian Islands. The aquatic resources of the Hawaiian Islands, Pt. II, Sec. 1. Bull. U. S. Fish Comm., Vol. 23, 1903. Wash- ington, 1905. 1906. Certain Scopelids in the collection of the Museum of Comparative Zoology. Bull. Mus. Comp. Zool., Harvard Col- lege, Cambridge, Mass., Vol. 46, No. 14, 1906. 1908. The lantern fishes. Rept. Sci. Res. Exp. Trop. Pacific “Albatross,” 1899-1900, X. Mem. Mus. Comp. Zool., Harvard College, Cambridge, Mass., Vol. 26, No. 6, 1908. 1911. Notes on lantern fishes from southern seas, collected by J. T. Nichols in 1906. Bull. Amer. Mus. Nat. Hist., Vol. 30, Art. 2, 1911. 1913. The lantern fishes of Japan. Mem. Carnegie Mus., Vol. 6, No. 2, 1913. 1915. Fishes collected by the U. S. Fisheries steamer “Albatross” in 1904. Proc. U. S. Nat. Mus., Vol. 48, 1915. Gilchrist, J. D. F. 1908. Description of fifteen new South Afri- can fishes. Mar. Inv. South Africa, Vol. 4, 1908. Goode, G. B. & Bean, T. H. 1895. Oceanic ichthyology. A treatise on the deep-sea and pelagic fishes of the world. Special Bull. U. S., Nat. Mus., Washington, 1895. Gunther, A. 1864. Catalogue of the fishes in the collection of the British Museum, London, Vol. V, 5. London, 1864. 1887. Report on the deep-sea fishes. Rept. Sci. Res. “Challenger,” Vol. 22, 1887. 1889. Report on the pelagic fishes. Ibid., Vol. 31, 1889. Holt, E. W. L., & Byrne, L. W. 1906. First report on the fishes of the Irish Atlantic slope. Fisheries, Ireland, Sci. Invest. 1905. Append. II, 1906. 1911. Fifth report on the fishes of the Irish Atlantic slope. Ibid., 1910, VI, 1911. Hutton, F. W. 1873. Contributions to the ichthyology of New Zealand. Trans. Proc. New Zea- land Inst., Vol. 5, 1872. Wellington, 1873. Johnson, J. Y. 1890. On some new species of fishes from Madeira. Proc. Zool. Soc. London, 1863. Jordan, D. S. & Evermann, B. W. 1896. The fishes of North and middle Amer- ica, Vol. 1. Bull. U. S. Nat. Mus., 1896. Washington, 1896. 1898. The fishes of North and middle Amer- ica, Vol. Ill, Bull. U. S. Nat. Mus., 1898. Washington, 1898. 1903. Descriptions of new genera and species of fishes from the Hawaiian Islands. Bull. U. S.. Fish Comm., Vol. 22, 1902. Washington, 1903. Jordan, D. S., & Jordan, E. K. 1922. A list of the fishes of Hawaii, with notes and descriptions of new species. Mem. Carnegie Mus., Vol. 10, No. 1, 1922-1925. Koehler, R. 1896. Resultats scientifique de la campagne du “Caudan” dans le Golf de Gascogne. Ann. Univ. Lyon, Paris, Fasc. 26. Paris, 1896. LeGendre, R. 1934. Annales de L’lnstitut Oceanogra- phique. Tome XIV, Fasc. VI. La Faune Pelagique de L’Atlantique au large du Golfe De Gascogne, Paris, 1934. Leydig, Fra. 1881. Die augenahnlichen Organe der Fische. Bonn, 1881. 94 Zoologica :,New Y ork Zoological Society [XXIX: 9 Lilljeborg, W. 1884-91. Sveriges och Norges fiskar. Upsala, 1884-1891. Lonnberg, E. 1905. The fishes of the Swedish South Polar Expedition. IFiss. Ergeb, Schwed. Siid- polar. Exp., Vol. 5, Pt. 6, Stockholm, 1905. Lowe, R. T. 1839. A supplement to a synopsis of the fishes of Madeira. Proc. Zool. Soc., London, Vol. 7, 1839. 1849. Supplement to “A synopsis of the fishes of Madeira.” Trans. Zool. Soc. London, Vol. 3, 1849. Lutken, C. 1892a. Spolia Atlantica. Scopelini Musei Zoo- logici Hauniensis. Bidrag til kundskab om det aabne Havs Laxesild eller Scopeliner. Kgl. Danske Vidensk. Selsk. Skrifter, Raekke 6, Bind 7. Cop- enhagen, 1892. 1892b. Korte bidrag til nordisk Ichthyogra- phi, VIII. Nogle nordiske Laxesild ( Scopeliner) . Vidensk. Meddel. Copen- hagen, Ser. 5, Vol. 3, 1891 (1892). Malm, A. W. 1863. Nya fiskar, kraft-och blod-djur for Skandinaviens fauna. Vetensk. Selsk. Handl. Goteborg, 1863. Matsubara, K. 1938. Studies on the deep-sea fishes of Japan, VII. On some rare or imperfectly known lantern-fishes in Kumano-Nada. Journ. Imper. Fisheries hist., Tokyo, Vol. 33, figs. 9-12. McCulloch, A. E. 1923. Fishes from Australia and Lord Howe Island, No. 2. Rec. Austr. Mu.s., Sydney, 14, 1923. Moreau, E. 1881. Histoire naturelle des poissons de la France. Paris, 1881. 1891. Supplement a l’histoire naturelle des poissons de la France. Paris, 1891. Norman, J. R. 1929. Notes on the Myctophid Fishes of the Genera Diaphus and Myctophum in the British Museum. Ann. Mag. Nat. Hist., Ser. 10, vol. IV, November 1929. 1930. Oceanic fishes and flatfishes collected in 1925-1927. Discovery Reports, vol. 2, Cambridge. Pappenheim, P. 1914. Die Fische der Deutschen Sudpolar- Expedition 1901-1903, II. Die Tiefsee- fische. Deutsche Sudpolar-Exp. 1901- 1903, Bd. 15, Zoologie VII. Berlin, 1914. Parr, A. E. 1928. Deep-sea fishes of the order Iniomi . . . with annotated keys to the Sudidae, Myctophidae, Scopelarchidae, Ever- mannellidae, Omosudidae, Cetomimi- dae and Rondeletidae of the world. Bull. Bingham Oceanogr. Coll., New Haven, Vol. 3, art. 3. 1929. Notes on the species of the Myctophine fishes represented by type specimens in the U. S. National Museum. Proc. U. S. Nat. Mus., vol. 76, art. 10. 1931. Deep-sea fishes from off the western coast of North and Central America . . . and a revision of the Macropterus group of the genus Lampanyctus. Bull. Bingham Oceanogr. Coll. New Haven, vol. 2. 1934. Studies of Myctophinae in the Museum of Comparative Zoology. Bull. Mus. Comp. Zool. Harvard College, Vol. LXXVII, No. 2, May, 1934. Peters, W. 1864. Sitzung der physikalisch - mathema- tischen Klasse, 20 Juni. Monatsb. Ak. Wiss., Berlin, 1864. Poey, F. 1865. Repertorio fisico-natural de la isla de Cuba. Havana, 1865-1868. Raffaele, Fr. 1889. Note intorno alle specie mediterranee del genere Scopelus. Mitth. Zool. Stat. Neapel, 1889. Regan, C. T. 1916. Fishes. British Antarctic (“Terra Nova”) Expedition, 1910, Nat. Hist, Rept. Zoology, Larval and Postlarval Fishes, Vol 1, 1914-1916. Richardson, J. 1844-48. Ichthyology of the voyage of H. W. S. “Erebus and Terror.” Lon- don, 1844-1848. Roule, L. & Angel, F. 1930. Larves et alevins des poissons prove- nant des croisieres du Prince Albert 1st de Monaco. Resultats Campagnes Sci. Albert I. Prince de Monaco. 1933. Poissons provenant des campagnes du Prince Albert 1 de Monaco. Resultats des Camp. Sci. Sci. Monaco, Fasc. 86, 1933. Sanzo, L. 1918. Stylophthalmoides lobianco Mazzarelli e St. mediterraneus Mazzarelli sono le respettive forme larvali di Scopelus caninianus C. e. V. e Sc. humboldti Risso, R. Com. Talass. Ital. Venice. Contribute alia conoscenza dello svi- luppo postembrionale negli. Scopelini Muller. Memoria 66, 1918. Smitt, F. A. 1893-1895. A history of Scandinavian fishes, 2nd ed. Stockholm, 1893-1895. Steindachner, F. 1881. 1. Uber einige neue und seltene Meeres fische aus China. 11. Iehthyologische Notizen, V, 4, Uber 1944] Beebe & Vander Pyl: Pacific Myctophidae 95 eine neue Scopelus und Monacanthus- Art aus China. Ichthyologische Beitrage (XI). Sitz. Ber. Kais. Ak. Wiss. Wien, Vol. 83, Pt. 1, Vienna, 1881. Tanaka, S. 1908. Descriptions of eight new species of fishes from Japan. Annot. Zool. Japon- enses, Vol. 7, 1908. 1911. Figures and descriptions of the fishes of Japan. Journ. Coll. Sci. Imper. Univ. Tokyo, Vol. 4, Taning, A. V. 1918. Mediterranean Scopelidae. Report Danish Oceanog. Exped. 1908-1910, Vol. 2, A 7, Copenhagen. 1928. Synopsis of the Scopelids in the north Atlantic. Vidensk. Medd. Dansk. Na- turh. Foren., vol. 86. 1932. Notes on the Scopelids from the Dana Expedition, I. Ibid., vol. 94. Thompson, W. W. 1916. Catalague of fishes of the Cape Prov- ince. Marine Biol. Rept., No. Ill, Prov. Cape of Good Hope. Cape Town, 1916. Townsend, C. H., & Nichols, J. T. 1925. Deep-sea fishes of the “Albatross” Lower California Expedition. Bull. Amer. Mus. Nat. Hist., Vol. 52, Art. 1. New York, 1925. Vaillant, L. 1888. Expeditions Scientifiques du “Travail- leur” et du “Talisman” pendant les annees 1880, 1881, 1882, 1883. Poissons, Paris, 1888. VlNCIGUERRA, D. 1885. Appunti ittiologici sulle collezione del Museo Civico di Genova, VII, Sopra al- cuni pesci nuovi del Golfo di Genova. Ann. Mus. Civ. Storia Nat. Genova, Ser. 2, Vol. 2, Genova, 1885. Waite, E. R. 1904. Additions to the fish-fauna of Lord Howe Island, No. 4. Rec. Australian Mus., Vol. 5, Sidney, 1903. Weber, Max 1913. Die Fische der Siboga-Expedition. Si- boga-Expeditie, Monog. 57, Livr. 65. Leiden, 1913. Weber, M., & Beaufort, L. F. 1913. The fishes of the Indo-Australian Ar- chipelago, Vol. 2. Leiden, 1913. Weill, R. 1926. Poissons abyssaux, a organes lumi- neux, du Detroit de Messina (demon- stration). C. R. Soc. Biol., Paris, vol. 94. WlNTHER, G. 1879-80. Prodromus ichthyologiae Danicae marinae. Fortegnelse over de i Danske farvand hidtil fundne fiske. Naturhist. Tidsskrift, Raekke 3, vol. 2. Copen- hagen, 1879-80. . Zugmayer, E. 1911. Diagnoses des poissons nouveaux pro- venant des campagnes du yacht “Prin- cess Alice” (1901-1910). Bull. Inst. Oceanographique Monaco, 1911. NEW YORK ZOOLOGICAL SOCIETY General Office: 630 Fifth Avenue, New York 20, N. Y. OFFICERS Fairfield Osborn, President Alfred Ely, First Vice-president Laurance S. Rockefeller, Second Vice-president Harold J. O’Connell, Secretary Cornelius R. Agnew, Treasurer William Bridges, Editor and Curator of Publications Zoological Park Lee S. Crandall, General Curator & Curator of Birds Leonard J. Goss, Veterinarian Claude W. Leister, Associate, Mammals John Tee-Van, Associate, Reptiles Grace Davall, Assistant to General Curator Aquarium Christopher W. Coates, Curator and Aquarist Ross F. Nigrelli, Pathologist Myron Gordon, Assistant Curator C. M. Breder, Jr., Research Associate in Ichthyology G. M. Smith, Research Associate in Pathology Homer W. Smith, Research Associate in Physiology SCIENTIFIC STAFF General John Tee-Van, Executive Secretary Jean Delacour, Technical Adviser Claude W. Leister, Education Department of Tropical Research William Beebe, Director Jocelyn Crane, Research Zoologist Henry Fleming, Entomologist William K. Gregory, Associate Gloria Hollister, Associate John Tee-Van, Associate Mary VanderPyl, Associate Editorial Committee Fairfield Osborn, Chairman William Beebe William Bridges Lee S. Crandall Jean Delacour Claude W. Leister John Tee-Van ZOOLOGICA SCIENTIFIC CONTRIBUTIONS of the NEW YORK ZOOLOGICAL SOCIETY VOLUME XXIX Part 3 Numbers 10-15 Published by the Society The Zoological Park, New York November 22, 1944 CONTENTS Page 10. The Psammocharidae (Hymenoptera) Taken at Kartabo and Other Localities in British Guiana. By Nathan Banks 97 11. Larval Crabs from Bermuda. By Marie V. Lebour. (Text- figures 1-19) 113 12. Some Venezuelan Aquatic Hemiptera. By H. B. Hungerford 129 13. Ocular Anatomy and Light Sensitivity Studies on the Blind Fish from Cueva de los Sabinos, Mexico. By C. M. Breder, Jr. (Plate I; Text-figures 1-3) 131 14. Field Notes on the Lizards of Kartabo, British Guiana, and Caripito, Venezuela. Part I. Gekkonidae. By William Beebe. (Plates I-VI; Text-figures 1-6) 145 15. On the Color Changes of Fiddler Crabs (Genus Uca ) in the Field. By Jocelyn Crane 161 Banka: Ps ammo char idae from British Guiana 97 10. The Psammocharidae ( Hymenoptera ) Taken at Kartabo and Other Localities in British Guiana.1 Nathan Banks. [During the eight years of occupancy, by the Department of Tropical Research, of the New York Zoological Society’s Station at Kartabo, British Guiana, extensive collections of insects were made. Most of these were taken in the quarter-mile area under intensive study — what may be called the Guiana Junglezone. Details and a general summary of this area may be found in “Studies of a Tropical Jungle,” Zoologica, Vol. VI, No. 1.] Contents. page Introduction 97 Synopsis of Subfamilies 98 Pepsinae Synopsis of Pepsis 99 Pepsis frivaldszkyi Mocz . 100 Pepsis strenua Erichs 100 Pepsis plutus Erichs 100 Pepsis grossa Fabr 100 Pepsis elevata Fabr 100 Pepsis julgidipennis Mocz 100 Pepsis margarete Lucas 100 Pepsis niplie Mocz 100 Pepsis excelsa Lucas 100 Pepsis xanthocerus Dahlb 101 Peps's ruficornis Fabr 101 Pepsis luteicornis Fabr 101 Pepsis fiavicornis M^cz 101 Pepsis ianthina Erichs 101 Pepsis sagana Lucas 101 Pepsis vicina Lucas 101 Pepsis iucuiida Lucas 101 Pepsis dimidiata Fabr 101 Pepsis compJeta Sm 101 Pepsis pretiosa Dahlb 101 Pepsis equestris Erichs 101 Peps's diabolus Lucas 101 Pepsis sp. near ruficornis 102 Pepsis hecote Mocz 101 Pepsis gracilis Lepell 102 Pepsis sp. near smaragdina .102 Pepsis f estiva Fabr 102 Pepsis nireus Mocz 102 Pepsis f estiva Fabr 102 Cryptocheilinae Table of Genera 102 Priocnemis ( Calicurgus ) nubilus Fox 102 Colopompilus vitreus Fox 102 Priocnemioides Radoszkowski 102 Key to Species 103 Priocnemioides gigas Fabr 103 Priocnemioides perpunctatus Fox 103 Priocnemioides purpureipes Cam 103 Priocnemioides bituberculatus Guer 103 Priocnemioides aurifrons sp. nov 103 Bolboana fulvipes sp. nov 103 Balbonna auripennis Fabr 104 PriochPus 104 Key to Females 104 Priochilus regius Fabr 105 Priochilus rhomboideus Fox 105 Priochilus nobilis Fabr 105 Priochilus scrupulus Fox 105 Priochilus diversus Sm 105 Priochilus sericeifrons Fox 105 Priochilus opacifrons Fox 105 Priochilus imperius sp. nov 105 Priochilus superbus sp. nov 105 Priochilus plutonius sp. nov 106 Priochilus formosus sp. nov 106 iContribution No. 692, Department of Tropical Research, New York Zoological Society. PSEUDAGENINAE Key to Genera 106 Alasagenia gen. nov 107 Alasagenia erichsoni sp. nov 107 Priophanes fabricii sp. nov 107 Phanochilus gen. nov 108 Phanochilus nobilitata Sm 108 Phanochilus gloriosa Sm 108 Phanochilus pilifrons Cam 108 Key to Aneniella females 108 Ageniella micans Fabr 108 Ageniella reversa Sm 108 Ageniella delila sp. nov 108 Key to Pseudagenia Females 109 Pseudagenia femorata Fabr 109 Pseudagenia smithi D.T 109 Pseudagenia tarsata Sm 109 Pseudagenia modesta Sm 109 Pseudagenia comparata Sm 109 Pseudagenia basalis Sm .109 Pseudagenia incrota sp. nov 109 Priocnemella eurytheme sp. nov 110 Priocnemella difformis sp. nov 110 PSAMMOCHARINAE Key to Genera 110 Batazonus decedens Sm Ill Batazonus fervidus Sm Ill Batazonus polistoides Sm Ill Arachnophroctonus crassidentatus Cam Ill Psammochares inculcatrix Cam Ill Psammochares echinatus Fox Ill Psammochares ornamentus Fox Ill Notiochares amethystina Fabr Ill Aplochares gen. nov Ill Aplochares imitator Sm Ill Sericopompilus exilis sp. nov Ill Planiceps pertyi sp. nov. 112 112 Anoplius varius Fabr 112 Pompilinus orthodes sp. nov 112 Notocyphinae Notocyphus vindex Lucas 112 Notocyphus tyrranicus Lucas 112 Introduction. During his trips to the Tropical Research Laboratory of the New York Zoological Society at Kartabo, British Guiana, Dr. William Beebe collected many specimens of Psammocharidae. These, together with specimens from Cornell University and those in the Museum of Comparative Zoology, form the basis of this paper. There are but few records of these insects from this country. In 1848 W. F. Erichson published seven species in Schomburgh’s Reise in Guiana; in 1913 Peter Cameron recorded or described 23 species in the Jr. R. Agric. Soc. of Demerara — and in 1928 F. X. Williams in his Studies in Tropical Wasps recorded three species. Many of the species described by Fabricius from “Amer. merid.” were probably from this general region. In the revision of the genus Pepsis by Lucas in 1895, there are hardly any records except those of Erichson. Since so NO l/2e ' i 98 Zoologica: New York Zoological Society [XXIX: 10 little has been published and much of it many years ago it is evident that there will be many new genera as well as species. In this paper 86 species are recorded; two of Cameron have not been identified. This is probably not half that occur in British Guiana. Fifteen species appear to be new, several in new genera. But one species, N otiochares amethystina, is widely distributed in tropical South America, but nearly all the previously described species are known from localities in northern South America and Brazil ; several are also known from Peru and Bolivia, almost none from Paraguay or Argentina. Two, Pepsis ruficornis and Priochilus opacifrons, are recorded from the West Indies. In 1897 William Fox described the Bra- zilian species collected by H. H. Smith; this work has been most useful to me, and three years ago I studied this collection in the Carnegie Museum, made many notes thereon, synoptic tables of the species, and comparisons with the types. The material sent by Dr. Beebe was taken mostly at the Tropical Research Laboratory at Kartabo, but other specimens are from the Penal Settlement, Kaieteur, Kalacoon, Kangaruma, Arakaka, Mt. Ever- ard, Georgetown, and west bank of the Demerara River. The Cornell material from British Guiana is mostly from Bartica, but many from Demerara River, others from Kartabo, Kaieteur, Penal Settlement, Georgetown, Tumatumari on Potaro River, and Rock- stone, Essequibo River, mostly taken by H. L. Parish, others by Prof. Forbes and Dr. Babiy. The Museum of Comparative Zoology material is chiefly from Kartabo, taken by Dr. W. M. Wheeler, others from Bartica. H. Lang took specimens at Kama- kusa and Georgetown which were presented by Dr. J. Bequaert. Dr. N. A. Weber gave a number captured at a Forest Settlement on the Mazaruni River, and from J. Ogilvie a few were obtained taken at the source of the Essequibo River. Synopsis of the Subfamilies. 1. Labrum wholly exposed, and often as long as the clypeus ; no hairs or bristles on the basal parts of maxillae; no groove on second ventral segment; no separated hind border to the propo- deum. In the fore wing the first re- current vein ends beyond the basal third of second submarginal cell and no pocket in base of second discoidal cell; in hind wing the anal vein ends before the origin of the radial sector. Legs with very small short spines, hind tibiae never with teeth or carina above; claws cleft 2 Labrum partly hidden under clypeus, if (some Pepsis ) it is wholly visible, then the first recurrent vein ends be- fore end of basal third of second sub- marginal cell and the anal in hind wing ends as far out as origin of radial sector 3 2. Inner margin of eyes emarginate above middle; face much narrowed below; propodeal spiracle about its length from the base of propodeum; often spines on under side of last joint of mid- and hind tarsi Ceratopalinae Inner margin of eyes not emarginate; face but little narrowed below; propo- deal spiracle twice its length from the base of propodeum; no spines on under side of last joint of mid- and hind tarsi N otocyphinae 3. In fore wing first recurrent vein ends before end of basal third of second submarginal cell, and in hind wing the anal vein ends as far out as origin of radial sector; basal parts of maxillae and the mentum lack hairs or bristles; second ventral segment in females at least with a transverse groove; hind tibia of female spined and toothed above; hind basitarsus of male com- pressed; claws never cleft; under side of last joint of mid- and hind tarsi with lateral spines (at least in fe- male) Pepsinae In fore wing the first recurrent vein ends beyond the basal third of second submarginal cell, (exceptions rare), and in hind wing the anal ends before the origin of the radial sector; basal parts of the maxillae or the mentum usually with hairs or bristles 4 4. No pocket in base of second discoidal cell; second ventral segment, at least in female, with a transverse groove or impression; last joint of mid- and hind tarsi sometimes with lateral spines ; spines at tip of the hind tibia if notice- able are short and nearly parallel ... 5 A pocket in base of second discoidal cell; no groove on second ventral seg- ment; hind tibia never with teeth or a carina above; spines at tip of hind tibia moderately long and inner one divergent; usually minute spine-pits above on hind femora near its tip; last joint of mid- and hind tarsi may have median spines below but not lateral ones; female sometimes with a comb on front tarsi; claws toothed or cleft; a separated hind border to pro- podeum Psam m ocharinae 5. Last joint of mid- and hind tarsus often with spines below, usually lateral; hind tibia often toothed above or with a carina; claws toothed or cleft; hind basitarsus of male not compressed; 1944] Banks: Psammocharidae from British Guiana 99 base of abdomen rather flat at connec- tion with propodeum, the first segment may be tapering, but not petio- late Cryptochilinae Last joint of hind tarsus usually bare beneath, rarely weak spines or bristles ; claws never cleft; hind tibia never with teeth above or a carina, legs rarely hairy or bristly; no spine-pits on hind femora; abdomen attached by a dis- tinct petiole Pseudageninae The subfamily Ceratopalinae was not rep- resented in the collection. Pepsinae. At present this includes but one genus, Pepsis, although it will be divided into sev- eral in the future. Synopsis of Pepsis. 1. Wings entirely deep black, without even a pale margin 2 Wings not entirely deep black, a pale margin or bands, or mostly rufous or brown 14 2. Antennae black, at most the extreme tip sometimes rufous 3 Antennae with at least several apical joints rufous or yellowish 6 3. Propodeum very hairy and at the turn from basal to apical part with a crest of dense shaggy hair; wings not es- pecially iridescent; mesosternal tuber- cle some distance from mid-coxal slope elevata Propodeum without a crest of shaggy hair; mesosternal tubercle at begin- ning of mid-coxal slope; wings plainly iridescent 4 4. Wings beautifully iridescent violet to reddish, changing to bluish near tips and greenish in places; male with mid- dle of fourth ventral with erect hairs denser on sides and with incurving tips margarete Wings iridescent bluish 5 5. Male with a large patch of long, erect hair on middle of fourth ventral seg- ment fulgidipennis Male with middle of fourth and most of fifth ventral segment smooth and shining, but on each side a dense row of incurving hairs nireus 6. Antennae rufous or yellowish from base of third joint to tip 7 Antennae rufous from tip of third joint or beyond; mid-ridge of propo- deum low or absent 11 7. Antennae pale yellow 8 Antennae rufous or orange-red 9 8. Male with a row of hairs on each side of posterior part of fourth seg- ment luteicornis Male with a few hairs on each side of anterior part of fourth segment flavicornis 9. Body extremely hairy above and be- low, except above on abdomen; espec- ially long on propodeum; clypeus with long hairs on basal part as well as apical xanthocerus Body but little hairy, on clypeus only on lower part 10 10. Propodeum with considerable long hair; teeth on hind tibia above well separated and tips rather blunt ruficornis Propodeum scarcely at all hairy above ; teeth on hind tibia above pointed and but little separated excelsa 11. Rufous begins on the fourth or fifth joint; wings often hardly deep black 12 Rufous begins on the sixth, seventh or eighth joint. Wings somewhat brown- ish 13 12. Basal and apical parts of propodeum plainly separated by a low mid-ridge; wings with a bluish to violet irides- cence; length of fore wing about 30 millimeters ; tibial teeth rather small ianthina Basal and apical parts of propodeum not separated, no mid-ridge; wings not iridescent; fore wing about 20 milli- meters long; teeth on hind tibia very small vicina 13. No distinct mid-ridge on propodeum; five or six joints of antennae pale; third cubital cell little longer than broad sagana A low but distinct mid-ridge on pro- podeum; four joints of antennae pale; third cubital cell much longer than broad diabolus 14. Wings almost wholly rufous or yellow- ish-red, no broad dark border; anten- nae almost wholly rufous 15 Wings with at least about half brown or black 18 15. Head and thorax with golden pubes- cence; mesosternal tubercle some dis- tance before the mid-coxal slope plutus No golden hair; mesosternal tubercle usually at coxal slope 16 16. Head with very high vertex, and back of eyes is longer than width of an eye; labrum often wholly visible frivcddszkyi Head not extended back of eyes to any such extent; vertex normal 17 17. Fore wings with darkened apical bor- der hecate 100 Zoological New York Zoological Society [XXIX: 10 Fore wings not darkened apically... gracilis 18. Almost apical half of wing yellowish, basal part black; antennae black; teeth on hind tibia very small strenua Apical half not yellowish 19 19. Wings wholly brown or blackish- brown 20 Wings mostly black, but with some yel- lowish, rufous, white or hyaline 21 20. Wings brown to dark brown, about nine joints of antennae pale; a low but distinct mid-ridge niphe Wings blackish-brown; no mid-ridge no. 12, 13 21. Antennae black; fore wing marked with hyaline or white at tip or base. 22 Antennae partly rufous, two to six joints; wings yellowish on basal part 25 22. White at base of wings, and tips nar- rowly equestris No white at base of wings 23 23. A broad white apical band, fully one- fourth length of wing pretiosa A narrow apical border hyaline or whitish 21 24. At tip of both fore and hind wings; mid-ridge high grossa A spot at tip only on hind wings; mid- ridge very low iucunda 25. The yellowish band separated from base by an equally broad black band; five or six joints of antennae pale completa The yellowish occupying basal half of both fore and hind wings, scarcely any black before it 26 26. Antennae with five or six joints ru- fous; not over one-half of wings yel- lowish; male with a brush on tip of fourth ventral segment dimidiata Antennae with but two or three joints pale; about three-fourths of wing yel- lowish, male with a tuft of hairs with incurved tips on each side of fourth ventral segment f estiva Pepsinae. Pepsi's frivaldszkyi Mocz. One from sources of Rio Essequibo (J. Ogilvie). Also recorded from Para, Brazil. This species has the vertex raised and the head extended back of eyes, and the lip is wholly visible, about as in Notocyphus; mesosternum densely long-haired; meso- sternal tubercle a little before the coxal slope, low and broad, hind tibia has small triangular teeth, well separated, the spines scarcely if any longer than teeth, but the slightly curved bristles are longer. Pepsis strenua Erichs. From Kartabo, March 1; Georgetown, February 15; Forest Settlement, Mazaruni River, August. The mesosternal tubercle is at base of coxal slope, broad and low. Peps is plutus Erichs. Penal Settlement, No. Hym. 287 ; Karta- bo, March 3, June 30; Forest Settlement, Mazaruni River, August-September. Also recorded from Para, Brazil. Mesosternal tubercle about at middle of mesosternum, long before coxal slope. Pepsi's grossa Fabl'. One from Kalacoon, No. Hym. 43. Also recorded from Colombia and Brazil. Mesos- ternal tubercle a little more than half way to end of mesosternum, fairly large. Pepsis elevata. Fabl*. Kartabo, January 23, June 13. No. Hym. 256. Also recorded from Cayenne and Brazil. Mesosternal tubercle at about one-third the length of mesosternum before tip, rather large. Pepsis fulgidipennis Mocz. Bartica, August 17, June 11; Kartabo, No. 20,136; Arakaka, December 21. Also re- corded from Colombia, Venezuela, Surinam and Brazil. Mesosternal tubercle at base of coxal slope, but this is very short; prominent and almost visible from above. Pepsi's margarete Lucas. Arakaka, December 20, 21 (mostly males) ; Waratuk, February 16. Also re- corded from Panama, Colombia, Venezu- ela and Brazil and Kartabo. Mesosternal tubercle near coxal slope, not prominent. Pepsis niphe Mocz. One female from Kartabo, May 8. Also recorded from Obydos, Brazil. Mesosternal tubercle at base of coxal slope. Pepsis excelsa Lucas. Kartabo, March 1, June 30, October 8, No. 201255; Arakaka, December 20; Georgetown, February 15; Demerara River, January 5; Bartica; Kalacoon No. 52; Penal Settlement, No. Hym. 226, September 20; and Rio Mazaruni, August. Also recorded from Brazil. The mesosternal hump is visible from above, and near coxal slope the teeth on hind tibia are large, sharp pointed, and rather close together, quite different from the widely separated, smaller teeth in P. ruftcornis. 1944] Banks: Psammocharidae from British Guiana 101 Pepsis xanthocerus Dahlb. From Kartabo, May 22, July; Bartica, April 15, May 20; Kaieteur, August 4; Up- per Essequibo River (Rucker & Romitti), February 24. Also recorded from Brazil. The mesosternal tubercle is situate near the coxal slope. It is quite possible that this is the ruftcornis of Fabricius rather than the form called rujicornis by Lucas. A male from Mackenzie, Demerara River, June 22 (Cornell), is probably this species. It is black, antennae rufous from base of second flagellar joint, venation similar to female, but the third submarginal cell a little shorter. Head, thorax and propodeum very hairy as in the female. The tibia and tarsus of hind leg are compressed and fringed, the basitarsus more than in latei- cornis. On the middle of the fourth ventral segment is a mass of erect, fairly long hair, leaving the extreme hind border of seg- ment smooth; on the hind part of the fifth ventrite is a row (two or three deep) of erect hairs, about one-half as long as those on the fourth, hardly dense enough to be a “comb,” and with fewer hairs in the middle; subgenital plate rather short and broadly rounded at tip, its apical half with fine, dense appressed pubescence. Pepsis rufico rnis Fabl’. Females from Kartabo, May 3, 6, April 20. Also recorded from Cuba, Haiti, Porto Rico, Honduras, Colombia and Cayenne. Mesosternal tubercle near coxal slope. Pepsis iuteicornis Fabl'. Kartabo, March 11, May 27, June 14, July-August, No. 130,179; Arakaka, Decem- ber 21, and West bank Demerara River, February 9; all males. Also recorded from Colombia, Surinam and Brazil. Pepsis flovicornis Mocz. One male from Kamakusa. Also recorded from Brazil. Pepsis ianthina Erichs. Kartabo, June 5, July 8, August 6; Bar- tica, December 23. Also recorded from Pan- ama, Surinam and Brazil. Mesosternal tubercle a little before coxal slope, elevated, prominent, and projects a little laterally. Pepsis sagana Lucas. Kartabo, January 11, March 23, summer of 1923, No. 21104. Also recorded from Bra- zil. Mesosternal tubercle placed laterally at base of coxal slope, prominent. Pepsis vicina Lucas. Kartabo, May 28, April 14. Also recorded from Surinam. Mesosternal tubercle near coxal slope, but not as lateral as in P. sagana. The teeth On hind tibia are very small, much smaller than in sagana. Pepsis iucunda Lucas. From Kartabo, July 21, and October; Forest Settlement, Mazaruni River, August- September. Also recorded from Surinam. The male has the tip of hind wings pale as in female; the fourth ventrite has on each side an oblique, rather dense row of hairs curving toward each other, the en- closed area smooth behind, punctate in front and with some short erect hairs; fifth seg- ment punctate, more densely so on sides and with erect hair most noticeable on sides. Subgenital plate rather short, rounded at tip, with some fine hair on surface. Pepsis dimidiate Fabl’. Kangaruma, February 15; Kaieteur, Feb- ruary 18, March 24, and Bartica District. All of the specimens seen from British Guiana as well as those from French Guiana and Surinam have the yellowish band scarce- ly beyond the hyaline spot of second dis- coidal cell, while those from Colombia have this band extending plainly beyond the hya- line spot; the hair brushes of males are the same. The mesosternal tubercle prominent, tip rounded, situate at base of coxal slope, projects a little laterally. Also recorded from Panama, Colombia and Brazil. Pepsis completa Sm. From Kalacoon, No. 227. Also recorded from Brazil. Mesosternal tubercle small, at base of coxal slope, projecting somewhat laterally. Pepsis pretiosa Dahlb. Kartabo, one male, No. 131. Also recorded from Caracas, Venezuela and Brazil. Pepsis equestrls Erichs. Described from British Guiana; have seen none in these collections but many from Co- lombia, Trinidad and Brazil. Pepsis diabolus Lucas? A female from Kartabo, May 24, runs to this in the Lucas tables but does not agree very well with description. Also recorded from Brazil. Pepsis sp. near ruficorn/s. Female from Kartabo, June 7. In size and appearance much like P. rufi- cornis; however, there is a pale yellowish area in hind wing from end of basal fourth to little beyond middle; the antennae are rufous from base of fourth joint; on the propodeum the striae are fairly even and rather numerous; a small mid-ridge, and small lateral tubercles; the mesosternal tu- bercle at base of coxal slope, rather large, and projecting somewhat laterally. On hind 102 Zoologica: New York Zoological Society [XXIX: 10 tibia the teeth are larger than in ruficornis, triangular, and closer together. The color is a brighter blue than ruficornis. Pepsis hecate Mocz. A female from Tumatumari, Potaro Riv- er, June 28. Also recorded from Brazil. Mesosternal tubercle of good size, at base of slope, but the slope is longer than usual. Pepsis gracilis Lepell. One from Bartica, March 4, which I iden- tify as this species not recognized in the Lucas monograph. It was described from Cayenne. A slender black species with red- dish wings; quite hairy on head, front coxae and pleura; mesosternum with only short hair ; the tubercle at base of coxal slope, tip rounded. Striae on propodeum fairly regu- lar, no mid-ridge, lateral tubercles rather sharp and prominent; hind tibia above with small sloping teeth; not as far apart as in ruficornis, spines also sloping, little longer than teeth; third submarginal cell about as in the Lucas figure 131. Pepsis sp. near smaragdina. One from Arakaka, December 21. Pepsis nireus Mocz. A male from Forest Settlement, Rio Ma- zaruni, August-September. Also recorded from Brazil. Pepsis f estiva Fabr. Pepsis thalassina Erichs., described from British Guiana, is a synonym, but no speci- mens in these collections. Also recorded from Surinam and Brazil. Cryptocheilinae. Table of genera. 1. Claws bifid, no lateral spines to last joint of mid- and hind tarsi; basal vein interstitial or only a little before trans- verse 2 Claws toothed 3 2. Last joint of mid- and hind tarsi with a median row of spines; last segment of abdomen not compressed, antennae situate well above clypeus, distinctly slender toward tip; pronotum very short, hardly any dorsal part Priochilus No spines on last joint of mid- and hind tarsi; last segment of abdomen compressed; antennae close to clypeal margin, distinctly somewhat clavate; pronotum moderately long, the distinct dorsal part nearly or quite flat, the front part vertical or almost so Balboana 3. No lateral spines to last joint of mid- and hind tarsi ; hind tibiae with more or less distinct teeth above Priocnemis Distinct lateral spines to last joint of mid- and hind tarsi 4 4. A stout slightly curved spine or tooth in front of each mid-coxa; venter of female (sometimes male) swollen, or with raised areas, or teat-like proc- esses Priocnemioides No such spine in front of mid-coxae; venter not modified; joints of flagel- lum moderately long; clypeus trun- cate Calopompilus Priocnemis ICalicurgusI nubilus Fox. Bartica, February 12, 24, May 26; Kar- tabo, July 19, December, 1920. Priocnemis sericeiornatus Cameron is, I believe, the same species. Calopompilus vitreus Fox. Kartabo, October 30. Also recorded from Rio, Brazil. It has a much shorter pronotum than the typical North American forms and is much less hairy. Priocnemioides Radoszkowski. This was separated from Priocnemis be- cause the third antennal joint was about as short as those beyond, and the hind tibiae were said to have two rows of strong spines above. He included two described species, Pompilus fulvicornis Cresson and P. flam - mipennis Smith, and described a new spe- cies which he supposed to be from Spain. Later Schulz noted that these species had two raised, somewhat teat-like, areas on the ventral side of the second segment. The length of the third antennal joint is plainly longer in some forms, otherwise similar, and which have very distinct ventral teats. Moreover the second ventral varies in shape of the raised areas, sometimes a broad curved ridge, sometimes two flat areas; but in all these forms there is a distinct some- what curved tooth or spine in front of each mid-coxa, this is not present in the genotype of Cryptochilus, nor in genotype of Salius, nor do these genera have the modified sec- ond ventral segment. I select Pompilus fulvicornis Cresson as the genotype. There are many species in South America. The five species I have seen from British Guiana are readily separable as follows: 1. Wings black or nearly so 2 Wings yellowish 3 2. Head and thorax with golden pile aurifrons Head and thorax without golden pile bitub erculatus 3. Head and thorax with much golden pubescence gigas Head and thorax not golden, black 4 1944] Banks : Psammocharidae from British Guiana 103 4. Body more or less purple, also much of the legs purpureipes Body black, no purple, abdomen above somewhat punctate; vertex coarsely punctate perpunctatus Priocnemioides gigas Fabr, Kartabo, No. 20593; April 29, March 3, 5, Penal Settlement, Kartabo, July, No. 192. Bartica, February 11. Priocnemis croesus Smith is, I believe, a synonym. Also re- corded from Brazil. Priocnemioides perpunctatus Fox. Kartabo, April 14. Also recorded from Brazil. Priocnemioides purpureipes Cam. Kalacoon, Georgetown, November 10 ; No. Hym. 251,280. Described from Demerara. Priocnemioides bituberculatus Guer. A male from Kamakusa which I presume is the male of this species. It is not the male of mammillatus; in the latter species the subgenital plate is less broad ihan in this specimen. In both species this plate and the preceding segment are clothed with short, erect black hair; the ventral segments are punctate ; in mammillatus the punctures are as dense in the middle of the segment as on sides, in this specimen there are few or none in the middle, but plenty on each side. In the hind wing the anal vein ends at fork- ing of cubitus. The black on antennae ex- tends only to tip of third joint; the body is black, without a blue tinge; the hair below the hind basitarsus is as short as in mam- millatus. Salius tuberculiventris Cameron from Demerara is probably a synonym of bituberculatus. Priocnemioides aurifr ons sp. nov. Black, clypeus and face densely covered with golden pubescence, also, but scarcely as bright, on pleura, coxae, middle of me- sonotum, scutellum, metanotum, and over entire propodeum, less distinctly on upper side of basal segment of abdomen. Wings dark brown, with a reddish iridescence, a little darker than in tinctipennis , tip nar- rowly pale; antennae black to end of fourth joint, upper side of fifth and partially on sixth, beyond rufous to yellowish, but last three joints brown. Structure much like other species; clyp- eus rather deeply emarginate below; ocelli in low triangle, hind ocelli almost twice as near each other as to the eyes ; vertex-width almost equal to third antennal joint, latter longer than fourth; propodeum coarsely ridged in posterior part, more coarsely than in mammillatus or bituberculatus ; second ventral segment with a narrow raised area each side behind ; subgenital plate about as broad as in mammillatus, with short dense erect hair, and also on the preceding seg- ment, but others with only scattered long hairs. Front femora with some short, fine hairs above; hind tibia with rows of short spines much as other species, but those above very numerous and fine; inner spur of hind tibia about one-third of basitar- sus, latter with a longer and denser fringe than in other dark-winged species. Venation about as usual; the first recur- rent vein, however, ends further from tip of second submarginal cell than usual, the second recurrent angulate near middle, end- ing much before middle of third submar- ginal cell, the latter much longer than the second ; in hind wings the anal vein ends just beyond the forking of cubitus. Length of fore wing 20 mm. One male from Demerara River, March 20 (Cornell). Balboana fulvipes sp. nov. Head, thorax, propodeum, and abdomen dull black; tip of abdomen and sides of first segment fulvous; antennae black, but basal joint fulvous below; legs fulvous, coxae also, tips of tarsal joints dark, spurs white; palpi pale. Fore wings hyaline, with two narrow black bands, one over basal and transverse veins, other occupying base of marginal, all of second and third submar- ginal, and upper part of third discoidal cells; hind wings hyaline. Clypeus about four times as broad as long, slightly convex below, surface with short hair; face scarce- ly narrowed above, ocelli in a broad triangle, lateral ocelli about as close to eyes as to each other; antennae plainly a little thickened toward tip; third joint not longer than fourth; pronotum arcuate behind; propod- eum finely transversely rugulose, no median groove ; abdomen rather slender, compressed at tip, last two segments hairy, venter with a few erect hairs. Head, thorax and abdo- men mostly covered with very short, ap- pressed white hair. Legs smooth on femora and tibiae, hind tibia has about eight or ten short fine bris- tles above, one or two on the sides, inner spur nearly equal to one-half of the basi- tarsus. Fore wings rather short, but slender, the marginal cell long, widest at end of second submarginal cell, from there sloping to acute tip, not one-half its length from wing- tip; second submarginal cell about as high as long, both ends oblique, receiving the first recurrent at basal third ; third submar- ginal fully three times its length from outer margin, a little longer than the second, nar- rowed about one-third above, outer side very oblique, receiving the second recurrent (strongly bent outward at middle) at about middle; lower basal vein slightly bowed for- 104 Zoologica: New York Zoological Society [XXIX: 10 ward near end, meeting end of transverse, which is not oblique, in hind wings the anal ends much before forking of cubitus. Length of fore wings 6 mm. One from Kartabo, July-August. Much smaller, and with less spiny legs and the dorsal part of pronotum longer than in typical Balboana, but agreeing in the clavate antennae, compressed tip of abdo- men, antennae at clypeal margin and vena- tion. There are several other species of this genus in South America. Balboana auripenni s Fabl'. Kartabo, April 16, June 1, November 4, March 13, 17, September 26; Bartica, De- cember 2. Also recorded from Brazil. The description of Pompilus moorei Cam. from Demerara fits auripennis. Priochilus. This is agenoid in form, but larger and readily separated by the cleft claws and the median row of spines below on last joint of mid- and hind tarsi ; in the smaller species these spines may be reduced to one or two. Moreover in the larger specimens there are one or two very small but distinct spine-pits on the hind femora toward tip. This with the tarsal armature would place the genus in the Psammocharinae, but there is no real “pocket” in the base of third dis- coidal cell. The hind tibiae have larger spines than in Pseudagenia and allies, mostly in one or two rows, in or near a carina. The venation is much like those Pseudogenias with a long third submarginal cell, but it is wider at tip; in hind wing the anal ends at or very near forking of cubitus; the basal segment of abdomen is not petiolate, and no distinct groove on the second ventral segment, but the segment is transversely impressed where the groove would be if present. Type, Pompilus nobilis Fabr. Includes at least fifteen or twenty neo- tropical species which were placed in Agenia or Salius. Females. 1. Fore wings black 2 Fore wings pale, clouded or banded with dark 6 2. Face, thorax, propodeum and abdomen with much silvery, propodeum hairy behind 3 Face silvery, little elsewhere 4 3. Mesosternum projecting laterally in an almost cone-like hump; carina of hind tibia shghtly undulate, upper spines very short; mid-tibia enlarged and with many short spines, second submarginal cell much longer than high superbus Mesosternum not projecting cone-like; carina of hind tibia not undulate, some upper spines nearly as long as width of joint, mid-tibiae slender, with fewer and longer spines ; second submarginal cell only a little longer than high ... regius 4. Second submarginal cell only a little if any longer than high, both ends ob- lique, propodeum hairy, face somewhat silvery rhomboideus Second submarginal cell plainly longer than high, ends less oblique 5 5. Propodeum with long black hair above ; elypeus very hairy all over, margin not polished; second submarginal cell al- most twice as long as high, third sub- marginal very wide at tip; anal in hind wings ends beyond forking of cubi- tus imperhts Propodeum with only very short fine hair, hardly noticeable; elypeus with only a few hairs in a row, margin pol- ished, shining; second submarginal cell about one and one-half times as long as high, third submarginal cell not so wide at tip; anal in hind wings at forking of cubitus plutonis 6. Abdomen yellowish-brown ; propodeum not hairy above; outer dark band on fore wings very broad, inner band extended basally . formosus Abdomen black 7 7. Propodeum hairy above, narrowly sil- very at base, broadly behind ; wing hy- aline between the bands, large spe- cies ..nobilis Propodeum not hairy, and little if any silvery, second submarginal cell often higher than long 8 8. Propodeum transversely striate be- hind; front with somewhat golden pu- bescence sericeifrons Propodeum not striate 9 9. No silvery nor golden at sides of scu- tellum; front opaque black, small spe- cies o pad f rons A silvery spot at each side of scu- tellum 10 10. Fore wings yellowish between the bands, no golden pubescence on front or notum scrupidus Fore wings hyaline between bands, some golden pubescence on front and notum di versus Priochilus regius Fabr. Kartabo, March 4, April 28, August, Nos. 146, 218, 1929, 20213, 20462, 201297; Ka- makusa, September, Forest Settlement, Rio Mazaruni August; Kalacoon; February 8, January 21, March 4; Penal Settlement, Oc- 1944] Banks: Psammocharidae from British Guiana 105 tober 7 ; Demerara River, March 10, 18. Also recorded from Brazil. Priochilus rhomboideus Fox. Bartica, March 5, May 8, 21, 24; Bartica, January 25; Arakaka, December 21; Turesi Falls, October 11; Mt. Everard, November 15; Georgetown, November 10, 15; Deme- rara River, March 8, 24. Also recorded from Brazil. Priochilus nobilis Fabr. Kartabo, April 4; Bartica, February 18, April 15; Tropical Research Laboratory. Also recorded from Brazil. Pompilus cos- mopteryx Cam. appears to be a synonym. Priochilus scrupulus Fox. Kartabo, August, Nos. 20558, 21139. Also recorded from Brazil. Priochilus diversus Sm. Kamakusa, September. Also recorded from Brazil. Priochilus sericeifrons Fox. Bartica, January 28, February 4. Pom- ■pilus harperi Cam appears to be this spe- cies; it is from Demerara. Also recorded from Brazil. Priochilus opacifrons Fox. Bartica, January 29, February 5, March 5, April 3, 14, July, Hym. 306. December 23; Demerara River, March 19, 20, 26, 29, Mackenzie, Demerara River, June 23. De- scribed from island of Jamaica, but these specimens appear to be the same and Fox records it from Brazil. Priochilus imperius sp. nov. Black throughout, no silvery on face or on the coxae; wings with violaceous and bluish reflections. Face and vertex rather more hairy than other species; propodeum with long hair, abdomen above hairy near tip, venter hairy on last four joints. In gen- eral structure similar to other species; an- tennae very slender, third joint much longer than width of vertex; ocelli in a low tri- angle, laterals nearer each other than to eyes; pronotum broadly arcuate behind; metanotum finely striate; propodeum even- ly convex, pleura hairy; legs slender, not hairy; hind tibia with earina above which is faintly cut into sections, hardly teeth, the bristles not one-half diameter of hind tibia; last joint of hind tarsus with a me- dian row of four or five rather large spines below; long spur of hind tibia not one-third of basitarsus. In fore wings the marginal cell is very long, hardly one-half its length from wing- tip; second submarginal cell longer than high, its basal side more oblique, receiving the first recurrent at about one-fourth from tip; third submarginal cell one-half longer below than the second, very much widened at tip, narrowed one-third above, the outer vein almost angulate in middle, receiving the second recurrent (faintly sinuous) near end of basal third. In hind wings the anal ends close to the cubital fork, and the cross-vein is beyond the middle of radial sector. Length of fore wing 22 mm. Females, one Kamakusa (Lang) (Be- quaert coll.) and other Kartabo, June 24 (Beebe), both British Guiana. It will not fit pilifrons Cam. (which does not belong to this genus) nor dives Lepell. (which may be a Priochilus), and is not in the H. H. Smith collection of Carnegie Museum. Others from Kamakusa; Moengo Boven Cottaca R., Surinam, May 14; Rio Itaya, Amazon, Peru, December; Guayaquil, Ecua- dor, May-June (Brues) ; Santa Cruz, Bo- livia (Steinbach) ; La Sombra, Peru. Priochilus superbus sp. 110V. In general very similar to P. regius, us- ually larger. Clypeus and face one-half way up to ocelli silvery; collar and lower sides of pronotum, propodeum above, apical half of basal abdominal segment, basal third of second, coxae and apex of the mesopleura strongly silvery. A spot each side at base of third segment, and more or less on foiuTh and fifth segments less strongly silvery. Wings deep black. Structure in general much like P. regius, ocelli in a broader triangle, third antennal joint proportionally not as long as in regius; propodeum slightly, evenly convex, hairy, especially behind; the mesosternum in front of mid-coxae projects outward in a prominent cone (only bulging slightly in regius) ; venter more hairy than in regius. The mid-tibiae are slightly enlarged, and with more numerous and shorter spines than in regius. In the fore wings the sec- ond submarginal cell is plainly longer than high, longer than in regius, otherwise ven- ation similar. Length of fore wing 15 to 18 mm. Holotype from Kamakusa, September. Paratypes from Villavicenc-io, Co’ombia; Chounazu, Peru, July 9; and several from Iquitos San Rogue, Peru, February, March, Mav; Santa Cruz, Bolivia. The slighPy enlarged, more spinose mid- tibiae, as well as the swollen mesosternum near mid-coxae, readily distinguish it. Priochilus plutonius sp. nov. Black, no silvery, wings somewhat viola- ceous. Clypeus short, very broad, truncate below, the margin polished and shining, a row of long bristles across clypeus, otherwise without hair; face bare, front with two long erect hairs each side near orbits ; ocelli 106 Zoologica: New York Zoological Society [XXIX: 10 in a moderately narrow triangle, hind ocelli little more than diameter apart, much far- ther from the eyes; antennae slender, the third joint very long, longer than vertex- width; pronotum with only a few short, fine hairs, angularly emarginate behind; mesonotum with one or two erect hairs each side. Propodeum without hair, opaque; tip of abdomen hairy on last few segments, above with some stiff bristles; legs long and slen- der, mid-tibiae with two rows of spines above, the longest about one-half the diame- ter of the joint, hind tibiae with a row of very short spines above, but no distinct carina, some spines on outer side one-half diameter of joint, long spur not one-third of basitarsus. In the fore wings the mar- ginal cell is very long and slender, sharp pointed at each end, and less than one-half its length from wing-tip; second submar- ginal cell hardly one-half longer than high, each end oblique, receiving the first recur- rent vein at about one-fourth from tip; third submarginal cell about as long above as second, much longer below, outer side sloping, receiving the second recurrent (al- most bent in middle) at about basal fourth; lower part of basal vein straight, ending a little before transverse vein, latter not oblique; in hind wings the anal vein ends at the cubital forking. Length of fore wing 13 mm. A female from Demerara, British Guiana, March 10 (Cornell). Priochilus formosus sp. nov. Black; abdomen brownish-yellow or rusty brown ; wings hyaline, a broad brown band out from basal vein and behind the cubitus extending back almost to base of wing; a still broader brown band over most of marginal cell, over both second and third submarginals, and most of the third dis- coidal cell; hind wings with the tip brown. Face and clypeus only faintly silvery, also faintly behind on propodeum, coxae more strongly silvery. Clypeus shaped much like that of regius, broadly truncate below; face a little nar- rowed above, third antennal joint equal to vertex-width, and much longer than the fourth joint; median groove reaching anter- ior ocellus, hind ocelli hardly more than diameter apart, much farther from the eyes, a few long hairs in a row across clypeus, and two erect bristles each side by upper orbits, mesonotum and scutellum also with few hairs, more on the propodeum. Pronotum broadly arcuate behind; propo- deum with a rather broad median groove, from side propodeum shows an even slope, hardly convex. Basal segment of abdomen from above with almost straight sides; tip of abdomen above and below with fine hairs, some above are stiff. Legs moderately slender; hind tibia with the longest spines about one-half diameter of the joint, inner spur not one-third of basitarsus. Fore wings with marginal cell long, at widest no broader than second submarginal, nearly equally pointed at each end, hardly half its length from wing-tip; second sub- marginal one and one-half times as long as high, both ends oblique, receiving the first recurrent a little before tip, third sub- marginal scarcely longer than second above, but longer below, receiving the second re- current (slightly sinuous) at before end of basal third; lower part of basal vein straight, ending scarcely before the trans- verse, which is hardly oblique; in hind wing the anal vein ends much before the forking of cubitus. Length of fore wing 13 to 15 mm. Holotype from Kamakusa, British Gui- ana, paratype from El Campanlentus, Col. Parene, Peru, July 1 (Cornell). It differs in many points from abdominalis Sm. PSEUDAGENINAE. Table of genera. 1. Mid- and hind tarsi with bristles and spines on lateral edges of last joint; hind tibia with only small spines above, and indistinct carina Phanochilus Mid- and hind tarsi without spines under last joint or not laterally 2 2. Hind tibia with teeth above (some- times very small) ; antennae situate hardly diameter of basal joint above clypeus 5 Hind tibia without teeth 3 3. Propodeum without hair; rarely a dis- tinct “beard,” sometimes a few short hairs; claws toothed Ageniella Popodeum more or less hairy; usually a distinct “beard,” several long stout bristles 4 4. Usually a carina above on hind tibia, and one or more rows of short spines; antennae situate rather high above clypeus; anal vein of hind wings ends beyond or at forking of cubitus Priocnemella No carina, few spines on hind tibia; in hind wings anal vein ends before the forking of cubitus Pseudagenia 5. A distinct “beard” under the head Alasagenia No “beard” under head Priophanes Alasagenia gen. nov. Abdomen with distinct petiole; venation as in Pseudagenia; a beard of six or more 1944] Banks : Psammocharidae from British Guiana 107 rather long hairs under head; in a trans- verse row last joint of mid- and hind tarsi bare beneath; claws toothed; basal vein ends before the transverse; in hind wings anal vein ends before forking of cubitus; hind tibia with a row of distinct teeth above, and on each side a row of short bristles. Type A. erichsoni sp. nov. Some species from the United States have been put in Priocnemis, but differ in the petiolate abdomen. Alasagenia erichsoni sp. nov. Body, antennae and legs deep black, cly- peus and sides of face strongly silvery, a silvery spot each side on hind border of pro- notum, also the hind border of mesonotum, sides of metanotum, a large square silvery spot each side on posterior slope of propo- deum and the base also silvery ; also silvery spots on the pleura, front coxae, part of basal abdominal segment, a large spot each side on second segment, basal half of third, and less distinctly across fourth and fifth. Fore wings hyaline, a narrow dark band over basal and transverse veins, a broader band beyond stigma, but not quite reaching across third discoidal cell, tip plainly brownish; in hind wing the tip is smoky. Head, thorax, propodeum with much fine pale hair, and some longer, erect ones. Clypeus nearly three times as broad as long, lower margin evenly, but slightly, con- cave, rather closely parallel to upper edge; face only a little narrowed above, vertex- width hardly as long as second plus third antennal joints; hind ocelli about equally near each other and to eyes; pronotum slightly angulate behind; propodeum rath- er low, evenly convex, no distinct dorsal groove; basal abdominal segment with a little short hair on petiole above, some fine hair on last few segments, and a few on venter. Legs with femora moderately stout, hind tibia above with a row of distinct teeth and an inner carina, the spines extremely short; inner spur of hind tibia not two-fifths of basitarsus. In fore wings the marginal cell is long, not as broad as the second submarginal, less than two-thirds of its length from wing-tip; second submarginal cell but little longer than high, not longer above, receiv- ing the first recurrent vein at apical third ; third submarginal about twice as long as second, the outer side oblique and rather strongly curved on lower third, receiving the second recurrent (much bent near mid- dle) at middle of cell, the cell about two- thirds its length from outer margin; basal vein ends much before the transverse, lat- ter a little oblique; in hind wing the anal vein ends at forking of cubitus. Length of fore wing 10 mm. From Kartabo, February 28. Named for W. F. Erichson, who described many in- sects from British Guiana. Priophanes fabricii sp. nov. Head black, clypeus yellow, face with short white pile, most noticeable at orbits; thorax and propodeum rufous; abdomen shining black, second, third, and fourth segments with a large spot of silvery pu- bescence each side; antennae black; legs yellowish, femora somewhat brown above, hind tibiae rather darker, coxae and tro- chanters black. Fore wings hyaline, stigma brown, a narrow brown band over basal and transverse veins, a broader brown band beyond stigma, but not extending behind third discoidal cell, tip of wing slightly brown. Hind wings not marked. Clypeus about three times as long as broad almost pointed at each end, upper and lower edges nearly equally evenly convex; palpi espe- cially third joint, very long. Face broadest in middle, more narrowed above than be- low, vertex-width longer than second plus third antennal joints; hind ocelli almost as near eyes as to each other; pronotum slightly angulate behind; propodeum even- ly convexly sloping, posterior half with some long pale hairs, no median groove. First segment of abdomen with pendent sides, tip with very fine hair, on venter a few short hairs; legs with femora very smooth, the hind tibiae with very minute, hardly visible teeth, the spines very short, black, three rows of them above, mid-tibia similar, inner spur of hind tibia fully two- fifths, but not one-half, of basitarsus. In the fore wings the marginal cell is moderately long, about two-thirds its length from wing-tip, equally broad at ends of sec- ond and third submarginal cells; second sub- marginal cell nearly one and a half times as long as broad, both ends oblique, receiv- ing the first recurrent vein before middle; the third submarginal about one-half its length from outer margin, above fully as long as the second submarginal, below one- third longer, outer end not strongly oblique, receiving the second recurrent (scarcely curved) before end of basal third ; basal vein ends much before the transverse; in hind wing the anal vein ends before forking of cubitus. Length of fore wing 6 mm. Type from Kartabo, July-August. Two paratypes, one St. Augustine, Trinidad, March 23, other Restrepo, Dept. Meta, Co- lombia. The one from Trinidad has the hind ocelli a little nearer to each other than in type, and the Colombian specimen has the hind ocelli a little nearer eyes than in type. Named for J. C. Fabricius, who first de- scribed various species from this region. 108 Zoologica: New York Zoological Society [XXIX: 10 Phanochilus gen. liOV. These have the general appearance of a large Pseudagenia or Priochilus, the basal abdominal segment being plainly petiolate and the venation is much like them; the basal vein is more separated than usual in Pseudagenia. In the hind wing the anal vein ends near or at forking of cubitus. The hind tibia may be toothed, or with a low carina which is broken by a row of very short spines; the last joint of mid- and hind tarsi have a few spines or bristles near each margin ; the claws are toothed ; the clypeus is often somewhat hexagonal ; the head seen from above is rather quadrate, being much more swollen behind than usual ; the meso- pleura projects laterally, into a large cone- like swelling, readily seen from above. Type Agenia nobilitata Smith. Phanochilus nobilitata Sm. Tropical Research Station No. 21265. Have also seen it from Rio Itaya, Amazonas, Peru. The Salius f us comar g hiatus Fox is very similar, but a heavier insect and the malar space is much broader. Phanochilus gloriosa Sm. One male from Kartabo, July 16. Also recorded from Brazil. Salms aureodecoratus Cam. described from Demerara, appears to be. this species. I see no difference between this species and the male of nobilitata ex- cept the dark cloud in wing of gloriosa. Phanochilus pilitrons Cam. Described from Demerara; specimens examined are from Kamakusa, September, Bartica, January 31, and Rio Essequibo ( J. Ogilvie). Ageniella. Females. 1. Body wholly yellowish; wings hyaline, with a large black spot at tip reversa Body not wholly yellowish, wings with two dark bands 2 2. Thorax reddish, legs black, apical part of antennae pale rnicans Thorax, black, antennae partly yel- lowish delila Ageniella micans Fabl’. Tumatumari, Potaro River, June 28, one male. The Pseudagenia pulchricornis Cam. described from Demerara is this species; I have also seen it from the Matto Grosso of Brazil. Ageniella reversa Sm. Tumatumari, Potaro River, June 22. Also recorded from Brazil. Ageniella delila sp. nov. Body, thorax, legs, abdomen dull black; antennae with basal joint pale beneath, beyond black to middle of sixth joint, thence yellowish ; a small silvery to golden spot each side at end of propodeum. Wings hyaline, fore wings with a moderately broad straight brown band across over basal and transverse veins; another brown band over basal part of marginal, most of second submarginal, basal part of third submar- ginal, and apical part of third discoidal cells, no broader than the other band, tip of wings scarcely fumose; hind wings un- marked. Clypeus about three times as broad as long, lower margin evenly rounded; ocelli in a small triangle, laterals much nearer to each other than to eyes; pronotum scarcely angulate behind; propodeum low, no median groove; abdomen petiolate, basal segment very slender just before attach- ment. No hairs on body except at tip of abdomen. Legs slender, smooth, mid- and hind tibiae with only extremely minute bristles above, inner spur of hind tibia about two-fifths of basitarsus. Fore wings with marginal cell broad at end of second submarginal cell, fully twice as broad as the second submarginal, about two-thirds of its length from the wing-tip; second submarginal cell very small, slightly broader at base which is about two-thirds of its length, receiving the first recur- rent vein at middle; third submarginal cell longer above than second below, and belo-vy twice as long as second, less than its length from outer margin, receiving the second recurrent (bent below middle) be- fore end of basal third; basal vein ends a short distance before transverse; in hind wing the anal vein ends much before fork- ing of cubitus. Length fore wing 7.6 mm. From Kartabo, July-August. The male has a more slender body; is similar in color except that the antennae is yellowish, with the fifth, sixth, and seventh joints dark. The clypeus is very broadly and deeply excavate on the lower margin and each end projects as a tooth; its surface has much short hair. The structure is also very similar to the female, but the basal abdominal segment is more slender, and with pendent sides; the mid- and hind tibiae have very minute smnes on sides and above, the inner spur of hind tibia is a little longer, but does not reach middle of Basitarsus; there are no hairs at tip of ebdomen. The venation is also similar, but the marginal ce'l is not quite so broad, and the second submarginal broader, but no longer than in female. Length 7 to 8 mm. 1944] Banks: Psammocharidae from British Guiana 109 From Demerara River, March 10, 12, 18 (Cornell) . PSEUDAGENIA. Females. 1. Mid- and hind femora black, mid-tibia rufous; clypeus hardly rufous below; pygidium black incrota At least hind femora reddish 2 2. Only hind femora reddish, thorax black com par at a Other femora at least partly reddish or yellowish 3 3. Thorax plainly bluish or greenish, metallic 4 Thorax black above, dull; mid- and hind tibiae black or dark; pygidium black smithi 4. Mid- or hind tibiae more or less pale, thorax somewhat greenish; propo- deum rather coarsely granulate, and on middle of basal part plainly transversely striate; pygidium dark tarsata Mid- and hind tibiae dark; propo- deum only finely granulate, punctate behind, not striate above 5 5. Basal joint of antennae pale below; third antennal joint rather short, body rather slender; pygidium pale sp. Basal joint of antennae black below 6 6. Thorax and propodeum more greenish; front and mid femora with much dark above especially toward base, pygidium brownish; propodeum with shorter hair, body rather slender sp. Thorax plainly bluish, femora wholly rufous, pygidium pale, propodeum with long white hair, body broader femorata Two species I have not been able to iden- tify, nor can I be sure they are new. Pseudagenia femorata Fabr. From Demerara River, March 12, 22 ; Kartabo, March 10; No. 1925, May 7, No. 1925; Bartica, January 29, February 1. Also recorded from Brazil. The basal vein ends plainly before trans- verse, latter scarcely oblique; third submar- ginal cell wider at tip than in tarsata; pro- podeum not coarsely granulate as in tarsata. Pseudagenia smithi D. T. Pseudagenia semisuavis Cam. from Demerara, is probably a synonym of this species. I have none from British Guiana but from northeastern Colombia, and it is recorded from Brazil. Basal vein nearly interstitial with transverse, latter oblique. Pseudagenia tarsata Sm. From Kartabo June 9, August 17. Basal vein ends only a trifle before the transverse, latter very oblique. Pseudagenia modesta Sm. Male from Bartica, February 10. Also recorded from Para, Brazil. Pseudagenia comparata Sm. From Kartabo, Tumatumari, March 10, Nos. 20419; 1925; June 28. Also recorded from Para, Brazil. Pseudagenia basalis Sm. Male from Bartica, February 26. Also recorded from Santarem, Brazil. Pseudagenia incrota sp. nov. Black, clypeus and lower face with silvery pubescence, antennae dull yellowish, last few joints brown, basal and ringjoint nearly black ; lower parts of pronotum silvery, and a silvery spot each side behind, four silvery patches on propodeum, the two at hind border the larger; abdomen with some fine whitish hairs near tip; wings hyaline, stigma black, venation brownish, tip of fore wing faintly brown; legs brown to nearly black, the front pair rufous to yellowish beyond femora, the mid-tibiae rufous, tarsi dull black, hind legs wholly dull black. Clypeus rather short, broadly rounded below; antennae rather slender, second plus third joints not equal to vertex width, hind ocelli nearer to each other than to eyes; pronotum broadly concave behind; propo- deum evenly convex, median line very faint, mesosternum laterally bulging, but not sharply so ; abdomen with fine bristles near tip; hind tibiae with a few minute bristles above, a median and shorter lateral rows, long spur about one-half of basitarsus. In fore wings the marginal cell is much less than its length from the wing-tip, second submarginal cell not twice as long as high, receiving the first recurrent vein a little beyond the middle, third submarginal about one-half longer than second, and much wider at tip, receiving the second recurrent a little before middle, basal vein ends plainly before the transverse; in hind wings the anal vein ends much before cubital fork. Length of fore wing 6 mm. Females from Kamakusa (Lang, Be- quaert), and Bartica, January 17, 1913 (Cornell), both British Guiana. Priocnemella eurytheme sp. nov. In general coloration, the short fourth antennal joint, the short inner spur of hind tibia, the shape of clypeus, and golden pubescence this species is close to Salius hexagonus Fox. The golden pubescence, however, is more extensive; it is on face, the clypeus when viewed from above, golden band on hind margin of pronotum, 110 Zoologica: New York Zoological Society [XXIX: 10 golden on mesonotum, middle spot on metanotum just behind the scutellum, also basal band and two large apical spots on the propodeum. A spot on pleura is golden to silvery, and there is a silvery sheen on sternum and coxae. The lower margin of clypeus is about equally three-sided, the middle portion trun- cate; fourth joint of antennae much shorter than the third, and shorter than the joints beyond ; face moderately narrow, orbits near- ly parallel, hind ocelli much closer to each other than to eyes; pronotum arcuate be- hind; scutellum convex; propodeum about as broad as long, somewhat narrowed be- hind, no median furrow, from side evenly convex, at the turn are a few rather large transverse ridges, from side some very fine and very short hairs are scarcely noticeable. Abdomen steel blue, metallic, some long hairs toward tip, venter with much long hair, and patches of dense and very short hair; legs black, front tarsi faintly brownish; hind tibiae with a row of very short bristles arising from a very low carina, but no distinct teeth; long spur of hind tibiae little more than a fourth of the basitarsus. Fore wings with marginal cell long and sharp-pointed, widest at end of second sub- marginal cell, latter a little narrowed towards tip, base oblique, almost one and one-half times as long as broad, receiving the first recurrent a little before tip ; third submarginal about twice as long and much wider at tip, narrowed one-third above, receiving the second recurrent vein (slightly curved) at about middle. Length of fore wing 15 to 17 mm. Two females, one from British Guiana, other Tropical Research Station at Kartabo, No. 117. Related to hexagonus Fox which probably belongs to this genus. Priocnemella difiormls sp. nov. Black throughout; head and thorax with bluish iridescence; wings somewhat viola- ceous; abdomen shining blue-black; clothed heavily with moderately long black hair on head above and below, pronotum, propo- deum, pleura, all coxae, and venter of ab- domen, a few on basal part of first segment above, on last segment with some long hairs, and with short hair above. Clypeus punctate, lower edge slightly convex, not three times as broad as long; face slightly narrowed above, antennae slender, second and third joints together equal to vertex-width, ocelli in a low triangle, hind ones very much nearer each other than to eyes, the beard is a small group of slightly upcurved bristles; prono- tum short, slightly angulate behind; meta- notum elevated in middle and with erect hairs, a few on hind part of scutellum, sides of metanotum not striate; propodeum rather low and evenly curved, from above as broad as long, with faint median groove. Basal segment of abdomen rather long, petiolate, the petiole with concave sides. Legs slender, hind tibiae almost without spines, above, only a few extremely minute ones nearer outer side, a low carina near inner edge, inner spur hardly more than one-fourth of the basitarsus; mid-tibiae with some stouter, more distinct spines. Fore wing with the marginal cell long, about two-thirds its length from wing-tip, hardly broader than the second submarginal cell, latter one and one-half times as long as broad, base curved, oblique, receiving the first recurrent vein near apical third; third submarginal almost twice as long as second below, at widest about twice as broad as second, narrowed one-third above, the outer side strongly oblique and faintly sinu- ous, its tip about one-third its length from outer border, receiving the second recur- rent (slightly curved) at middle; basal vein but little oblique, ending much before the transverse; in hind wing the anal vein ends at or just beyond forking of cubitus. Length of fore wing 16 mm. Holotype from Kamakusa, paratypes from Iquitos San Rouge, Peru, April; and Buenavista, Santa Cruz, Bolivia. In appear- ance much like Phanochilus pilifrons Cam. and Priochilus imperius but differing very much in characters of mid- and hind tibiae and last tarsal joint. It is larger than in- dicated by Fabricius for captivus, and he does not mention bluish body nor violace- ous wings. PSAMMOCHARINAE. The genera can be tabulated as follows: 1. With a line or sigma on middle of pos- terior part of pronotum 2 No such sigma 3 2. Posterior part of propodeum humped each side A rachn o ph roc tonus Posterior part of propodeum evenly rounded Batazonus 3. Pronotum elongate, front femora thick- ened, abdomen compressed, but two sub- marginal cells, no spines under last joint of mid- and hind tarsi Planiceps Pronotum not elongate 4 4. Clypeus emarginate in middle 5 Clypeus not emarginate 6 5. Fore wings folding lengthwise; no spines under last joint of mid- and hind tarsi; thorax without hair Aplochares Fore wings not folding lengthwise; spines under last joint of mid- and hind tarsi Notiochares 1944] Banks: Psammocharidae from British Guiana 111 6. No spines under last joint of mid- and hind tarsi; tip of abdomen with a few fine hairs; female with a tarsal comb; male with claws cleft Sericopompilus Spines under last joint of mid- and hind tarsi; tip of abdomen with some stiff bristles 7 7. Propodeum bare; marginal cell short Pompilinus Propodeum hairy 8 8. Female with a tarsal comb Psammochares Female without tarsal comb Anoplius Batazonus decedens Sm. From Kartabo, No. 2121, February 26, March 4, June 3. Also recorded from San- tarem and Para, Brazil. Batazonus fervidus Sm. Kartabo, March 24, June 30, No. 24688; Bartica, February 4. Also recorded from Brazil. Batazonus polistoides Sm. Kartabo, Nos. 1920(2); 282, 20945, May 14, 29; Rio Mazaruni, Forest Settlement, September, Demerara River, February 18, March 8, 19; Bartica, January 16, 27, Feb- urary 12. Also recorded from Brazil. Arachnophroctonus crassidentatus Cam. Described from Demerara. One from Kar- tabo, June 11, appears to be the same as A. vulpes Fox. Psammochares inculcatrix Cam. Described from Demerara; specimen be- fore me from Surinam; it is close to and perhaps identical with more southern P. triquetrus Fox. Psammochares echinatus Fox. Kartabo, No. 22136; July 23, October 10, May 18, March 27; Bartica, January 29, April 14; West Bank Demerara River, Feb- ruary 9. Also recorded from Brazil. The male is structurally very close to P. oma- mentus, and with the fore wings entirely black. Psammochares o rnamentus Fox. Bartica, February 11, 24, 26, January 28; Kartabo, May 27, July-August; Essequibo River, June 26. Also recorded from Brazil. The size of the pale area in middle of wing varies, and so might be a variety of echina- tus; there are slight differences in venation which I doubt will be constant; all speci- mens are males. Notiochares amethystina Fabr. Kartabo, August 5, 1920 (2 spp) ; July 29, 1920; September 1, 1922; April 28, 1919; Penal Settlement, Nos. 320, 329; Tropical Research Station, Nos. 2045, 2048, 20306, 20307, 21105; Bartica, May 21, 1924. This species is widely distributed from Paraguay to Colombia, and there are numer- ous specimens in the collection from British Guiana. Cameron’s record of imperialis doubtless refers to this common species. In the forked subgenital plate and the short band of velvety hair it is extremely similar to N. philadelphica of the United States. A variety, less blue and covered with sericeous pile especially prominent on the propodeum, from Arakaka, does not differ from the typ- ical in other characters. Aplochares gen. nov. Clypeus elevated, convex, lower margin with a deep rounded emargination in mid- dle; female with a comb, spines slender; mid- and hind tibia with slender spines; claws toothed ; last joint of tarsus below with a few teeth in middle; wings tend to fold lengthwise as in Episyron; marginal cell not its length from tip of wing, basal vein interstitial with the transverse, thorax, pleura and propodeum without hair; head bare, except two or three erect hairs each side on vei'tex; abdomen with fine hairs at tip; antennae slender. Type Pompilus imi- tator Smith. Differs from Lophopompilus and Notiochares (which also have emargi- nate clypeus) by absence of stiff bristles at tip of abdomen, the folding wings and ab- sence of hair on much of body. Aplochares imitator Sm. One from Kartabo, October 11. Also re- corded from Brazil. Sericopompilus exilis sp. nov. Male. Black with much sericeous pubes- cence especially on face, clypeus, pronotum and propodeum; pronotum with white band behind; abdomen with whitish spot each side on base of second and third segments, and last dorsal segment also white; legs mostly black; front tarsi paler, hind tibia with a whitish streak on basal half above, mid- and hind spurs white, front ones more rufous. Wings hyaline, fore wings with dark band over basal and transverse veins, and a broader one occupying all but the base of marginal cell, apical half of second submar- ginal, apical part of third discoidal, and extending to tip of wing, where, however, it is much less dark than the beginning; hind wing with extreme tip dark; antennae often somewhat rufous toward base. Lateral ocelli about as close to eyes as to each other; antennae thick, short; not reach- ing tip of scutellum ; hardly any erect hairs on head, thorax, or propodeum, only the ap- 112 Zoologica: New York Zoological Society pressed pubescence; propodeum with medi- an furrow ; abdomen compressed particu- larly toward tip, no long hairs at tip; mid- and hind tibiae with a few moderately long spines above, and between the two main rows there is a more or less complete row (sometimes but two or three) of very much smaller spines, inner spur of hind tibia three-fourths as long as basitarsus, claws cleft. In fore wing the marginal cell is pointed, and hardly its length from wing-tip; but two submarginals, no trace of third, the second fully half the length of the marginal cell below, but narrowed each end above, receiving the first recurrent plainly before middle, and the second recurrent (straight) a little before tip; basal vein interstitial or nearly so with the transverse. In hind wing the anal vein ends much before fork- ing of cubitus. Length of fore wing 6.5 to 7 mm. Type from Bartica, British Guiana, March 20 (Cornell) ; paratype, Zanderij, Para district, Surinam, April 27. Planiceps pertyi sp. nov. Black; head and thorax with fine short pale pile which on hind part of pronotum, posterior part of mesonotum, and a broad streak each side on propodeum is yellowish, in some views almost golden. The second, third and fourth segments of abdomen have a pale yellowish spot each side, and the last segment has a median streak, almost white. The fore wings are yellowish, the hind pair less distinctly so; veins, except costal, mostly pale ; in fore wings is a brown mark over basal vein, extending almost half way out in the second discoidal cell ; another broader brown band occupies all but base of marginal cell, half of second submarginal and a like distance beyond, the two bands separated by more than the width of the outer band; tips of both wings faintly brown. Clypeus proportionally a little longer than in herberti; front ocellus in a deep pit; hind ocelli about as near eyes as to each other; vertex fully as broad as the third plus fourth joints of the antennae; prono- tum long, sides rounded, with distinct scat- tered punctures above; propodeum more slender than in herberti; abdomen very slender, with a few very fine hairs toward tip, and some along venter, mostly toward tip, front femora very broad. Venation of fore wings much as in her- berti, but the basal vein rather more bulg- ing basally; hind wings also similar. Length of fore wing 10 mm. From Kartabo, April 13. Paratypes from Demerara River, March 24 (Cornell) and Tropical Research Station, No. 20655. (Beebe.) Anoplius varius Fabr. One male from Kartabo. Also recorded from Brazil. The abdomen differs from that of P. echinatus and P. ornamentus in lacking the pale band on fourth segment as noted by Fabricius. The fore wings are dark only at tip; the male has a hair brush on the fourth ventral segment. Po mpilinus orthodes sp. nov. Head, thorax, legs, antennae black, wings dark fuscous, abdomen with first three seg- ments rufous, others black, extreme base of first segment black. Clypeus, and lower face silvery, coxae also somewhat silvery; few short hairs on vertex and pronotum, a row of longer hairs on clypeus, tip of abdomen above with stiff black bristles, venter with a few hairs. Clypeus about three times as broad as long, lower edge truncate; face rather long and nai’row; vertex-width hardly equal to second plus third antennal joints; pronotum an even curve to front, behind angulate in middle; propodeum a trifle longer than broad, from side an even curve, no hair. Abdomen rather long, tip not compressed ; legs smooth, a few very short, stiff bristles above on front femora, spines on basitarsus very short ; mid-tibia with two rows of spines, outer row very stout ; hind tibia with two rows of spines, inner row the longer, inner spur equal to half of basitarsus; claws toothed. In fore wings the marginal cell triangular, more than its length from wing-tip, second submarginal cell as broad as long below, shorter above, receiving the first recurrent vein at apical third; third submarginal cell nearly as long below7 as second, almost to a point above, fully twice its length from margin of wing, receiving the second re- current (almost straight) at middle; basal vein interstitial with transverse; in hind wing the anal vein ends at forking of the cubitus. Length fore wing 8.5 to 10 mm. Two females, Georgetown, September 29 (Bequaert), and Kartabo, July-August (Wheeler) . Notocyphinae. Notocyphus vindex Lucas. One from Kaieteur, August 11. Also re- corded from Brazil. Notocyphus tyrrani cus Lucas. Recorded by Williams from British Guia- na (Studies in Tropical Wasps, 1926, p. 133). Also recorded from Brazil. Besides the above, Cameron describes two species from Demerara which I have not recognized in any material available to me, Priocnemis curtispinus , Pompilus clemeraraensis. Lebour: Larval Crabs from Bermuda 113 11. Larval Crabs from Bermuda.1 Marie V. Lebour. Naturalist at the Plymouth Laboratory. (Text-figures 1-19). A certain number of crabs in berry were collected on the shore and their young hatched out in the laboratory of the Ber- muda Biological Station. Also a few were taken in dredgings and hatched out and some megalopae were reared to young crabs. The larvae were not reared through series of stages or to the megalopae. It is however useful to describe these larvae, hitherto un- known, some of which show interesting fea- tures. A colored drawing was made from each zoea hatched from the egg and color notes are given here. A few breeding records were also noted. PORTUNIDAE. The zoea was hatched out from the eggs of Portunus sayi and P. depressifrons. The mcgalopa of Portunus anceps was obtained from the plankton and reared to the young crab. Both zoeae are typical portunids and so also was the megalopa. Portunus sayi (Gibbes). (Text-fig. 1). This species lives among floating Sar- gassum weed and is commonly found in the open water around Bermuda, sometimes drifting into the sheltered shallow bays in the Reach. It was found in berry from July to September. The eggs were 0.32 mm. across when nearly ready to hatch and were hatched July 1, 1938. The first zoea mea- sured 1.2 mm. in length (from the front of the head region to the tips of the telson fork.) The body has a faint purplish-pink tinge and there are black chromatophores in the mouth region, in the thorax and ab- domen and on the maxillipedes. Portunus depressifrons (Stimpson). (Text-fig. 2). This species is very commonly dredged from sand in shallow water in the Reach, or obtained by pushing a hand net. In 1 Contribution, Bermuda Biological Station for Re- search, Inc. dredgings taken in July and August, 1938, it was not breeding but in April, 1939, sev- eral were in berry and the zoeae were Text-fig. 1. Portunus sayi. a. Side view of first zoea, 1.2 mm. long. b. Telson. 114 Zoologica: New York Zoological Society [XXIX: 11 u- Text-fig. 2. Portttnus depressifrons. a. Side view of first zoea, 1.6 mm. long. b. Antenna, c. Abdomen and telson. hatched on April 23. The eggs measured 0.36 mm. across when ready to hatch. They have a pinkish color in the early stages and are dark brown later. The first zoea measures 1.6 mm. in length. The lateral spines on the telson are very small, especially the poster- ior spine. The body is very transparent and colorless with concentrated black chromato- phores in the mandibular region, in the center of the thorax and on the abdominal somites. Portunus anceps (Saussure). (Text-fig. 3). This species is not common, but a mega- lopa was taken in the plankton from a night haul in the Reach, June 18, 1938, which changed to a crab. This and other species were kindly named by Professor Garstang. The megalopa was large, the cast skin measuring 7 mm. long from the tip of the rostrum to the end of the telson. When liv- ing it was pale greenish-yellow with a very few dark chromatophores. The uropods bear 16 setae. The rostrum is long and straight. The first crab stage from the megalopa was pale yellowish with very powerful paddles. Unfortunately it died in casting its skin to the second crab stage. The carapace measured 6 mm. across with a very long lateral spine and 8 teeth in front of it. Grapsidae. The zoeae of Planes minutus, Pachygrap- sus transversus and Percnon gibbesii were hatched from the egg. Planes and Pachy- grapsus are typical grapsids, but Percnon, belonging to the Plagusiinae, has some in- teresting and peculiar features. Planes minutus (L.). (Text-fig. 4). The zoea of this little Sargassum Crab has been described by Hyman (1924), who hatched it from the egg. It is extremely common in the Sargassum weed all round Bermuda and was breeding freely from June to October and from March to June, occasionally in other months, so that there is hardly a month in which it cannot be 1944] Lebour: Larval Crabs from Bermuda 115 3. Portunus anceps. a. im. long. b. Carapace of first crab from mega- found in berry. The eggs were hatched out on June 13, 1938, and several times after- wards. They are dark brown in the mass, and measure 0.36 mm. across when nearly ready to hatch. The first zoea, 1 mm. in length, has much black pigment thickly spread on the thorax and in patches on the abdomen, the whole body being a very pale yellow so that the appearance is greenish. There are no lateral spines on the carapace and the dorsal spine is slightly longer than the rostral, both being short. There are for- wardly directed knobs on the second abdomi- nal somite, outwardly directed knobs on the third and backwardly directed knobs on the fourth. The third and fourth somite end postero-laterally in a process with 2 teeth. This is slightly different from Hyman’s de- scription and figure. The telson ends in a fork only slightly widened posteriorly with 3 teeth externally at the end of the forks and 6 long spines in the center. The an- tenna is a simple long process with spinules along its distal two-thirds and a minute pro- cess representing the exopod which bears a single hair. The megalopa is common in the plankton, and is greenish spotted with pink chromatophores, the carapace measuring 4 mm. in length. Thus it is large compared with the first zoea. The fifth pair of legs end in 3 long feelers. The front of the car- apace bends down as a rounded process. There is no true rostrum and there are no spines on the carapace. The uropods bear 19 spines. The first young crab from the megalopa measures 4 mm. across the car- apace. It is very like the adult. Pachygrapsus transversus (Gibbes). (Text-fig. 5). This species is very common under stones on the shore above high water mark, some- times in the water slightly lower down. All along the Reach and along the south shore of Long Bird Island it is abundant. It breeds from April to October. The eggs measured 0.28 mm. across and were hatched June 25, 1938, and later. The first zoea 116 Zoologica: New York Zoological Society [XXIX: 11 Text-fig. 4. Planes minutus. a. Side view of first zoea, 1.1 mm. long. b. Abdomen and telson. c. Cast skin of megalopa, carapace 4 mm. long. d. Abdomen, e. Uropod. f. Chela, g. Carapace from side. h. Carapace of first young crab from megalopa. measured 0.9 mm. in length, and was green with much black. It is very like Planes and of the same type but is smaller and has a shorter and straighter dorsal spine. Cano (1891) describes and Hyman (1922) quotes his description of the zoeae and megalopa of Pachygrapsus marmoratus which in appear- ance is very like the present species. It differs however in the telson and in having no knobs on the third abdominal somite. Two megalopae are described as belonging to P. marmoratus, but one cannot be certain if these really were two stages as no moults were obtained. Only one megalopa stage was 1944] Lebour: Larval Crabs from Bermuda 11? found in the Plymouth Brachyura (Lebour, 1928) but several workers have stated that two occur from other parts of the world and Aikawa (1937) notes them in Plagusia. Goniopsis cruentatus (Latreille). The Mangrove Crab is very common in rock crevices above but near water. Two fe- males were obtained in berry, Aug. 29, 1938, from rocks around the fish-pond in the Station grounds. This pond is connected with the sea. The eggs were dark brown, ap- pearing black in bulk, and measured 0.32 mm. across. Unfortunately they did not hatch, as it was difficult to keep them alive. Percnon gibbesii (Milne Edwards). (Text-fig. 6). This crab is common under stones on the south shore of Long Bird Island and on the other islands near. It runs very fast when disturbed and usually lives very near high water mark. It was in berry in July and some eggs were hatched July 15, 1938, but the zoeae died and the color was not noted. The eggs were bright red and measured 0.36 mm. across. The first zoea measured 1.9 mm. in length. It differs very much from Planes and Pachygrapsus but is somewhat similar to Plagusia dentipes described by Aikawa (1937), although differing in several points. Thus they both have lateral spines on the carapace but the antenna differs and there are knobs on abdominal somites 2-4 in Plagusia as there are in Planes and Pachy- grapsus, while in Percnon they are only on 2 and 3. The telson in both Plagusia and Percnon has no lateral spine, while in Planes and Pachygrapsus there are 2. Percnon differs from all zoeae so far known in hav- ing the posterior part of the carapace bent up to form a flange on each side. The an- tenna has a minute unarmed knob represent- ing the exopod. There is a very large and swollen anal papilla. A megalopa of this species found by Mr. John Armstrong of Harvard University among rocks on Cooper’s Island is interest- ing as it is very large, the carapace mea- suring 5 mm. in length. It has the abdomen outstretched and the uropods of a typical megalopa armed with 22 setae, but the last legs end in denticulate dactyls without any feelers, thus resembling a Spider Crab. There are 3 large teeth in front of the cara- pace and the walking legs are very long. It is possible that this is a megalopa in the second stage. Xanthidae. Gurney (1938 a) discusses the known zoeae of the family and shows that the char- acters used for the grouping of these (an- tennal exopod, lateral spines on the telson and spines on the carapace) do not corre- Text-fig. 5. Pachygrapsus transversus. a. Side view of first zoea, 0.9 mm. long. b. Abdomen and telson. c. Telson. spond with the present arrangement of the adults. Three more zoeae from Bermuda can now be added to this list (Gurney, 1936 d, having already described the first zoeae of Heteractaea ceratopus) — Panopeus bermu- densis, Leptodius parvulus and Xanthodius denticulatus. Of these the first two have the antennal exopod as long as the spine and the third has a short exopod with 2 terminal setae and appears to be a typical zoea of the sub-family Xanthinae. There are six known zoeae of this family which have the antennal exopod as long as, or very nearly as long as, the spine and no terminal setae. These are Pilumnus hirtellus (described by Lebour, 1928), P. minutus and P. vespertilis (described by Aikawa, 1929), Panopeus her- 118 Zoologica: New York Zoological Society [XXIX: 11 Text-fig. 6. Percnon gibbesii. a. Side view of zoea, 1.9 mm. long. b. Antennule and an- tenna. c. Posterior end of carapace, d. Abdomen and telson. e. Megalopa. f. End of last leg. mudensis and Leptodius parvulus (described here) and Heteropanope glabra (described by Aikawa, 1929). Cano (1892 c) describes three species of Pilumnus including P. hir- tettus but does not differentiate between them. These are quoted by Hyman (1925) and are all of the same type as hirtellus. The Pilumnus species (excepting those of Cano) were all hatched from the egg and agree except for the fact that P. minutus has only 2 lateral spines on the telson in- stead of 8. They all have 3 kinds of spines on the carapace. Leptodius parvulus agrees with Pilumnus hirtellus. Heteropanope gla- bra lacks the dorsal and rostral spines on the carapace and has only one lateral spine on the telson: Panopeus bermudensis comes between the Pilumnus species and Hetero- panope glabra, having very short rostral and lateral spines on the carapace, as in Pilumnus minutus, but agreeing with Heteropanope glabra in having only one lateral spine on the telson. These 6 species seem to form a group characterized by the antenna as op- posed to all the remaining zoeae known in the family which have the exopod shorter than the spine, either vestigial or with terminal setae. As Gurney (1938 a) points out, the difference between the zoeae of Heteropanope glabra and H. tridentata (the latter described by Tesch, 1922) is so strik- ing that if the description and parentage are correct any classification according to larval characters on these lines is upset. 1944] Lebour: Larval Crabs from Bermuda 119 Text-fig. 7. Leptodius parvulus. a. Side view of first zoea, 1.12 mm. long. b. Abdomen and telson. c. Telson. Leptodius parvulus (Fab.) Rathbim. (Text-fig. 7). The zoea is of the Pilumnus type but in form and color is very like that of Xantho- dius denticulatus described below. This spe- cies is common in rocks and under stones between tide-marks on the south shore of Long Bird Island. In berry in June. The eggs were hatched July 2, 1938, and were brown, 0.32 mm. across. The newly hatched zoea measures 1.12 mm. in length. It is a pale grayish-blue all over; it is not so transparent as Xanthodius denticulatus, with rather more red and black pigment, which besides being on the thorax and ab- domen also occurs on the maxillipedes and horns of the telson. There are dorsal, rostral and lateral spines on the carapace, all well developed, the dorsal and rostral being long. There are 3 lateral spines on the telson and knobs on abdominal somites 1 and 2. The antennal exopod is as long as the spine and has no terminal setae. Panopeus bermudensis Benedict & Rathbun. (Text-fig. 9). This species was dredged off the north shore cf Long Bird Island and was in berry June 10, 1938. The eggs measured 0.32 mm., very red. They were hatched June 14 The newly hatched zoea, 1.4 mm. long, is green- ish with black chromatophores and at first sight looks very like Planes minutus and Pachygrapsus transversus. It is peculiar and unlike other xanthid zoeae in having spine-like knobs on the fourth and fifth ab- dominal somites as well as On the second and third. There are denticulations postero- laterally on the carapace and only one lat- eral spine on the telson. The antennal exopod is as long as the spine. Panopeus occidentalis Saussure. From Spanish Point. In berry Aug. 11, 1938. The eggs were black, 0.32 mm. across. The crab died and the eggs were not hatched. Micropanope spinifer Milne Edwards. From a rock crevice on the south shore of Long Bird Island. In berry June 20, 1938. The eggs were dark red, 0.32 mm. across. The crab died and the eggs did not hatch. Xanthodius denticulatus (White) Rathbun. (Text-fig. 8). This crab was obtained fairly commonly on the south shore of Long Bird Island and was in berry in July. The eggs were hatched out July 24 and July 26, 1938, and measured 120 Zoologica : New York Zoological Society [XXIX: 11 b Tcxt-fig. 8. Xanthodius denticulatus. a. Side view of first zoea, 1.3 mm. long. b. Abdomen and telson. 0.34 mm. across when ready to hatch. The newly hatched zoea measured 1.3 mm. in length and was very transparent with black and red chromatophores on the thorax and abdominal; somites and in the region of the mandible. The antennal exopod is about a quarter the length of the spine, with 2 terminal setae. The dorsal and rostral spines are fairly long and of about equal length, the lateral spines short. The telson has the usual 3 outer spines. There are lateral spines on abdominal somites 3-5 and lateral knobs on 2-3. This appears to be of the Xantho type. OXYRHYNCHA. Majidae. The crabs hatched out belong to the sub- families Acanthonychinae: Acanthonyx pe- tiverii, and Majinae: Mithrax forceps and Microphrys bicomutus. The zoea of the first is like that of Acanthonyx lunula described by Cano (1893 b) and Boraschi (1921), the other two come between the Pisa type of zoea (Lebour, 1931 c) and that of Maia (Lebour, 1927 and 1928 b). Acanthonyx petiverii Milne Edwards. (Text-fig. 10). These crabs are abundant, living among seaweeds (including Fucus ) in rocky pools around Cooper’s Island and one was dredged in sand just outside it. I am indebted to Dr. F. A. Chace, Jr., for its identification. The seaweed appears to be the natural habitat as they have pieces of the weed on the ros- trum and sometimes about the body and closely resemble the weed itself. This is a new record for Bermuda. Females in berry were obtained in April and the zoeae hatched out May 2, 1939. The newly hatched zoea measures 1.8 mm. in length and corresponds in form with that of Acanthonyx described by Cano (1893 b), having a short dorsal spine, a very short rostral spine and no lateral spines, no lateral spines on the tel- son and the antennal exopod as long as, or slightly longer than, the spine, with no terminal setae, but 2 setae near the tip. There are knobs on the second abdominal somite. The body is a pale pinkish-brown with much black pigment, especially in the thorax and ventrally on each abdominal somite. The early eggs were golden yellow, 0.56 mm. across, the later eggs slightly larger and appearing black in the mass. Epialtus bituberculatus Milne Edwards. This crab was dredged from sand off Cooper’s Island with Acanthonyx and was also obtained by Dr. Klitemann in Hungry Ill 1944] Lebour: Larval Crabs from Bermuda Text-fig. 9. Panopeus bermudensis. a. Side view of first zoea, 1.4 mm. long. b. Antenna, c. Abdomen d. Telson. Bay, in berry, both in April. Unfortunately the eggs did not hatch, as the crabs died. The eggs are few and large, 0.72 mm. across, pinkish with a bright orange-red yolk. Mifhrax forceps Milne Edwards. (Text-fig. 11). This crab occurred not uncommonly on the south shore of Long Bird Island and To- bacco Bay near St. Georges, in rocks be- tween tide-marks. In berry in July. The eggs hatched July 6, 1938, and were 0.64 mm. across and dark brown. The newly hatched zoea measured 1.8 mm. in length. The dorsal spine is short, the rostral very short, and there are no lateral spines. There is one lateral spine on the telson, the forks of which bear spinules, and there is a pair of lateral knobs on the second abdom- inal somite. The antennal exopod is not quite as long as the spine and bears 3 terminal setae. There are lateral spines on the third, fourth and fifth abdominal 122 Zoologica: New York Zoological Society [XXIX: 11 Text-fig. 10. Acanthonyx petiverii. a. Side view of first zoea. 1.8 mm. long. b. Antenna, c. Abdomen and telson. somites. Thus in the type of antenna it resembles Maia but in most of its other fea- tures it is more like Pisa. The color is pale yellowish with pink on the abdomen and on the maxillipedes, black in the thorax and in patches on the abdominal somites. Microphrys bicornutus (Latreille). (Text-fig. 12). This crab is common under stones along the shore between tide-marks and was ob- tained from the north and south shores of Long Bird Island and from a sandy bay on St. David’s Island. In berry June and July. The eggs, 0.48 mm. across and dark brown, were hatched June 29, 1938. The zoea is very like that of Mithrax but the eyes are much larger and of a peculiarly vivid blue. The dorsal spine is short, the rostral very short, even shorter than that of Mithrax, and there are no lateral spines. There is one lateral spine on the telson, and the forks bear spinules. There are lateral knobs on the second abdominal somite and somites 3-5 have lateral spines. The antenna is like Mithrax except that the exopod is shorter in proportion to the spine. The body is a transparent yellow with red c-hromato- phores behind the carapace and at the base of the second maxillipede and black on the thorax in patches and ventrally on the ab- dominal somites. The first zoea changed to the second (last) in a small glass bowl. It is very little larger than the first, but the pleopods are very long and there are 6 setae on the maxillipedes. Parthenopidae. (Text-fig. 13). The megalopa of a species of Parthenope was occasionally taken in the outside tow- nets. This had a characteristic form with very long chelae. The carapace, including rostrum, measured 3.5 mm. in length. There is a long dorsal and a long rostral spine, the abdominal somites 3-5 have lateral spines and the uropods bear numerous setae (ex- act number not ascertained). As is found in all the known megalopae of the Oxy- rhyncha, there are no feelers on the dactyl of the last leg, but this is armed with teeth and hairs OXYSTOMATA. Leucosiidae. Several undetermined zoeae of Ebcdia- like form were obtained in the outside tow-nets, one of which was occasionally abundant. Three species, A, B and C, were distin- Text-pig. 11. Mithrax forceps, a. Side view of first zoea, 1.8 mm. long. b. Abdomen and telson. guished. All of them had much black pig- ment on the body. Species A (Text-fig. 14). This was much the commonest. There was much heavy black pigment in the thoracic region. There are long dorsal and rostral spines and fairly con- spicuous lateral spines. The second and third abdominal somites have lateral knobs and the telson has one lateral tooth near the angle. The first zoea measured 0.80 mm. in length; the third, which had 6 setae on the maxillipedes, and fairly conspicuous pleo- pods, measured 1.9 mm., and the fourth (last) measured 3 mm. The last had 8 setae on the maxillipedes and very long pleopods. Species B (Text-fig. 15). Has a fairly long dorsal spine (broken in the figure), a short rostral spine and no lateral spines. This measured ca. 2.2 mm. in length in the last stage. The telson is similar to A. Species C (Text-fig. 16). Measured 2.2 mm. in length in the second stage and is like B but with a longer rostral spine and no outer spine to the telson. It is worth while recording these forms, as species belonging to this family are ex- tremely rare in Bermuda and probably sev- eral have yet to be found there. Calappidae. Calappa flammed (Herbst). (Text-fig. 17). The megalopae of this species have al- ready been described by Smith (1880 b, as Calappa marmorata ) , but not figured. They are found occasionally on the outside tow- nets and swimming at the surface in the shallow water of the Reach at night, the adult being common around the coasts. A 124 Zoologica: New York Zoological Society [XXIX: 11 Text-fig. 12. Microphrys bicornutus. a. Side view of first zoea, 1.7 mm. long. b. An- tenna. c. Abdomen and telson. d. Telson. e. Side view of second zoea. Text-fig. 13. Parthenove sp. a. Megalopa cara- pace, 3.5 mm. long. b. Chela, c. End of last leg. d. Abdomen and telson. peculiar and striking feature of this mega- lopa, not noted by Smith, is the presence of very large oil-like spherical globules in the thoracic region, always 12 in number. Al- though they may be arranged in a slightly different manner they are always symmetri- cally placed. The body is a pale yellowish all over. There are 20 setae on the uropods, and, as Smith has pointed out, there are 3 feel- ers on the end of the dactyl of the last leg. This changed to a young crab, June 21, 1938. The carapace measured 4.2 mm. across. The specimen died and was slightly dam- aged, therefore the drawing shows the carapace distorted. A specimen from deeper water, presumably the same species, differs slightly from the others (Text-fig. 18). Cycfoes bairdii Stimpson. (Text-fig. 19). The megalopa, hitherto undescribed, of this species, which in many ways resembles that of Calappa, was obtained occasionally in the outside tow-nets. It is pale yellow all over with small red spots. The carapace measured 3 mm. long and is rather narrower than that of Calappa. There are here also large globules in the thorax but there are 14 of them. The last leg ends in 3 feelers. The uropods bear 17 setae. There are tubercles on the dactyl of the chela and along the propodus. This changed to a young crab which has a very round carapace, measur- ing 3 mm. in length. The chelae are of the characteristic shape of the adult. Bibliography. The lite^atm-e is all referred to Gurney’s Bibliography of the Larvae of Decapod Crus- tacea. 1939. Ray Society. 1944] Lebour: Larval Crabs from Bermuda 125 Text-fig. 14. Ebalia-Wke zoea A. a. Side view of Species A, 0.8 mm. long, b. Back oblique view. c. Stage 3. d. Stage 4 (last), e. Abdomen and telson of stage 4. 126 Zoologica: New York Zoological Society [XXIX: 11 Text-fig. 16. Ebalia- like zoea C, second (?) stage, 1.5 mm. long. Text-fig. 15 Ebalia- like zoea B. a. Side view of Text-fig. 18. Rough sketch of another mega- last stage, 2.2 mm. long. b. Abdomen and telson. lopa from deeper water. 1944] Lebour: Larval Crabs from Bermuda 127 Text-fig. 17. Calappa flammea. a. Megalopa, 4.5 mm. long. b. Cast megalopa, dorsal, c. Cast megalopa, ventral, d. First crab from megalopa, carapace 4.2 mm. long. e. Chela. Text-fig. 19. Cycloes bairdii. a. Megalopa, carapace 3 mm. long. b. Cast megalopa. e. Chela, d. End of last leg. e. Telson and uropod. f, g. First crab from megalopa, 3 mm long. h. A chela. Hungerford: Some Venezuelan Aquatic Hemiptera 129 12. Some Venezuelan Aquatic Hemiptera.1 H. B. Hungerford University of Kansas. [This is a contribution from the Forty-third or Venezuelan Expedition of the Department of Tropical Research of the New York Zoological Society made under the direction of Dr. William Beebe. The expedition was sponsored by grants from the Committee for Inter-American Artis- tic and Intellectual Relations and from four trustees of the Zoological Society, George C. Clark, Childs Frick, Laurance S. Rockefeller and Herbert L. Satterlee, and by invaluable assistance from the Standard Oil Companies of New Jersey and Venezuela.] By the courtesy of Dr. William Beebe I have been permitted to examine a small col- lection of aquatic hemiptera from Caripito, Venezuela. In the matter of distribution this list is of importance, as eight out of the eleven species have not heretofore been re- corded from Venezuela. 1. Lethocerus grandis (Linn.). Brazil, Paraguay, Colombia. A new- record for Venezuela. 2. Lethocerus annulipes H.S. U.S.A. (Florida), Mexico, Honduras, Guatemala, Nicaragua, Costa Rica, Panama, Cuba, Jamaica, Colombia, Venezuela, British Guiana, French Guiana, Puerto Rico, Paraguay, Uru- guay, Brazil, Argentina. 3. Lethocerus mello-leitaoi DeCarlo. Brazil, Paraguay. A new record for Venezuela. 4. Be.lostoma . dentatum (Mayr). Venezuela, Bolivia, Brazil, Argen- tina, Paraguay. 5. Belostoma asiaticum (Mayr). Mexico, Peru, Argentina, Brazil. A new record for Venezuela. 6. Tenagogonus duolineatus Kuitert. Bolivia, Peru, Paraguay. A new rec- ord for Venezuela. 7. Limnogonus celeris magnus Kuitert. Bolivia, Brazil, British Guiana. A new record for Venezuela. 8. Potamobates unidentata Champion. Panama, Colombia. The Venezuela specimen has a different color pattern and must be at least a new variety. 9. Rhagovelia insidaris Champion. Trinidad, Brazil, Colombia, Panama. A new record for Venezuela. 10. Rhagovelia tenuipes Champion. Mexico. A new record for Venezuela. 11. Ranatra macrophthalma H. S. Colombia, Peru, Bolivia, Dutch Guia- na, French Guiana, Brazil, Venezuela. 1 Contribution No. 693, Department of Tropical Research, New York Zoological Society. Breder: Studies on Blind Fish from Cueva de los Sabinos, Mexico 131 13. Ocular Anatomy and Light Sensitivity Studies on the Blind Fish from Cueva de los Sabinos, Mexico. C. M. Breder. jr. The American Museum of Natural History (Plate I; Text-figures 1-3). Introduction. The discovery of another blind characin from the Mexican caves of the State of San Luis Potosi, which is advanced beyond the conditions of eye degeneration found in the specimens of La Cueva Chica, calls for an examination of this new form in regard to its ocular anatomy and reactions to light. The present contribution is thus a continuation of the studies of Gresser & Breder (1940) and Breder & Gresser (1941a, b and c). The finding of this fish, under taxonomic study by Dr. C. L. Hubbs, has been re- ported by Tafall (1942 and 1943). This second cave, Cueva de los Sabinos, is about fifteen miles distant from La Cueva Chica. Chemitropic reactions of this fish, the La Cueva Chica specimens and the deriving river fish, Astyanax mexicanus (Philippi), have been reported by Breder & Rasquin (1943). The current status of the whole problem and a preliminary no- tice of some of the conclusions of the pres- ent paper have been given by Breder (1943c). The experimental part of the study has been carried on in the Department of Animal Behavior of the American Museum of Natural History. The histological sec- tions of the head have been made by Miss Priscilla Rasquin. The cleared and stained skulls have been prepared by Miss A. M. Holz. Dr. Ralph Meader, of Yale Univer- sity, examined and interpreted certain of the sections and provided the photographs of Plate I. Dr. W. K. Gregory gave advice in connection with interpretation of the osteological changes. Dr. E. B. Gresser, of New York University, gave helpful aid in connection with the comparisons of the ocular details. To all of these people the author wishes to express his grateful ap- preciation. Structure of the Eye. Compared with the eye remnant of the fishes from La Cueva Chica, described by Gresser & Breder (1940) and Breder & Gresser (1941a), the optical architecture of the fishes from Cueva de los Sabinos is still further reduced. Hardly anything is left of the eye capsule found in the more advanced of the mixed fishes from La Cueva Chica. In addition to this great re- duction of the capsule there is no optic nerve connecting with the brain. Unsatisfied with the difficulties in find- ing and interpreting these remnant struc- tures, the slides were referred to Dr. Ralph Meader of Yale University who is engaged in studying the brain tracts of these fishes. His careful examination of the slides brought out numerous fine details about which he wrote as follows, for inclusion in this paper. “An optic capsule is present embedded in the fat which fills the region of the head normally occupied by the eye and orbit in eyed river fish. The capsule consists of a sphere of moderately heavy cartilage, in- complete on its medial wall where it is replaced by a lamina of dense connective tissue. Within the capsule is a more or less loose network of connective tissue in which lies a small heavily pigmented sphere or cyst, which presumably repre- sents the choroid coat. Although some stain- able material is visible within this cyst and it contains some cells of unidentified ori- gin, no recognizable elements of a retina can be found. There is no lens nor any of its associated structures. Outside of the pigmented sphere are a few vascular chan- nels which are probably the remnant of the choroidal gland. A few strands of stri- ated muscle are seen in the region of the connective tissue which completes the me- dial wall of the cartilaginous outer layer of the' capsule. The muscle fibers do not show well the normal histology of voluntary muscle. Near them, attached to one cap- sule, can be seen a small bundle of tissue which in its shape and position resembles the atrophic optic nerves of some blind specimens from La Cueva Chica. The bun- dle can be traced medialward only a short 132 Zoological New York Zoological Society [XXIX: 13 distance from the capsule before it be- comes very small and then ends. No sim- ilar bundle can be seen attached to the other capsule. “From the anterolateral aspect of each optic capsule a stalk of tissue extends towards the skin. On one side this stalk is small and hollow near the capsule. A few cells with large nuclei line the cavity which is filled with amorphous debris. Nearer the skin the stalk becomes smaller and solid. Its final attachment cannot be certainly made out because of torn sec- tions, but it probably attaches to the con- nective tissue underlying the skin that covers over the obliterated orbital cavity. On the opposite side of the head, the stalk is relatively large, hollow, and has definite attachment to the skin. The same type of cells with large nuclei lines its cavity and the cells which form the rest of its thick wall resemble somewhat those of the skin. At one place they nearly become continu- ous with the surface epithelium. These stalks resemble those described by Gresser & Breder (1940) and Breder & Gresser (1941a) for the blind fish from La Cueva Chica. They are most like that shown in their figure 2 (1941a). “The condition of this optic capsule in the Sabinos fish seems to be an exaggera- tion of the reduction in size and in the loss of retina seen in occasional specimens from La Cueva Chica. For example one of the latter specimens prepared here ex- hibits no retinal structure although in one pigmented cyst is evident a multipolar nerve cell whose presence I cannot explain. This specimen, however, does have an optic nerve that can be traced to the brain whereas in the Sabinos specimen, no op- tic nerve can be certainly identified and the strand tentatively named the optic nerve peters out in the orbital fat. No optic nerve can be found connected with the brain.” Plate I shows sections illustrating the above description and is comparable to the plates in Breder & Gresser (1941a) of the La Cueva Chica material, and taken to- gether they clearly indicate the increasing reduction of architectural detail. Modifications of the Skull. With such a reduction of the eye it is to be expected that the osteological elements surrounding this area should show some modification. These include progressive changes accompanying the reduction of the globe and involve conspicuously a closing in of the circumorbital series of bones. The details of these changes in the conformation of the skull are indicated in Text-figure 1, which has been based on camera lucida drawings of cleared and stained heads. The sutures in the outer series of bones were easily evident, but some of the elements of the chondrocra- nium were discerned with difficulty or not at all, leading to the omission of certain bone names in the figure. The normal-eyed river fish, Text-figure la, shows a series of five suborbital bones and a well ossified skull, as was evidenced by the taking of the stain (alizarin). This skull is a generalized type of characin skull as may be noted by a comparison with the figures of Gregory (1933) and Gregory & Conrad (1938). It resembles clearly that of Cheirodon, which these au- thors, following Eigenmann (1917), con- sider basic to the group containing Asty- anax and which they designate the Cheiro- dontinae. A specimen from La Cuevva Chica with a reduced eye shows an essentially similar structure, but with the circumorbital series of bones closed in and with the ad- dition of three new elements. These are evidently fragmentations of so2 and so3, or new centers of ossification induced pri- marily by the shrinkage of the eye diam- eter. The arrangement of this series of bones is indicated in Text-figure lb, to- gether with the diameter of the globe and the larger anterior extension of the orbit. A typical blind fish from La Cueva Chica shows a still further closing in of the suborbital series, with the elimination of the globe proper, Text-figure lc, and what appears to be a fragmentation of so.e> into three elements. This is one less than the number of new elements shown by the fish with merely a reduced eye diameter, and there are other changes to be discussed later. A second specimen of the same type showed a similar fragmentation of this bone, but the arrangement of the separate elements is rather different. This series of circumorbital bones is seen in Text-figure Id. Both these fishes showed minor varia- tions in these details from side to side. Such variations in the mixed fishes from La Cueva Chica should probably be ex- pected in any structure associated with the features of reduction and are evidently equivalent to the variations in the eye it- self already discussed by Breder & Gresser (1941a). A specimen from Cueva de los Sabinos shows a still further closing in of the sub- orbital series, as is indicated in Text- figure le. Along with this but not indi- cated in the figure is an evident reduction of ossification, indicated by the manner in which the stain is retained. It will be noted that the circumobital series has re- turned to five in number; thus the fragmen- 1944] Breder: Studies on Blind Fish from Cueva de los Sabinos, Mexico 133 0) i m 3 a ° 'g'H o 0J P ft g* u c3 — i " -5 3 .S 3 bo Om U U . . fi -M -2 -M 5- ft M CD ■ " p^jQ ® O ti „ £ O oo . o — asn - o •a)^UJ-c!-r|r2rc3'£3 i m 01 S+J B On" _ft,S3 M-HC3 «W — ^ ^ w . U C3 . I I h^h.1 w O' o> ^ 1 > ^ i. ' H3 h o X VI Z .. . X o ft. &. be cp Or-f >, TO f-i I ~-£ a> "t -£ fi +? © CU Q mSE 2 £ I a W«H W £ > H ftia iii n b fl M > «c -a a % * « ft O I=J 0) J f, , £ r£ JJ o S4" O+J l/l «4_| fi 3 m O c3 '£ fig 3 oj'm £ = o - Zn-^rZ WrO , .v .S^S^is T— I «h o> O . © rH r~| _Q r-1 t-t I E g ° §? § „ w <2 .-s °-f£r£i ft Hc^SftHgig _Q O o U2 o £ d I w w ^ o I ° A^J ag- £ w ,_ r - 1 S-S.S3 ft ft supraclei thrum, suporb — supraorbital. 134 Zoologica: New York Zoological Society [XXIX: 13 tation or new elements, is evidently a tran- sient phenomenon, which reaches its high- est development immediately on the reduc- tion of the eye size and is then gradually reduced to the original state, perhaps as a new level of organizational stability is at- tained. This specimen, on the other side, was not so far reduced, soa being repre- sented by two elements which occupy an area nearly equivalent to that of the single so3 - on the side illustrated. The increased exposure of the chondro- cranial elements is evident from the series of three skulls in Text-figure 1, as the sur- face bones crowd forward as the eye is reduced. If there is any involvement of the basic nature of the jaw musculature, as found by Petit (1940) for Typhleotris madagascariensis Petit and Coecobarbus geertsi Boulenger, it is not evident from these specimens. Text-figure 1 well indi- cates the reduction of the size of the max- illa, premaxilla and dentary and the near loss of the prefrontal. Otherwise there is a general shifting around of all the head bones in more or less minor fashion, as they evidently spread or contract to fit the new conditions imposed by the removal of the eye. The greatly increased size of the supraorbital is notable. The sclerotic bones, not illustrated, are normal and ossi- fied in the river fish but they degenerate and disappear along with the globe, ex- cept as they are possibly represented by some small cartilaginous remnants. The curious backward extension of the angle of the properculum in the Cueva de los Sabinos fish and its median separation from the operculum were hardly to be ex- pected. The pointed posterior process of the supraoccipital is reduced in size ma- terially in this series, moving forward as it and the parietal usurp the place occupied by the reducing frontal on the top of the skull. The above-mentioned changes are hard to comprehend and give the impression of being hit-or-miss and that they might just as well have taken on a variety of other arrangements. On analysis it is seen, how- ever, that actually there is a large amount of regularity in these transforming skulls and that the osteological elements are mov- ing and rearranging themselves under the force of some general regulating control. This may be most readily visualized if we make use of the methods of Thompson (1942) in which some system of coordi- nates is established on the original form and allowed to distort itself as the lines follow the elements that they intersect. In such construction it is a more or less general practice to start with a set of right angled rectilinear coordinates. This pro- cedure will yield all the usual data from such a series, but in the present case, in- stead of applying the simple Cartesian system, it was found to be easier to fol- low the transformations by the use of a system of polar coordinates with its origin at the center of the eye. The reasons for this are manifest. In the generalized fish skull the surface bones are arranged in a series of essentially concentric elements about the eye. This is obvious from Text- figure la. This same skull is shown in Text-figure 2a with such a grid of polar coordinates superimposed upon it.. These coordinates are purely arbitrary in dimen- sion and any other distances between the lines would have been just as suitable. Following these through to La Cueva Chica blind fish, Text-figure 2b, the regu- larity of the changing elements in their basic topography is evident and this stage is clearly intermediate between the river fish and the further advanced Cueva de los Sabinos fish, Text-figure 2c. From even a cursory examination it at once becomes evident that the whole skull, as a unit, is shifting and what would at first seem to be capricious changes in bone shapes and sizes are actually mandatory on each element if it is not to do violence to some master plan that controls the architectural changes. It is curious that the radial ordinates in all still find a common center in the orbit and it is possibly more than accidental that the central circular ordinate becomes more or less bean-shaped, with the con- cave side uppermost, which is in agree- ment with the shape that the reduced eye capsule approaches. It should also be noted that it is only by treating the so3 in its three parts as a unit that the integrity of the regularity of the intersecting coordinates can be retained. This could be taken to suggest that this increase in number is truly due to a process of fragmentation. Likewise in Text-figure 2c, where so3 is again a single bone, if the other two continued to exist and were treated as before, as a unit, the regularity of the superimposed grid would be destroyed. In a similar manner the other cranial elements could be discussed, as all subscribe to this general proposition. That is, if they were disposed in any other way, independently of the rest, and there is adequate room from such changes, a smooth flowing grid of parallel or regu- larly diverging lines would not fit. For example, the greatly expanding supraorbi- tal and the contracting frontal would seem to have no independent reason and the two could have remained very much more as they were and still jointly cover the same space, but as is clear from the figures 1944] 135 Breder: Studies on Blind Fish from Cneva de los Sabinos, Mexico they do change their relative sizes, as if merely to meet the inscribed coordinates in a regular manner. If it is objected that there are actually small divergencies in each of the two dis- tortions of the initial grid it must be borne in mind that such divergencies can be of two sorts: partly organic irregularities and partly artifacts due either to the diffi- culties involved in handling these small skulls or to the ineptness of the draughts- man. The remarkable part of the construc- tions seems to the author to be, not small discrepancies but rather that there is such a large amount of profound agree- ment between the distorted grids and the skull elements themselves. It must be borne in mind in this connection that simi- lar constructions have been made between related species which had all their organs intact and merely varied in dimension. When, in a case such as the present, a dominating organ is merely subtracted, it would seem to be of more than passing interest that the remaining parts do show such a fundamentally simple basic plan in reconstituting themselves. It also goes far to indicate the nature of the influence of the eye on the whole pattern of develop- ment of the fish skull. In making these constructions the ac- tual method employed was to point off, item by item, the intersections of a given line corresponding- to the grid placed on the normal-eyed skull. Then a line was drawn through these points freehand as one might draw such a curve from any biological data. Thus, some of the points fall exactly on the line and others scatter slightly to either side of it. It would have been a greater refinement to use some method of curve fitting such as the method of least squares, but the limitations of ac- curacy of the basic material clearly do not warrant such elaborate treatment. When these lines were inked in with the aid of draughtsman’s curves it was discovered that the radial ordinates, whatever the nature of their curvature, were all clearly not logarithmic, but that the concentric ordinates partook of this character to a very large degree. That is to say, it was found that for the latter the so-called draughtsman’s “Logarithmic spiral curve” fitted these freehand curves perfectly for much of their length while for the radials it was necessary to employ the so-called “Irregular or French curves” to obtain a fit. Those portions of the concentric ordin- ates anterior to the diagonal passing through the coracoid, sos and supraorbital, maintained their logarithmic feature al- most completely. Those posterior to this line took on a distortion, increasingly so as they approached the opposite diag- onal, passing through the mandible, so2, so5 and the scale bone. Nevertheless these curves were nearly all drawn by means of the logarithmic spiral. 136 Zoologica: New York Zoological Society [XXIX: 13 All this would seem to indicate the very great influence of the eye on the whole architectural plan of the skull. It is inter- esting to compare this essentially exponen- tial rearrangement of a skull with its cen- tral element removed with the logarithmic growth forms of mollusc shells discussed at length by Thompson (1942). The fact that in a structure as complicated as a fish skull such logarithmic elements are clearly present, as have been so simply and ele- gantly demonstrated in a clam, is evidently due to the fundamental nature of growth, which in an intricate structure becomes involved and complex but does not obliter- ate the basic regularities inherent in or- ganic development. If we look upon this polar grid as a field- of-force diagram in which the concentric ordinates represent equipotential lines, the eyed fish would seem to have its ele- ments disposed very much in accordance with the grid centered in the eye as ar- bitrarily laid down. A notable exception to this is the disposition of the supraoccipi- tal, which has a strong radial trend. In the other two fishes it will be noted that the greatest distortion of these grids is close to, but not identical with, the seem- ingly radial trend of the supraoccipital. The grids conceived as fields of force change in such a way as to suggest that there is only one pole of any significant magnitude in the area covered by the grids, namely, that centered in the eye. If an- other were present it would be expected to lead to some dumbbell-like curves, which are clearly absent. On this basis the curi- ous outward extension through the scale bone becomes explainable on the basis that it is due to another field of influence lying outside of the area of the limits of the grid and which if extended might produce such dumbbell-shaped figure characteristic of a bipolar field. The only evident struc- ture that this diagonal and the “points” of the concentric ordinates are directed toward is the base of the dorsal fin, which, in advance, would be expected to exert such an influence. This is not reflected in the original “monopolar” field which is com- pletely arbitrary, but shows good agree- ment with the circular disposition of the skull elements, except in the supraoccipital area where they appear to be radial in arrangement as previously noted. Clearly, then, when these circles distort with the shifting skull elements, they become affec- ted by any other influence, such as that of a second polar field, which is ignored by the very nature of the original grid. This is actually what has evidently happened in these transformations. It thus may not be far from the truth to say that these diagrams indicate the almost complete dominance of the eye except as modified by a dorso-caudad influence probably rooted in the dorsal fin. Considering the concentric circles of Text-figure 2a not as polar coordinates, nor as equipotential lines of force but as simple geometrical figures expressible in formulse on ordinary graphic treatment, it is possible to express the first by the equation (x-p)2 + y2 = e2x2 This, of course, is the basic formula for conic sections, of which the circle is merely a special case, and which is more simply expressed as X2 Y2 M2 m2 in which M — the major axis and m = the minor axis. The other two grids should be expressible by some exponential expan- sion of this expression. Then the difference between the three expressions should yield a mathematical measure of the basic na- ture of these differences. Obviously with the present case the labor involved would hardly justify making the involved calcu- lations necessary and this is mentioned merely to indicate an approach that might have considerable utility in other but sim- ilar connections. Lest it be thought that some artifact has been introduced by centering polar coordinates in the eye, that is, on a struc- ture in which the elements are obviously arranged more or less concentrically, other constructions have been made. One on an orthogonal basis shows essentially the same thing, as it necessarily must, but in not so clear a fashion; it consists of both horizontal and vertical ordinates ap- proaching each other in the eye region and suffering considerable distortion in the temporal region ; its verticals and hori- zontals in this region both bend toward the dorsal fin. Another construction on polar coordinates centered in the scale bone shows approach of the concentric ordin- ates in the eye region while three of the radials, the anterior horizontal, lower verti- cal and the included diagonal, approach each other in the eye region and pass through the scale bone and with slight angular divergence also point to the dorsal fin region. Instead of discussing the matter in terms of centers of influence which was based primarily not on these constructions but on the long recognized concentric ar- rangements of the surface bones of the head it might be more appropriate to con- sider the matter in terms of lines of 1944] Breder: Studies on Blind Fish from Cueva de los Sabinos, Mexico 137 strain; the large obvious one in this case being that which runs diagonally upward and back from the mandible through the eye and occipital region. Since the skull is under various strains from muscular at- tachments it is perhaps to be expected that such a distortion line should appear in a direction involving both the powerful man- dibular musculature and to the point of attachment of the occipital condyle which involve also the whole locomotor body musculature. Such considerations, how- ever, would carry us beyond the province of this communication. A still further analysis can be made by these methods for the purpose of trying to understand their full implications. If any sets of the constructions be drawn superimposed on one paper, then the homo- logous intersecting ordinates may be con- nected by lines of varying length with an arrow at their forward end indicating di- rection of travel of this point on the trans- forming skull. Very complicated diagrams result that do not look unlike weather- maps with their arrows of wind direction, but in which the length of the arrow indi- cates the proportional distance of travel. By combining visually each of the three sets of diagrams, using different ordinates, it becomes apparent that all show the same features in different terms; this agrees nicely with what has already been set forth. But again this is not the place for- extended discussions of such matters. As further evidence of the similarity of these constructions to those of fields of force it may be mentioned that Hartridge (1920) and Ponder (1925a and b) applied the equipotential curves of Cayley, devel- oped to describe magnetic fields, to the shapes of erythrocytes, and Breder (1943a) discussed them in connection with the shapes of fish eggs. With the information at hand, based on the construction of the polar grid, it would be of interest to re- calculate the whole matter on a basis not of equally spaced concentric lines but on the locus of the equipotential curves of Cayley; but this would clearly carry us out of the bounds of the present contribu- tion and no further into an understanding of the eye itself. Obviously, what has been dealt with here is a projection of one of the three dimen- sions and for a rigorous ana’ysis all should be considered, just as the equipotential curves of Cayley are descriptive of sur- faces and not of lines on a plane. Breder (1943a) discusses these relations in con- nection with fish eggs and Thompson (1942) gives an extended dissertation on the whole matter. The preceding discussion has centered on the dermal elements of the skull and suffice to indicate the essential regularity of the modifications. No attempt has been made to analyze the rearrangements of the chondrocranial elements which among other things would practically demand a difficult three-dimensional treatment. It was evident from the sections however, that the bones forming the eye socket un- derwent extensive modification. These did not close in, as might be thought, but be- came thinner and lost their solid attach- ment to the brain case, to some extent actually increasing the tissue-debris-filled eye socket in the Cueva de los Sabinos fish. Experiments on Light Reactions. In order to determine the extent of light sensitivity in the Cueva de los Sabinos fish, an experimental approach was made similar to that employed by Breder & Gresser (1941a and c). For this purpose two troughs were prepared with one end covered and the other under the influence of a 25-watt light. They each measured thirty-nine by seventeen and one-quarter inches and were filled with water to the depth of one and one-half inches. Thus there were two equal compartments each nineteen and one-half by seventeen and one-quarter inches and identical except that one was illuminated and the other under a solid cover one-half inch over the water’s surface. At five-second intervals the positions of the fishes were noted in reference to which compartment they occupied. From this work the data of Table I were compiled and directly compare to the earlier work on the other forms. It is clearly evident from this Table that the fish from Cueva de los Sabinos show no clear reactions to light under the conditions of the experi- ments, as compared with the earlier stud- ies or as compared with the La Cueva Chica specimen used as a control against the Cueva de los Sabinos material. Be- cause of the agreement of the latter with the more extensive earlier studies, it was possible to reduce the number of tests, as all the earlier work can be considered as a control for the present experiments. Text-figure 3 shows these results in graphic fashion and compares the present data with that of the earlier work. This is indicated by dotted lines. It will be seen that the present control fish, La Cueva Chica stock, gave readings of between 60 and 80 per cent, light negativeness in terms of random expectation. As pointed out by Breder & Gresser (1941a), the smaller of these approach 60 per cent, of expectation while the largest approach 80 138 Zoologica. New York Zoological Society [XXIX : 13 Text-fig. 3. Comparison between light reactions in fishes from La Cueva Chica and from Cueva de los Sabinos, shown by solid lines, and between La Cueva Chica speci- mens with and without optic capsule, shown by dotted lines, from Breder & Gresser (1941b). New data indicated by underlining. All expressed in terms of expectancy of random movement. See text for full explanation. per cent., the reason clearly being due to the greater absolute mass of tissue over- lying the blind but light-sensitive capsule. Its removal, as they showed, places the fish close to 100 per cent, of expectation, as indicated in Text-figure 3. It is close to this point that all the readings on the Cueva de los Sabinos material is located. This was, of course, to be expected since it has been shown that these fish lack all remnants of retinal structure and there is no connecting optic nerve. They are thus normally comparable to the condition produced in the La Cueva Chica material by operation. The La Cueva Chica stock is thus sepa- rated from that from Cueva de los Sabinos as well as from operates by a range of nearly 20 per cent, as measured by this system of notation, checking perfectly with what might be deduced from the morphol- ogy of the situation. Discussion. Although in all the experiments on light reactions in this and the two earlier papers care was exercised to avoid thermal differ- ences in the light and dark chamber, the direct effects of radiant energy of course could not be eliminated. Although all ex- amples with some light sensitivity reacted negatively to light, see Breder & Gresser (1941a and b), it will be noted that those forms without an optic nerve or with the capsule removed show a mean value slightly on the positive side. Since in all there have been 72 such experiments made, the present data and that of Breder & Gresser (1941b), this may well indicate a significant difference. Expressed another way, all but one of the sets of experiments were somewhat over 100 per cent, on the scale adopted. It is suggested that this may be reflecting a slight positive bias on the part of these apparently light-indiffer- ent fish to accept the warmth of the radiant heat of the light bulb, perhaps very much as a blind man can detect the edge of a shadow if the sun is bright, but on a much finer scale of discrimination. Since these fish move readily into the direction of warmth, as has been indicated before, this effect is surely to be expected. To measure this in any great degree of refinement a 1944] Breder: Studies on Blind Fish from, Cueva de los Sabinos, Mexico 139 number of difficult obstacles present them- selves and the present methods are not delicate enough to certainly determine it. Such an attempt has not been made, for with other facts already at hand it seems to be not of sufficient significance in pres- ent connections. Scharrer (1928) has shown that blinded Phoxinus laevis are light sensitive. This sensitivity he found located directly in the midbrain, made possible, at least, by the thin and not completely opaque over- lying structures. He concluded that possi- ble skin sensitivity was definitely not in- volved. It might be that the effect above considered is actually a measure of a sim- ilar condition in these fishes. Experiments along the line of those conducted by Scharrer would have to be undertaken in order to determine these effects, although it does not seem that he eliminated the possible effect of radiant heat, as above considered, from his experiments. In connection with this it should be noted that very few fishes have been un- equivocally .shown to possess integumen- tary photosensitivity. It is evidently pres- ent in ( Chologaster ) Forbesella, Eigen- mann (1909), and in Amblyopsis spelaeus DeKay, Eigenmann (1909), Parker (1909a) and Payne (1907). In addition to these fishes proper it is present in larval Petromyzon (or Lethenteron) Parker (1909a). Although absent in Amphioxus, photosensitive cells are located in the nerve tube and exposed by the transparency of the creature, Boeke (1902) and Parker (1908a and b). A variety of species, Fundulus heterocli- tus (Linnaeus), Long (1904), Parker (1909a and b), Fundulus majalis (Wal- baum), Long (1904), Mustelus canis (Mitchill), Anguilla rostrata (LeSueur), Stenotomus chrysops (Linnaeus), Tauto- golabrus adspersus (Walbaum), Tautoga onitis (Linnaeus), Chiloynycterus sclioepfi. (Walbaum), Opsanus tau (Linnaeus) and Microgadus tomcod (Walbaum), Parker (1909b) were shown to have no such sensi- tivity. It would thus seem that dermal photosensitivity is largely absent in fishes, only the Amblyopsidae having had such demonstrated. Other kinds of such sensi- tivity, as in Phoxinus, would seem to be incidental to the translucency of struc- tures overlying the central nervous system. Although the La Cueva Chica material will avoid light except if it is overridden by sufficiently large thermal or other dif- ferences, it is evident that the Cueva de los Sabinos fish indicate no such negative reactions to light. Thus it follows that while the first are probably held in their cave by this means, the latter are not pro- tected from blundering into a surely fatal environment by any such mechanism. Until much more is known about the geography of Cueva de los Sabinos we can only specu- late as to whether these fish are prevented from entering outside waters by physical barriers or whether there is a continual drain off of that population. Although di- rect tests have not been made it is to be expected that the fish from Cueva de los Sabinos will be found to react to current and water temperature in a manner essen- tially identical with those of the fishes from La Cueva Chica and eyed river fish. The essential similarity of their gross re- actions as well as minor differences in behavior in reference to other senses are discussed by Breder & Rasquin (1943). It is possible that physiological methods of greater refinement could show light sen- sitivity on a “micrometric” basis by some modification of the methods of Kurodo (1932), who showed by a kymograph trace that certain fishes gave an immediate res- piratory response to various chemical stim- ulse. Such a method was found inadequate to the peculiarities of these fishes as is reported by Breder & Rasquin (1943). As with the development of chemical sensory changes discussed by the above, the eye structure, the skull architecture and behavior of the Cueva de los Sabinos stock as reported here also suggest that these fish are a further development of the conditions found in La Cueva Chica mate- rial rather than that the two are independ- ent developments. Whether or not the present La Cueva Chica material should be looked upon as “hybrids” between some small and isolated group of Cueva de los Sabinos fish and the Rio Tampaon stock made possible by a new contact the pres- ent studies do little to illuminate. Hatch (1941) considers that degenera- tion occurs in caves, because in the ab- sence of selection, mutations of various kinds can survive and that degeneration follows because most mutations are those of loss. Walls (1942) holds similar views, writing as follows: “Just how the eyes of any blind fish spe- cies were led to disappear, we cannot say. An old idea was that where the eye had become useless, there was a positive incen- tive for eliminating the organ, since this would save energy both in adulthood and — especially — during growth. This notion seems ridiculous nowadays, for the pro- portion of a growing animal’s food-intake which goes to enlarge the eye is negligible. Most of the energy released from food goes for motor and secretory activity, and only a very small part of the food is converted into new protoplasm. Nor does the disap- 140 Zoologica: New York Zoological Society [XXIX: 13 pearance of an eye leave a hole in the head — -its volume is occupied by tissues (mainly muscle) which consume just as much energy as the eye had done. “Though a normal eye is excess baggage to a cavernicolous or limicolous fish, there appears to be no urgent reason why he should get rid of it. Useless organs do not always promptly disappear simply because they have become useless — as witness the human appendix, coccyx, platysma, tonsils, wisdom teeth, et al. We are left to suppose that in the immediate outside ancestors of most cave species the eye was ‘trying’ to disappear anyway, but was prevented from doing so, by natural selection, because it was useful and necessary. The usefulness once removed by the assumption of cavern- icolous life, the inherent tendency for the eye to shrink was allowed to express it- self, even onto the logical end-result — complete loss. “This explanation does not tax the imagination of ichthyologists as severely as one might think. In many open-water fish species, reduced-eyed individuals appear as soon as the food supply is made abundant and predatory enemies are removed. Lack of competition then permits the full de- velopment of individuals which, since their germ-plasm has undergone ‘mutations of loss,’ would formerly have been suppressed by starvation or capture. Loss-mutations are known particularly to affect the more complex organs of vertebrates, such as the eye. A species or family in which such mutations occur with especial frequency has of course no advantage, over others, in any attempt to become adjusted to a hab- itat in which illumination is reduced or absent. But if a group which throws loss- mutations also produces an unusual num- ber of other trial-and-error modifications (as seems likely), then such a group might readily evolve the dermal sense organs, barbels or what-not required to cope with a dim-light environment. Once adapted to dim-light existence, such a group would actually be better off in a cave, if it hap- pened to find one, than outside where there were predators to be dodged. And once in- side the cave for good, a rapidly-mutating species would inevitably lose what re- mained of its eyes, though without being under any positive necessity of doing so.” Pike (1943a and b) would refer such degeneration to thermodynamic irrevers- ibility and states his interesting views as follows : “Confusion has arisen concerning cer- tain processes which have been called re- versible, in living organisms. For four dec- ades, biologists have searched for so-called reversible processes comparable to the so- called reversible processes of general chemistry. The object was to disprove the existence of supposed vital characteristics of living organisms. A number of these processes have been found, although there has been scepticism concerning their actual reversibility under biological conditions. In these four decades, we have come to recognize more clearly that, in inorganic nature, processes which are not strictly mechanical are irreversible, thermodynam- ically. “Two costs must be reckoned in non- mechanical processes — that in free energy, and that in entropy. Unless both can be kept at zero when the operation is put through in the reverse direction, the process is irreversible thermodynamically. ‘If an ir- reversible process can occur, it will occur.’ (Houstoun). Unless there is some mysteri- ous mechanism in living organisms to for- bid irreversible processes, they will oc- cur. Experimentally, we have found many such, but we have never found one which is reversible thermodynamically. To show the existence of reversible processes in living organisms would be the equivalent to a demonstration of vitalism. “These considerations bear on certain hypotheses of heredity, the transmissibility of the effects of the environment, and the degeneration of specialized tissues in changed environments, some of which hy- potheses postulate thermodynamically re- versible reactions, or their equivalent in thermodynamic efficiency. “It has been asserted that degeneration of eyes of cave forms is due (1) sole’y to chromosomal changes, and (2) to inheri- tance of the effects of the environment. Since photo-receptors appear only in snecies which have some time lived in the light, it seems permissible to assume some causal relationship between light and photorecep- tors until demonstration of their complete independence. “The chromosomal hypothesis of degen- eration of the eyes of cave forms postulates that the process of heredity will go on for generation after generation, with no dissi- pation of free energy or gain in entropy, and, hence, be a truly thermodynamically reversible process, or a process of equivalent thermodynamic efficiency, such as is found nowhere else in nature. “Since darkness is merely the absence of light, and cannot be a form of energy, it can have no positive effect, and can produce nothing. Since it has no positive effect, the effect of darkness could not be transmitted hereditarily. “The observed result of living in per- petual darkness is what we would expect when organisms experience a failure of the 1944] Breder: Studies on Blind Fish from, Cueva de los Sabinos, Mexico 141 driving force concerned with the develop- ment of photoreceptors. “When the phenomena of recession and degeneration, of use and disuse, in on- togeny and in phylogeny, are regarded from the dynamical point of view, there appears to be some fundamental relationship to their irreversibility, in the thermodynamic sense of irreversibility.” There is clearly no basic conflict between these two views or the view that genes carrying eye defects can be perpetuated in such environments and spread through the population by some such means as ex- pressed by Hubbs (1938) of which Breder (1943b) wrote, “An interesting speculation on the possible mathematics of the genetic reduction of other than blind white forms coupled with repeated separation of small population groups is discussed by Hubbs.” In this connection the interesting report of Ogneff (1911) showed that gold fish kept in darkness up to three years suffered his- tological degeneration of the retina. He was satisfied that his fish were blind. Unfortu- nately he says nothing about light sensitiv- ity. The eyed specimens from La Cueva Chica were mentioned by Breder (1943c) as follows: “All fish brought freshly from the cave act like the blind fish in regard to feeding. This has been checked twice, all forms from the cave performing circling movements for food finding for at least a couple of months and then those with struc- turally good eyes apparently learning the meaning of a retinal image and taking on the feeding habits of the river fish. These forms at all times, from the first, ‘jump’ at the passage of a shadow, like those with merely an exposed retina.” It is proposed to check this material histologically to de- termine if this inability to find food by vis- ual means had to do with retinal degenera- tion and later reorganization under the in- fluence of light or on a basis of learning. The work of Ogneff (1911) strongly suggests that the former may be playing a large part in this connection. The nature of such changes in either direction should yield data bearing on the views of Pike (1943a and b). Summary. 1. The blind characins from Cueva de los Sabinos, San Luis Potosi, Mexico, dif- fer from those of La Cueva Chica, which is about fifteen miles distant, in that the former show a still further structural loss in eye architecture. 2. Nothing is left of the retinal tissue and all that can be found in the socket is a much reduced pigmented cyst which may represent the last of the coroid. 3. The last vestige of a connecting optic nerve is gone. 4. Along with the eye loss has come modi- fication of the skull, involving a closing in of the circumorbital series of bones, and some de-ossification. 5. Great regularity is shown in the rear- rangement of the bones of skull with the subtraction of the eye, clearly in- dicating a reconstitution of the skull as a whole in which the separate elements are without independent action. 6. Experiments with a gradient trough show that these fish are fully indiffer- ent to light as compared with the nega- tive reactions of La Cueva Chica speci- mens. 7. Current views on the biology of such degenerative changes are discussed. Table I. Tabulations of Experiments on Light Sensitivity. Each experiment represents 100 notations at 5-second intervals. Exp. Type of Number Observed % of random no. fisn of rish in light expectation 1 La Cueva Chica1 1 30 60 2 “ tt tt 1 24 48 3 (( u u 1 28 56 4 “ a a 1 63 126 5 “ “ “ 1 27 54 6 1 77 154 7 “ “ “ 1 12 24 8 (< u n 1 49 94 9 tt u a 1 53 106 10 u “ u 1 17 34 Mean 37.8 75.6 11 Cueva de los Sabinos 2 1 49 98 12 u it it 1 63 126 13 a u a 1 51 102 14 1 u u a 1 45 90 15 ° l u a a 1 80 160 16 “ “ u a 1 44 88 17 it tt tt u 1 16 32 18 “ a a a - 1 17 34 19 it a a tt 1 36 72 20 a n tt 1 89 178 Mean U9 98 21 Cueva de los Sabinos3 4 158 79 22 “ “ “ 4 156 78 23 i{ a tt tt 4 226 113 24 a tt tt 4 223 111.5 25 “ ■ a a a 4 240 120 26 tt a it tt 4 257 128.5 27 a a a 4 273 136.5 28 a a n 4 233 116.5 29 ; a tt a 4 212 106 30 u a n tt 4 161 80.6 Mean 213.9 106.9 + Mean of Exps. 11 to 30 102.5— iThis fish seventh generation of tank-reared stock, 49 mm . in standard length. 2 This fish from Cueva de los Sabinos, 46 mm. in standard length. 3 These fish from Cueva de los Sabinos, between 46 and 49 mm. in standard length. 142 Zoologica: New York Zoological Society [XXIX: 13 Bibliography. Boeke, J. 1902. On the structure of the light percep- tion cells in the spinal cord, on the neurofibrillae in the ganglion cells, and on the innervation of the striped muscles in Amphioxus lanceolatus. Proc. Sect. Sci. Kon. Akad. Wet. Am- sterdam 5: 350-358. Breder, C. M. Jr. 1942. Descriptive ecology of La Cueva Chica, with especial reference to the blind fish, Anoptichthys. Zoologica, 27 (31 : 7-15. 1943a. The eggs of Bathygobius soporator (Cuvier and Valenciennes) with a dis- cussion of other non-spherical teleost eggs. Bull. Bingham Oceanographic Collection, 8 (3) : 1-49. 1943b. Apparent changes in phenotypic ra- tios of the characins at the type local- ity of Anoptichthys jordani Hubbs and Innes. Copeia, (1): 26-30. 1943c. Problems in the behavior and evolution of a species of blind cave fish. Trans. N. Y. Acad. Sci., ser. 2, 5 (7) : 168-176. Breder, C. M., Jr., & E. B. Gresser. 1941a. Correlations between structural eye defects and behavior in the Mexican blind characin. Zoologica, 26 (16) : 123-131. 1941b. Behavior of Mexican cave characins in reference to light and cave entry. Anat. Rec., 81: 112 (Abstract No. 216). 1941c. Further studies on the light sensitivity and behavior of the Mexican blind characins. Zoologica , 26 (28) : 228-296. Breder, C. M., Jr., & P. Rasquin. 1943. Chemical sensory reactions of the Mexican blind characins. Zoologica, 28 (20) : 169-200. Eigenmann, C. H. 1909. Cave vertebrates of America, a study in degenerative evolution. Carnegie Instit. Wash. Pub. (104) : 1-241. 1917. The American Characidae. Mem. Mus. Comp. Zool., 43 (1) : 1-102. Gregory, W. K. 1933. Fish skulls: a study of the evolution of natural mechanisms. Trans. Amer. Phil. Soc., n.s. 23 (2) : 75-481. Gregory, W. K., & G. M. Conrad. 1938. The phylogeny of the characin fishes. Zoologica, 23 (17): 319-360. Gresser, E. B., & C. M. Breder, Jr. 1940. The histology of the eye of the cave characin, Anoptichthys. Zoologica, 25 (10) : 113-116. Hartridge, H. 1920. Shape of red corpuscles. Jour. Physiol. 53: 81. Hatch, M. H. 1941. An important factor in evolution. Science, 93 (2408) : 182-183. Hubbs, C. L. 1938. Fishes from the caves of Yucatan. Carnegie Instit. Wash. Pub., (491) : 261-295. Kuroda, R. 1932. Preliminary studies on the gustatory response in fishes. Acta Psychologica Keijo, 1 (4) : 157-161. Long, J. A. 1904. The reaction of eyeless (blinded) fish to light. Science, 1 9 : 248. Ogneff, J. F. 1911. Ueber die Aenderungen in den Or- ganen der Goldfische nach dreijahr- igem Verbleibein in Finsternis. Anat. Anz., 40: 81-87. Payne, F. 1907. The reactions of the blind fish, Ambly- opsis spelaeus, to light. Biol. Bull., 1 3 : 317-323. Parker, G. H. 1908a. The sensory reactions of Amphioxus. Proc. Amer. Acad. Arts Sci., 43: 415-455. 1908b. The origin of the lateral eyes of verte- brates. Amer. Nat., 42: 601-609. 1909a. The receptiveness of the verterbrate skin for light and the origin of the vertebrate eye. Science, 29 : 432. 1909b. The integumentary nerves of fishes as photoreceptors and their significance for the origin of the vertebrate eyes. Amer. Journ. Physiol., 25 (2) : 77-80. Petit, G. 1941. Sur la morphologie cranienne de deux poissons cavernicoles aveugles: Typhleotris madagascariensis G. uetit et Coecobarbus geertsi Blgr. Comp. Rend. Trav. Fac. Sci. Marseille, 1 (2/3) : 36-40. Pike, F. A. 1943a. The thermodynamic irreversibility of processes in living organisms. Anat. Rec., 87 (4) : 38. Abstract. 1943b. Entropy and the degeneration of pho- toreceptors in perpetual darkness. Anat. Rec., 87 (4) : 38. Abstract. Ponder, E. 1925a. The shape of the mammalian erythro- cyte and its respiratory function. Journ. Gen. Physiol., 9: 197-204. 1925b. The shape of the mammalian erythro- cyte and its respiratory function. Journ. Gen. Physiol., 9: 625-629. Scharrer, E. 1928. Die Lichtempfindlichkeit blinder Elrit- zen. Zeit. Vergl. Physiol., 7 (1) : 1-38. 1944] Breder: Studies on Blind Fish from Cueva de los Sabinos, Mexico 143 Tafall, B. F. O. 1942. Diaptomus (Microdiaptomus) cokeri, nuero subjenero y especie de diapto- mido de las cuevas de las region de Valles (San Luis Potosi, Mexico). Ciencia, 3 (7) : 206-210. 1943. Observaciones sobre la fauna acuatica de los cuevas de la region de Valles, San Luis Potosi (Mexico). Rev. Soc. Mex. Hist. Nat., 4 (1-2): 43-71. Thompson, D. W. 1942. On growth and form (a new edition). First Edition 1917. Camb. Univ. Press, 1116 pp. Walls, G. L. 1942. The vertebrate eye and its adaptive radiation. Bull. Cranbrook Institute Science, (19) : 1-785. 144 Zoologica: New York Zoological Society EXPLANATION OF THE PLATE. Plate I. Sections through the left orbit of specimen from Cueva de los Sabinos. Fig. 1. The central body represents the carti- laginous capsule within which is a pig- mented cyst surrounded by loose con- nective tissue. Magnification 80 X. Fig. 2. A slightly more posterior section show- ing the medial wall of the capsule to be here composed of connective tissue. At about the center of the connective tissue portion is a mass made up of muscle and possibly an atrophic nerve. Magni- fication 62 X. I BREDER. PLATE I. FIG. 1. FIG. 2. OCULAR ANATOMY AND LIGHT SENSITIVITY STUDIES ON THE BLIND FISH FROM CUEVA DE LOS SABINOS, MEXICO. Beebe: Geklconidae of British Guiana and Venezuela 145 14. Field Notes on the Lizards of Kartabo, British Guiana, and Caripito, Venezuela. Part 1. Gekkonidae.1 William Beebe. Director, Department of Tropical Research, New York Zoological Society. (Plates I-VI; Text-figures 1-6). [This contribution is a result of various ex- peditions of the Department of Tropical Re- search of the New York Zoological Society to British Guiana and to Venezuela, all made under the direction of Dr. William Beebe. The Guiana expeditions were made during the years 1909, 1916, 1917, 1919, 1920, 1921, 1922, 1924 and 1926, and the Venezuelan trips in 1908 and 1942. The latter was sponsored by grants from the Committee for Inter- American Artistic and Intellectual Relations and from four trustees of the Zoological Society, George C. Clark, Childs Frick, Laurance S. Rockefeller and Herbert L. Satterlee, and by invaluable assistance from the Standard Oil Companies of New Jersey and Venezuela.] Contents. Pag 4 Introduction 145 Gekkonidae 145 Gonatodes albogularis (Dumeril and Bibron) 146 Gonatodes annularis Boulenger 148 Gonatodes beebei Noble 152 Gonatodes booni Van Lidth de Jeude 152 Gonatodes caudiscutatus (Gunther) 152 Gonatodes humeralis (Guichenot) 153 Gonatodes ocellatus (Gray) 153 Gonatodes vittatus (Lichtenstein) 153 Sphaerodactylus molei Boettger 153 Hemidactylus mabouia (Moreau de Jonnes) ....... 155 Thecadactylus rapicaudus (Houttuyn) 155 References 158 Introduction. In the year 1909, and from 1916 to 1926, eight expeditions went out from this depart- ment to British Guiana, and in 1908 and again in 1942 field work was carried on in Venezuela. Throughout the course of these expedi- tions many field notes, color plates and pho- tographs were made of tropical vertebrates, and the object of this present series of papers is to assemble and publish these notes and illustrative material. Any change or alteration of the original notes is placed between brackets. The chief value of these data is that they are concerned with living or recently killed specimens. 1 Contribution No. 694, Department of Tropical Research, New Yoi'k Zoological Society. The observations in Guiana were made in one-quarter of a square mile of jungle at Kartabo, and those in Venezuela at or close to Caripito, which is only 528 kilometers northwest of Kartabo. In addition to numerous technical papers in Zoologica and several popular volumes, there have been published the following general ecological summaries: Zoologica : (Kartabo) Vol. II, No. 7, 1919, pp. 205-227; Vol. VI, No. 1, 1925, pp. 1-193 ; (Caripito) Vol. XXVIII, No. 9, 1943, pp. 53-59. Also see “Tropical Wild Life In British Guiana” by Beebe, Hartley and Howes, published by the New York Zoological Society, 1917, pp. 1-504. My hearty thanks go to Dr. Charles M. Bogert of the American Museum of Natural History and Dr. Karl P. Schmidt of the Chicago Natural History Museum for iden- tification and for bringing up to date my out-worn names of many years ago. My original field numbers and other data have been appended to descriptions, breed- ing and other notes. These specimens are either in the collections of the Department of Tropical Research or in those of the American Museum. In the latter case the original field numbers are still attached, with the additional catalogue numbers of the Museum. Whenever the term total length is used, a perfect, unregenerated tail is understood. Most of the figures in the plates are black and white reproductions of original color paintings, so only the pattern is preserved. The following are from photo- graphs of living specimens taken in the jun- gle: PI. II. Fig. 4; PI. IV, Figs. 9, 10; PI. V, Fig. 14; PI. VI, Fig. 16. Family Gekkonidae. ■ Eleven species of geckos were found at the two localities of study, Kartabo and Caripito, ten at the former and three at the latter. Only one of the Venezuelan spe- cies was absent from Kartabo. One of the 146 Zoologica: New York Zoological Society [XXIX: 14 latest check-lists of South American lizards (Burt & Burt, 1933) records nine species of this family from northeast South America, of which we found every one in our re- stricted areas. In addition we studied two others at Kartabo, one of which ( Gonatodes humeralis) was previously recorded only from the basin of the Upper Amazon, and the other (G. caudiscutatus) from Colombia and Ecuador. Some species of this family of small, very primitive lizards are familiar to every visi- tor to the tropics. They are not uncommon, they usually are provided with voices, and several are almost certain to be found in houses and camps situated near woods or jungle. Although perfectly harmless these geckos are universally feared by the na- tives, who call them by such names as Fathers of Leprosy and Poison Shooters. Most are nocturnal and are without movable lids to the eyes, and furnished with vacuums of sorts on the soles of the feet, enabling them to cling to and run over smooth verti- cal surfaces. Geckos breed for the most part in the long rainy season and the eggs, one or two in number, are deposited in mould or the dust of decayed logs and other suitable places in the jungle. Patient study of the habits of geckos might very probably yield interesting re- sults for they are very primitive even for reptiles. No fossil forms of the family have been found but the remains of closely re- lated lizards are known from the early Mesozoic, and we know that more than one hundred million years ago their tails were being shed as a means of safety first. Even today, geckos carry around with them such indelible proofs of their primitiveness as amphicoelous vertebrae, the presence of in- tei'centra and an occasional well-developed second branchial arch. Gonatodes albogularis ( Dumeril and Bibron, 1836). Names: Stripe-shouldered Gecko; Black- and-white-tailed Gecko. Range : Northeastern South America and the Dutch Leeward Islands. General Account: In Caripito, both in abundance, in habits and in superficial ap- pearance, this little gecko takes the place of G. annularis as we had found the latter at Kartabo, 528 kilometers to the southeast, south of the Orinoco. In notes on a collec- tion of lizards from Surinam, Van Lidth de Jeude lists ‘“Three specimens [of albogu- laris] with coloration of upper parts re- sembling closely those of G. annularis.” I found, however, no record of G. albogularis either in the Georgetown Museum or in my eight years in the field in British Guiana. My field name for temporary use was “stripe-shouldered” as compared with “spot- shouldered” for the other species and sug- gests one good distinction, this very charac- teristic marking being in albogularis in life rather a large, irregular blotch of brown or black with a vivid white streak across the middle, than the round black ocellus, more or less bounded with lighter color. The fully developed males usually possess some dorsal red, and the markings of the back and tail recall the very pronounced pattern of the females and young of annu- laris. All the dorsal markings, especially the terminal black and white caudal bands, are emphasized in newly hatched individ- uals. (PI. I, Figs. 1 and 2). Many of these lizards were taken in our pits, showing that they were active on the ground at night. The majority thus cap- tured were in Pit 13, which was the only one actually in open savanna, about thirty feet from low second growth. No other species was found in this particular pit, but for several weeks stripe-shouldered geckos fell in by ones and twos. When two were found at once, they were, except in one instance, both males or both females, emphasizing the solitary or at least unpaired habits of these lizards. An unusually large number were in Pit 13 in early April, several weeks before the rainy season began. Three indi- viduals lived in the laboratory, visible mostly at night high up on the walls. Male, not breeding, body length 30 mm., total length 66 mm. (No. 30,043, Caripito, April 11, 1942, Color Plate 1538) : Color in Life: Upper parts grayish- brown, with paler gray lines on the head and a wide vertebral band to beyond the base of the tail. A dark brown humeral blotch bisected by an irregularly crescentic, transverse band of pure white, almost meet- ing on the mid-back. Back and limbs mar- bled with shades of brown and freckled with white. At the posterior insertion of the hind limbs is a small paired imitation of the humeral marking, small jet black spots flanked posteriorly with white. These are repeated six times down the tail, spread- ing out posteriorly into broad, black and white bands. Lower labials dull lemon, chin and throat bright lemon yellow, fading posteriorly into yellowish-gray. On the chin are three large spots of rich orange, and behind these, three converging bands of the same color, freckled with brown. Iris mottled with two shades of brown. Breeding male, body 36 mm. (No. 30,100, Caripito, Pit 4, May 14, 1942) : Color in Life : Dark brown ; tip of snout white; faint patch on occiput; white line at insertion of fore limb, extending up and 1944] Beebe: Gekkonidae of British Guiana and Venezuela 147 Text-fig. 1. Gonatodes albogularis. Hyoid of 24-hours-old lizard. X15. obliquely back, almost meeting its fellow at mid-back. Lores reddish. Indian red line around antero-inferior circumference of eye; a broad red line back from eye below white stripe; another red line on side neck back to shoulder stripe; a few indistinct red spots on body. Below cold gray. Breeding female, body 35 mm., total length 75 mm. (No. 30,093, Caripito, Pit 4, May 6, 1942) : Color in Life : Pale brown on head, back and limbs. A very wide dark brown band from ear to half down tail each side of the body, the upper side of this band is zig- zag, cutting into the dorsal light ground color with a series of angled points. Legs mottled brown. In front of the shoulder is a large white blotch, surrounded by a wide light area, and this in turn by a black band, all these frames being open in front toward the head. Chin and throat dead white, sparsely dotted with dark, with a series of larger spots along the sides. Lower body and tail warm orange. On the posterior half of the tail, the brown of the body gradually changes to rings, which form the proximal boundary of a series of pale brown bands. On the under side of the tail these light bands become brilliant white. The reason for this is ap- parent when I watched this individual lizard frightened. The tail curled far up and for- wards over the back, forming a conspicuous, banded black-and-white banner. An enemy would certainly snatch first at this bright colored moving structure, and the tail then breaking off and beginning its frantic, isolation dance, the owner would escape. In the oviduct was a large, soft-shelled egg nearly ready for deposition. Text-fig. 2. Gonatodes albogularis. Hyoid of adult, breeding female. Xll. Hyoid: Newly hatched, 24 hours old, (Text-fig. 1), No. 30,018, total length 30 mm. (KOH No. 2,500, Caripito, March 1, 1944). Glossohyal long and slender (1.7 mm.). It is hyaline except for one-quarter of its length beginning near the tip, this area showing considerable calcification. The clarity of the glossohyal continues to the division into the basihyals, all of which elements show solid calcification. At the lower part of this area is a distinct shoulder or socket from which spring the first two gill arches. The hypohyal arises from the upper rim of the shoulder, extending hori- zontally as an almost straight element, slightly larger in caliber at the base, and expanding distally into two, widely sep- arated points. From the lower point a short, obliquely backward directed rod is seen, from which there arises the ceratohyal. So firmly is this pair of bones attached to the auditory apparatus that in dissection, both were torn away from their hypohyal at- tachment, remaining firmly fixed to the otic area of the skull. This extremity shows a solidified, large, flat, diaphragm-like expan- sion. From the hypohyal socket itself arises the first ceratobranchial (1.7 mm.), the junc- tion being by an enlarged, rounded, hyaline head, with considerable calcification imme- diately behind it, suggesting a hint of sep- arateness which might justify considering it a hypobranchial relic. The distal extrem- ity which is slightly calcified, gives rise to a slender epibranchial curved into a hook at the end. Although the hypohyal calcification ends abruptly at the socket, the arch ex- tends posteriorly with no change in curva- ture and only a slow narrowing, as the sec- 148 Zoological New York Zoological Society [XXIX: 14 ond ceratobranchials. These are elaborately developed and end in elongated, angular ex- tremities, and at the very tips there is a minute, clawlike bit of cartilage, probably a vestigial second epibranchial. This entire third arch is quite devoid of calcification. Adult breeding female, (Text-fig. 2), No. 30,093, total length 75 mm. (KOH No. 2501, Caripito, May 6, 1942). The hyoid of this adult female is actually only 32 per cent, or one-third larger than that of the day old specimen, and in eyeball diameter there is even less difference, an increase in the adult of only one-fourth or 25 per cent. Opposed to these organs so valuable to the newly hatched young lizard, are the relatively less important total lengths of young and adult, 30 and 75 mm., respectively, an increase of two and one-half times, or 250 per cent. The hyoid of the adult gecko differs from that of the twenty-four-hour lizard in only a few important ways. The glossohyal is rel- atively longer and more slender, and the outward curve of the hypohyals is somewhat more pronounced. In my description of the juvenile hyoid I purposely italicized the horizontal position of the hypohyals, as be- ing radically unlike that in the normal hyoid arch of adult geckos. In the present speci- men these elements have swung forward quite 90 degrees into the specialized gek- konid position. This forward shift includes the entire hyoid arch, and seems to have brought about or perhaps may be said to have been the result of a loss of direct con- nection with the auditory apparatus, so that the end of the ceratohyal is -well anterior to the end of the first ceratobranchial, and the extreme distal tip instead of being closely united to the diaphragm-like transverse element, dies out, and shows now, instead, a short separate bit of cartilage lying along- side its tip. The first branchial arch shows little change except a shortening of the epibran- chial, and the same is true of the terminal, irregular filaments of the second cerato- branchial. Breeding : On March 1, 1942, at Caripito a gecko hatched from one of six eggs de- posited under the bark of a rotten log by three or more females of this species. The eggs measured 6 by 7.5 mm., the newly hatched lizard 30 mm. over all. The color in life was dark brown with a series of orange- edged black spots down each side of the back. The tail ended in conspicuous bands, two white and two black. The tail was held high in the air from the first step after hatching, and waved from side to side when walking. March 2nd the lizard had shed and eaten its skin. From this night on, the change from dark brown to very pale body and intense black-and-white-banded tail was very marked. A second egg contained an embryo only three-fourths developed, very lively but hampered by a considerable amount of yolk. Its most conspicuous markings were the shoulder stripes. It was 28 mm. over all, and the tail was very active. The other eggs were almost fresh, so that the entire six must have represented a com- munal laying on the part of several females. Gonatodes annularis Boulenger, 1887. Names: Spot-shouldered Gecko, Yellow- throated Gecko; Wood Slave (Guiana Creole); A-tah-zick (Akawai Indian). Range : The Guianas. General Account : This is the most abun- dant of the small species of geckos at Kar- tabo. Its favorite haunts are old stumps, hollow and fallen trees and the debris be- hind half-rotten leaf spathes of palms. I once found two at a height of thirty feet in a bromeliad. They were seen now and then in the laboratory but were not as conspicuous as S phaerodactylus , nor as skil- ful in climbing vertical surfaces. They are less nocturnal than Thecadactylus and SpKaerodactylus although, like the latter, they have round pupils. They are active on cloudy days but I never saw them in bright sunlight. The food of this species consists of small insects, especially termites, as well as col- lembolas, diptera and ants. A pet monkey caught and ate one without much zest, and repeated this on three occasions. The only other observed enemy was a large marine toad which devoured two, and a young trum- peter which killed but did not eat a spot- shouldered gecko which escaped from a vivarium. These geckos are decidedly solitary and I never found a pair close together, but in three cases (twice in the vicinity of eggs) a male and a female were present in the same log. Twice I found a female actually touch- ing the eggs, one and four eggs respectively. While only a single egg is deposited at a time, I am reasonably certain that as many as four may be laid at considerable intervals in the same cavity by the same individual. Eggs are always hidden beneath soft, fine debris well below the surface. When the female was found with the eggs, she too, was buried out of sight in the damp saw- dust. Four-fifths of the eggs found were laid during the long wet season from April to September. One egg, kept under natural con- ditions, hatched after 52 days from the time of discovery, apparently indicating this re- markable length of time as a minimum period of development. In one case I found two eggs buried in a termite nest, and a newly hatched dead liz- ard, showing that unlike tegus ( Tupinam - 1944] Beebe: Gekkonidae of British Guiana and Vejiezuela 149 bis) in corresponding conditions, these weak geckos cannot free themselves when built in. Whether the throat in this species of gecko is immaculate or distinctly lined is a character quite independent of age or sex, but in the adult male it seems invariably yel- low. In sexually active males there is al- ways a large amount of red on the dorsal surfaces. The dorsal irregular spotting or banding (see figure of coloration of type, Proc. Zool. Soc. London, 1887, p. 153) is often apparent in adult females and in the very young of both sexes. In five such indivi- duals, females and young, very distinctly marked in life, not a trace remains in the preserved specimens. Male adult (No. 189, Kartabo, June 27, 1919) : Measurements : Length 68 mm., head 9, tail 34 (renewed portion 31), eye diameter 1.7, snout 4, eye to ear 3.2, fore limb 11, hind limb 16 mm., weight 1 gram. Color in Life: Head in general greenish- yellow, the lip scales all around above and below reed yellow (colors from Ridgeway’s nomenclature), face markings olive yellow. (PI. I, Fig. 3). These latter include a snout spot, an irregular line in front and another behind the eye, the head below the eye and back to and including the ear, and a small spot well above the ear. Dorsal body sur- face violet gray thickly granulated with Brazil red back to mid-body, where the red concentrates and forms two dorsal and two lateral lines of three to five large round spots. In front of the insertion of the fore limb is a large ocellus, black, banded with smoky, and with a wide vertical line in front and another behind of yellowish- white. Tail amber (broken off twice near base and grown completely again). Limbs tawny olive, the scales tipped with black. Chin, throat and lower neck bright olive ochre. Ventral body, upper arm and leg olive buff, tail from just beyond vent vinaceous fawn. Scalation: The ventral scales are flat, well separated and with numerous fine black dots. A large area, rounded, abruptly marked on the posterior belly at equal dis- tances from the vent and the insertion of the thighs, and an elongated area on the under side of each thigh, consists of spe- cialized scales, much swollen and immac- ulate. Those on the thigh number about eight long and three or four scales wide. The ventral scalation of the renewed tail is very irregular, three or four transverse scales extending clear across, being followed by a broken irregular series of small scales, then more transverse scales. Male adult, length 71 mm., tail 36, weight 1 gram (No. 782, Kartabo, May 17, 1922) : Color in Life : General color dark brown, head variegated with red and green. Dark shoulder spot bordered behind by white line. Two light brown spots at base of tail above. :Chin and throat and under parts greenish- yellow. Female breeding (No. 558, Kartabo, Au- gust 9, 1922) : Measurements: Length 75 mm., head 11, eye 2, fore limb 10, hind limb 12, tail 41.5, weight 1.1 gram. Egg about to be laid. Male adult, length 83 mm. (No. 531, Kar- tabo, March 8, 1922, Color Plate 330) : Color in Life: Head and back dark olive green with broad, rather irregular bands of carmine. The lores are wholly carmine, and three lines extend back from the eye, the lowermost curving up over the humeral ocel- lus. Labials and all four limbs light olive green. Chin and throat immaculate, apricot yellow. There is a large, roundish black spot on the shoulder with a slightly angular, very conspicuous broad white line extending transversely from the lower insertion of the fore limbs almost to the mid-back. In the same individual, preserved, only the black spot remains with all the red gone and the white line as an obscure gray shade. Other- wise the lizard is monochrome dull brown. Hyoid: Male adult, length 65 mm. (KOH No. 2018, Kartabo, June 24, 1920. (Text fig. 3). Description and drawing checked with identical hyoid of No. 2020, same length and sex) : Glossohyal or lingual process a slen- der, needle-like rod 3.5 mm. in length to the base of the forked but otherwise undif- ferentiated basihyals. The anterior 1.7 mm. of the glossohyal is covered with a broad, spear-shaped tongue core, 1.5 mm. wide Text-fig. 3. Gonatodes annularis. Hyoid. X 10, 150 Zoologica: New York Zoological Society [XXIX: 14 with two transparent posterior cornua. From their point of divergence, to the shoul- der articulating with the arches, the basi- hyals measure 1.3 mm. The proximal ele- ment of the hyoid arch, the hypohyal, is rep- resented by a slender rod of 2 mm. extend- ing obliquely forward from the basihyal, curving slightly outward for the distal half millimeter, and connecting by a close-fitting joint with a curious, skate-shaped affair. From the inner side of this, two-thirds to- wards the rear, arises the long, slender, backwardly curved ceratohyal 2.8 mm., en- larging slightly but evenly throughout its length, and ending in a truncate, wide- flanged tip close to the ventral aspect of the auditory apparatus. The articulation of the first ceratobranchial shows a distinct, strongly marked, curved head, the remains of the hypobranchial. The ceratobranchial itself follows the general direction and length of the ceratohyal, but is slightly longer. The end of this part of the hyoid apparatus is strongly curved and distinctly divided into two distal segments, which may represent the epibranchial and perhaps the pharyngobranchial elements of the first branchial arch. A third arch is well developed in this prim- itive little lizard. These second ceratofiran- chials extend back as two strong spines from the articular area of the basihyals, completing with them a general wishbone shape. From the end of each depends a long, delicate, thread-like strand of tissue, of per- haps degenerate epibranchial origin. A cleared hyoid of Gonatodes caudiscuta- tus is identical with the above, except for a reduction in the size of the tongue core. Breeding : On August 23, 1919, an egg of this gecko was found in a broken palm stub close to the laboratory at Kartabo. An adult female lizard was a foot away and was ac- cidentally killed when captured. The egg must have been deposited some time before, as it hatched the same day. I watched the lizard (Cat. No. 229) break the shell, uncurl and walk away along the edge of the glass dish. The egg was broken off irregularly, about one-quarter of the shell coming away in two large pieces. The egg measured 7 by 8 mm. and five minutes after hatching the lizard was 38 mm. in total length, the head being 8 mm. When slowly approached the lizard watched intently, its eyes moving independently, but the little creature remained motionless, except for a to and fro motion of the tail tip. This never ceased. When I moved, it scurried off with short, quick darts, the last remains of the yolk sac falling away, while a small bit of shell stuck to its throat for some time. When running, the tail was lifted in a Jong high curve. The color was dull bluish-gray, faintly marked with small grayish-white spots on the head, and a series of ten larger sepa- rated spots extended in pairs down tne back, joining on the rump and becoming bands on the tail. All these spots were pre- ceded by dark areas. The most conspicuous marking, as in the adult, was the transverse prehumeral whitish lines, enclosing a large black spot. Beneath, the color in general was uniform bluish-gray, with the sides of tne head, chin and throat marbled with shades of gray. The labials were very conspicuous, black with small, light-colored centers. The pupil was round, the iris with a bright, nar- row outer ring of gold, the rest of the iris finely mottled with dull golden brown. The new-hatched lizard seemed to get grayer and this was seen to be due to tne loosening of the epidermis. Within half an hour after emerging from the egg, the skin began to peel off, beginning at the digits and working up the limbs and toward the tail, revealing beneath, intensely contrast- ing shades of gray and white. Within the first hour the entire skin had been pulled off and swallowed, finishing with a long shred from the tail. Whenever a bit of skin or dirt got on the eyeball, the tongue was run out sideways and quickly passed over the whole eye, thus taking the place of the eyelids. On July 18, 1920, five eggs were uncov- ered in three lots in the decayed stem of a spiny palm. (PI. II, Fig. 4). The shells were hard, smooth and dull white. Weights aver- aged .15 of a gram, sizes 7 by 7.5 mm. In the case of the two pairs, each egg was separated from its fellow by about a week’s development. In the two older ones which were about one-third developed, the embryo, except its head, was completely embedded in the thin yolk. The opaque blue-black eyes were of enormous size, (Text-fig. 4), pear- shaped, half as large as the entire head, which in turn was equal to or larger than all the rest of the embryo. The tail was rolled tightly, the four feet with their short, blunt digits all pressed close together. - On April 29, 1922, an egg was washed out of a hole among roots in the river bank up-river from the laboratory. The embryo was within five days of hatching. This egg was unusually small for this species, 6.7 by 7 mm. The color pattern of the embryo was like the newly-hatched specimen (Cat. No. 229) with the shoulder spot even more em- phasized, and in addition a pair of post- sacral, black-and-white ocelli. (Text-fig. 5). On July 23, 1922, four eggs of this species -This specimen shows, in addition to the large eye, two irregularly linear groups of dark chromatophores, one at the anterior part of the body, behind the eye, and the other at the beginning of the tail. These may very possibly be the first hint of the two pigmental concentrations so dis- tinct in the late embryo. 1944] Beebe: Gekkonidae of British Guiana and Venezuela 151 Text-fig. 4. Gonatodes annularis. Early embryo Text-fig. 5. Gonatodes annularis. Late embryo. were found close together in the debris at the bottom of a decayed bamboo stub close to the laboratory at Kartabo. The shell of these eggs is thicker and tougher than that of a hummingbird. The average size was 6.5 by 7.4 mm. and the weights .25 of a gram. The first egg hatched August 17, and when 24 hours old the lizard itself weighed .25 of a gram. It had not fed, except on its own skin, and the increase in weight must have been from the absorption of water. A second egg, opened on the same date, was one-third developed. The third egg, also opened on the same date, was four-fifths developed. In this embryo the shoulder ocel- lus was clearly marked. The fourth egg hatched on September 13, at least 52 days after the egg was laid. Unlike the newly hatched gecko of August 23, 1919, the skin of this one did not loosen until six hours after emergence, and the method of sloughing was quite different. The epidermis of the entire head, body, limbs and tail became detached, the general color changing quickly to a pale smoky gray. When first observed the loose epidermis was pushed back from the head, standing out in a ruff close behind the jaws. The gecko crawled about his cage, creeping under the lumps of sawdust and bits of bark in which the eggs had been found, pushing and work- ing till the skin hung in tatters from his body. By turning his head far around, he at last got a piece of this in his mouth and tore off as much as would come and munched it down. He thus gradually freed his body and disposed of the debris, then pushed and rubbed again till the skin at the base of the tail was frayed. He made a tremendous effort, coiling around even beyond a com- plete circle in order to get a firm grip on the tail covering at its free edge ; then he slowly straightened out and the skin peeled cleanly off in one piece, which made his next course. His legs were still completely encased, but he stripped them free very quickly and neatly as gloves are pulled off backward, each little toe covering being perfect as the owner gulped it down. The entire process occupied a quarter of an hour. I kept the gecko alive for two weeks in a vivarium, during which time I fed it on eight termites. It grew very little. The com- 152 Zoologica: New York Zoological Society [XXIX: 14 parative measurements are as follows, the first taken when twenty-four hours old, the second when two weeks of age: Total length 38 and 43 mm., head 6 and 7 mm., tail 20.3 and 22 mm., weight .25 of a gram and 3 grams. Go natodes beebei Noble, 1923. Name: Blue-eyed Chestnut Gecko. Range: British Guiana. General Account: This field name, which we used before identification, gives an ex- cellent general idea of the appearance of this lizard. It is known from two specimens and two color plates. The lizards are in the collection of the American Museum of Nat- ural History. Type: N.Y.Z.S. No. 311, American Muse- um No. 21251, adult male, total length 94 mm., August 26, 1920. Color Plate 215, dorsal view, nat. size, head and neck en- larged; Color Plate 1175, iris. Topotype: No. 38972. (PL III, Fig. 5). Noble’s account of the coloring of the type after three years in preservative is as follows: “Uniform reddish brown above, whitish immaculate below, except for a slight suffusion of brown on the abdomen and appendages, this suffusion tending to form dark edges to the scales of the ventral surface of the thigh.” His measurements of the type are as follows: “Tip of snout to vent 47 mm., tip of snout to ear 11.5, tip of snout to orbit 5.5, greatest width of head 7, vent to tip of tail 47 mm.” Color in Life of Type: Above chestnut shading through orange rufous on the sides to cadmium yellow on the chin and throat, and vinaceous russet on the ventral surface. A scattering of small black dots on back, tail and upper limbs. Iris light cerulean blue, and except for an inner pupil ring stippled everywhere with dark gray. Pupil round. Caught on a fallen tree trunk in high jungle, five hundred yards from Kartabo laboratory. Kept alive for four days. Quiet, feeding well and amenable to handling. It frequently cleaned its eyes by running its tongue out and over them. Gonatodes booni Van Lidth de Jeude, 1904. Name: Yellow-spotted Gecko. Range : Known only from Surinam and British Guiana. General Account: On April 15, 1924, I saw a species of gecko new to us, appearing brown and covered with small yellow spots. It escaped all our efforts at capture, running beneath an overhanging bank of earth in the jungle about three hundred yards from the laboratory at Kartabo. A month later, on May 16, I caught the same or another yellow-spotted gecko in the same bank, dig- ging it out of a twelve-inch hole into which it had rushed. A week later, on May 24, I saw another individual of the same species only ten feet away from where we had taken our specimen. It eluded us. (PI. Ill, Fig. 6). Male breeding (No. 2833, Kartabo, May 16, 1924, Color Plates 740 and 741) : Measurements: Length 88 mm., head 12, body 31.5, tail 44.5, head width 8.5, body width 9, eye 2.5, fore leg 17.5, hind leg 28.5, testes 4.5 mm., weight 2.5 grams. Color in Life: Background of upper parts of head and body blackish-brown, covered on head, face, throat, back and upper sur- faces of fore limbs with rather closely spaced, irregular spots of empire yellow. These become duller and somewhat smudged on snout and anterior chin. Tail, hind legs and all four feet light russet-vinaceous, mot- tled with darker. The tail coloration extends up the mid-back in a wide, fairly definite line. Ventral surface primuline yellow under shoulders, shading backward to uniform tawny on under body and limbs, and russet vinaceous under tail. Inner rim of iris ceru- lean blue, remainder black flecked with light blue. (PI. Ill, Fig. 7). A changing phase of color shows all the spots much more brilliant and intensely orange, and the russet areas become decid- edly orange in tone. The only other specimen known is the type, collected in Surinam and now some- where in Holland. Its color after preserva- tion is described as “Upper parts dark green, covered all over with larger or smaller lighter spots composed by one, two or more granules of a light green color. Belly lighter without spots; gular region with large greenish spots. Tail without spots, below lighter than above.” Gonatodes caudiscutatus (Gunther, 1859). Name: Ladder-back Gecko. Range. Colombia, Ecuador and British Guiana. General Account: No males were seen or secured and only three females. One of these was about to lay an egg on August 25, 1920. The dorsal markings were sufficiently dis- tinct from those of related species to war- rant our tentative field name of Ladder- backed, given long before I knew the actual taxonomic species. The vertical white shoul- der stripe seems to be placed more posteri- orly than in other species. One specimen possessed a very marked ability to change the shade of pattern, the black dorsal mark- ings at night becoming almost invisible. On being annoyed the pattern returned in full strength. Female adult (No. 598, Kartabo, May 21, 1922, Color Plate 366) : 1944] Beebe: Gekkonidae of British Guiana and Venezuela 153 Color in Life: General color above gray- ish-olive, with irregularly longitudinal streakings of dark brown on head. (PI. IV, Fig. 8). A broad vertebral stripe beginning at base of neck vinaceous buff, extending to insertion of tail. Two small, square patches of black almost meet over the shoulders across the dorsal stripe. The stripe is cut into along each side by nine symmetrical pairs of black points. Shoulder ocellus a large roundish spot of black, surrounded by a band of the general olive background, and this in turn by a still wider black frame, broken only on the posterior rim. Here a crescent-shaped narrow white line extends from the insertion of the fore arm to the mid-back. Side of head olive with fine darker markings and a lateral black line from lores through eye and on backward in a curve above tympanum to shoulder spot. Another black line from upper posterior eye back- ward and upward to top of head. Lips light olive with a line of black mottling along the lower labials. Limbs strongly mottled with dark brown. Tail with fine, irregular longi- tudinal markings of dark brown becoming black at tip. Sides of body finely mottled with olive, with two longitudinal lines of dark brown grouped in lengthwise dark spots. Ventral surface dirty white on chin, cream color on under body, mottled with gray under limbs and between fore arms. Under tail salmon buff. Eyeball ochraceous buff, pupil round, iris dark mottled walnut brown with orange tawny border. Gonatodes humeralis (Guichenot, 1855). Name : Black Wood Slave. Range : Peru, Brazil and British Guiana. General Account : Only five specimens were found at Kartabo and in the field they were confused with other species of the genus. No unusual habits were recorded, but the dark color was observed as setting them apart. In spite of considerable hand- ling, the tails of this species seemed less deciduous than those of their fellow species. Color in Life : Two types of patterning were observed, apparently distinguishing the sexes. An adult male, length 80 mm. (No. 3084, Kartabo, September 7, 1919) was very dark brown above, slightly paler on chin and throat, becoming still lighter un- der body and tail. The humeral spot was very faint in life and quite lacking in the preserved specimen. A female, 60 mm. long (268, No. 3117, Kartabo, August 3, 1922) was light brown with a pronounced humeral ocellus, a round, central black spot, sur- rounded by a ring of pale brown ground color and an outer frame of grayish-white. The back was plain but with a pair of light brown sacral spots, and six wide, pale caudal bands. Below grayish-white with faint, con- verging bands on throat. Gonatodes ocellatus (Gray, 1831). Name: Ocellated Gecko. Range: Northeastern South America, To- bago and Trinidad. General Account: No. 519, and a second in the American Museum of Natural His- tory, No. 38784, appear to be the first re- corded from British Guiana. I have no defi- nite notes in regard to habits. Male, body length 50 mm., tail gone (No. 519, Kartabo, March 21, 1922) : Color in Life : General color of head above and on sides and neck Brazil red. Chin and throat brilliant orange. Top and sides of head covered with elongated patches (two running up and back from eye) and mark- ings of capucine yellow edged with black, while on the ventral surface chin, throat and neck these markings are rounded spots. A line of the same yellow extends across the back from shoulder to shoulder, touched with black and slightly broken in the mid- back. General color of sides oil yellow, finely mottled with Brazil red and black. One large and one small ocellus on each side of the body. Broad band down the middle of back and the limbs seal brown. At night the back and limbs change to a frosty gray. Ventral surface dusky slate violet, except for a small patch of terra cotta at base of neck and a median line under the tail of pompeiian red. Pupil round, iris rim apricot buff, remainder hazel with dark stippling. Gonatodes vittatus (Lichtenstein, 1856). Name: Gray-striped Gecko. Range: Colombia to British Guiana, Trin- idad and the Dutch Leeward Islands. General Account: These geckos were of so indefinite a pattern and coloration that in the field they were not given a name or set apart from the females of annularis. At least five individuals were recorded, taken March 28, 1919 (2 specimens), July 4, 1920, July 16, 1920 (2 specimens), and June 4, 1922. Sphaerodactylus motei Boettger, 1894. Names : Black-and-white-headed Gecko (male), Orange-tailed Gecko, Gray-headed Gecko (female); Striped Wood Slave (Gui- ana Creole name) ; Shallee-shallee -(Akawai Indian) . Range: Northeastern South America and Trinidad. General Account: This gecko shares with the larger Thecadactylus whatever advan- tages are yielded by life in human buildings. It was often found in both our Kartabo and Caripito laboratories, running at night over the walls and up the legs of our tables. It 154 Zoologica: New York Zoological Society [XXIX: 14 creeps very slowly when stalking prey and at other times runs with short, quick darts. When frightened its short legs become in- adequate, and progression is by a frantic sinuous wriggling, almost Leposoma- like. The tail breaks off very easily and the break is clean with almost no muscle ends show- ing, and on the body end there is usually sufficient overlapping skin to fold in and almost hide the fracture. The tail end, reasonably enough, is less protected by ex- cess skin. I made no detailed observations on the shape of the pupil in this species, taking for granted that it is always round, but in several photographs taken immediately after death one or both pupils appear as slightly vertical. Male breeding, body 25 mm., tail 29 (No. 209, Kartabo, April 5, 1919, Color Plate 163). Color in Life : Top and sides of head and neck jet black variegated with creamy white, chiefly in the form of two broken lines from the nostrils back over the eyes, forming an indefinite “Y” on the neck. (PI. IV, Figs. 9, 10 and 11). Also a solid broad white line from the eye back to the abrupt ending of the black nuchal area in front of the fore limb. Body buffy green. Limbs very light, dotted thickly and irregularly every- where with dark brown. Tail abruptly och- racheous-orange, the color resulting from many large spots set close together both above and below. Rostral, all labials, lower side of head and neck, the chin and throat yellow-green, remainder of ventral surface of body creamy-white. Pupil round, iris ivory white, finely dotted with large, dark brown crescentic marks, especially dense in front of and behind the pupil. A 49 mm. male taken April 11, 1919, wholly lacks the central dorsal cephalic white spots, and the body freckling is con- fined to the sides and pelvic region. Male, not breeding, body 25.5 mm., tail 27.5 mm., (No. 538, Kartabo, April 13, 1922, Color Plate 358) : Color in Life : General color above deep olive buff tinged with apricot orange on head and merging on the tail into capucine yellow. Groups of tiny picric yellow dots on orbits and before and behind orbits in vague lateral lines. All upper surfaces covered with scattered small spots of pecan brown, very faint on neck and fore limbs. These spots become heavier and darker brown on posterior body, and still heavier and more irregularly spaced and elongated on anterior tail. Beyond mid-tail they die out, the re- mainder being faintly mottled with gray. The dorsal spots arrange themselves either side of a plain narrow vertebral line, giving the effect of a dorsal body stripe. Side of head primuline yellow. Broad lateral stripe of picric yellow from eye diagonally upward to sides of neck. This marking is strongly edged with black along each side, which color in turn shades into orange outer bands. The lower band extends forward through the eye, becoming fainter on the snout. Chin lemon chrome; ventral surfaces maize yel- low. Under tail zinc orange. Toes grayish. Pupil round. Inner iris rim ivory white. Rest of iris white with grayish stippling on outer edges. Fore and aft a wide band of walnut brown continues through the iris the loral-postorbital band. Female, not breeding, length 46 mm. (No. 209a, Kartabo, April 11, 1919) : Color in Life : In general buff with the head and neck rather indefinite gray. Ap- parently unmarked except for two light flank lines, down each side of the dorsal pel- vic region. In the right light, however, all the broad cephalic lines of the male are vis- ible as lighter brown— i. e., the two lateral lines and the central “Y”. The lateral lines extend very faintly down the body and, be- coming strong at the pelvic zone, die abrupt- ly at the caudal change to coarser imbricated scalation, which marks the area of regener- ation. Color in Life of Young: Young geckos show a uniform pattern. Many, such as one 30 mm. in length, taken May 4, 1922, are light buffy brown, with no trace of the dark cephalic bands of the adult. Two wide, pale brown bands start from the eye and extend back just above the limbs to the tip of the tail. These are faintly bounded with black. Here and there down these lines and also down the center of the back, are very small, brilliant, isolated yellowish-white scales. On the posterior half of the tail these in- crease in number, gather some black scales about them and form four light bands, the tip of the tail being almost white with a black sub-terminal band. A very young gecko, 23 mm. total length, May 16, 1919, has the head dark grayish, body and tail olive. The eye to tail tip bands are very pronounced. There is a row of small, pale vertebral dots. Near the base of the tail are two pale lateral spots; beyond these a second pair almost join, and still farther are three pale creamy white bands, the most conspicuous markings on the young lizard. At night all markings except these caudal bands disappear. Food : Small, winged wood roaches, red mites, ants and termites. Enemies: On the night of July 3, 1920, by flash-light I surprised a ghostly white Thecadactylus on a jungle tree trunk with a small gecko in its mouth. My quick grab secured only the victim and the wriggling tail of the larger lizard. The unfortunate 1944] Beebe: Gekkonidae of British Guiana and Venezuela 155 gecko proved to be a 28 mm. Sphaerodac- tylus molei. Breeding : No. 1934 was an oval egg, 5 by 7 mm., collected July 30, 1919, in a rotten stump, together with the female lizard which was about to deposit a second fully- formed egg. The egg in the sawdust was very stained and slightly dented, so I thought it was bad, but the embryo proved to be in perfect health and would have hatched in three or four days. It was packed tightly in the shell, and measured 33 mm. in length over all. The color was a dull mottling of several shades of brown with the two lateral body lines and caudal black and white bands well marked. Four separate times I placed it in absolute dark- ness in the photographic room, examined it after returning it to the light, and then re- examined it after exposure to strong day- light. In each case the gray mottling and to a less extent the lateral body lines became so faint that they were barely discernible. There was no change in the caudal bands. The embryo lived for five hours, and the color change tests were carried on in the first half hour, before I removed the lizard and measured it. In the case of this female there was a single large egg in the oviduct, and no trace of a less developed second. This was the case in three other instances, but not with gecko No. 2652, adult female, body length 28 mm., taken March 16, 1924, which was about to shed its skin. When I dissected this specimen I found two fully formed eggs about to be laid, 5.5 by 7 mm. The lizard was of typical color, buffy brown with gray head, and faint head lines. Hemidactylus mabouia (Moreau de Jonnes, 1818). Name : White Gongasacka. Range: Northeastern South America in- cluding Brazil, and the Antilles. General Account: No notes were made of this species except that at least two speci- mens were taken, one of them in July, 1920, in the laboratory at night, appearing dead white unmarked even by black tail bands. No. 195 was found on a tree trunk Septem- ber 25, 1917. In general large size and in the peculiar character of the vertical pupil this gecko recalls Thecadactylus rather than any of the genera of small lizards. Thecadactylus rapicaudus (Houttuyn, 1782). Names: Large House Gecko, Cat-eyed Gecko; Gongasacka (Guiana Creole Name) ; Cang-gah-sah or Kingasah, “One-who-calls- in-the-house” (Akawai Indian). Range: South America except southern part; Central America and the West Indies. General Account: Almost anywhere in the warmer parts of the neotropical region, when human beings first occupy a house they will find that two small native creatures have preceded them. One is the little tree frog, Hyla rubra, who considers the kitchen sink or a bath tub or rain barrel the equal of any jungle home. The other is the gray house gecko, Thecadactylus rapicaudus. Both are perfectly harmless, they are nocturnal, they can cope with smooth perpendicular surfaces by means of vacuum soles to their feet, and both have pleasant voices which are heard, now and then, throughout the night. The cat-eyed gecko is common both at Kartabo and Caripito. Its normal haunts are decayed logs and hollow trees, and it also approves of the cavities behind the leaf stalks of palms. Half a dozen were always to be found in the laboratory at Kartabo, hidden away in solitai'y obscurity during the day, but creeping forth when our insect- attracting lamps were lit. Out of dozens of these geckos collected, the following measurements of a male with a perfect tail are average. They are per- centages of total length. No. 201, Kartabo, adult male, Dec. 23, 1920: Total length 144 mm., body 53 per cent., head 18, head width 10, snout to eye 7, eye diameter 3, fore leg 16, hind leg 21 per cent., weight 9.3 grams. Color in Life: Adult. The thirteen words with which Boulenger ( Catalogue of the Lizards in the British Museum, I, p. 112) characterized this gecko can hardly be im- proved upon. “Brown or greyish-brown above, variegated with darker and lighter; lower surfaces whitish, immaculate.” The variety of mottling and spotting in adult geckos is infinite. The most consistent marking is the olive buff of the labials, and a broad, irregular, pale olive brown stripe from the eye to the shoulder, bounded above and below by equally wide bands of dark brown. The complexity of the pigmental characters in general is increased by the occasional temporary appearance of fairly regular bars and spots down the back, at times when the creature is excited or alarmed, with a subsequent return to the heterogeneous pattern of calmer emotions. The tail is the most variable part, if such a superlative is possible, and no two are alike. Some are lined with a multitude of thin, hairlike, dark streaks, or waved in- definitely with several shades of gray or brown, or a dark background will have broken bands of lighter. Regenerated tails are usually without definitely contrasting markings. Immature geckos often show very regular markings, such as No. 30,009, under the next heading. Change of Color: Temperamental change of color is sometimes very marked, but is never so radical as that from diurnal to 156 Zoologica: New York Zoological Society [XXIX: 14 nocturnal color and pattern. This is especally pronounced in young and half-grown indivi- duals. Typical was the change in No. 30,009 (Caripito, March 15, 1942, Color Plate 1508.) This immature specimen, which measured 125 mm. over all, was captured in the jungle under bark. In general it may be called vinaceous, with the top of the head, limbs and dorsal markings of body and tail light wine-colored, elsewhere darker. The two most brilliant areas are the golden eyes, and a narrow, single-barbed dead white line from the eye back along the side of the head. The head is coarsely dotted, the limbs blotched with dark vinaceous. Down the back are very conspicuous mark- ings, light violet, divided narrowly into three, each of which may be described as a broad arrow head, directed posteriorly, with two extra pairs of barbs down its shaft, all surrounded and emphasized by a frame of black. The basal half of the tail has three wide bands of light violet, then an exceed- ingly wide band of black and a pale tip. (PI. V, Figs. 12 and 13). The lizard escaped in the laboratory and when caught late in the evening bore no pigmental resemblance to its diurnal phase. The whole animal had become pale pinkish- white, except for the dorsal markings which had paled to ivory white. The two conspicu- ous exceptions to this fading were the eyes and the broad black tail band which re- mained a broad black tail band. Color change in another individual was less marked. A female, 112 mm. in total length, every night at 8 o’clock turned a brownish-white, with several small, irreg- ular dark marks down the tail, marks which were quite absent in the diurnal patterning. After having been covered all night, at 6 in the morning the color was unchanged except for a decided tinge of olive green. After ten minutes’ exposure to daylight the green increased in intensity, and a dark ground color began to appear, a deep olive brown, lighter colors being confined to ir- regular streaks on the head and a series of haphazard spots down the back. This sequence of color change was always the same whether the lizard was uncovered at 8 A.M., noon or 3 P.M. The change was un- affected by the regular rhythmic change of the twenty-four hourly shift from light to darkness, but influenced only by direct illumination on the gecko. Tail: This organ is strongly prehensile, and when the lizard is clinging motionless to glass, the belly and tail are pressed so closely against the surface that it seems as if they must give material aid to the vac- uum feet. Yet this gecko can walk on a dry, vertical glass surface with the body and tail raised clear. At moments of excitement the tail is curved high in the air and waved slowly to and fro, doubtless as lure or bait to any enemy; an easily dislocated “pound of flesh” offered in exchange for the chance of escape. When the tail is detached there ensues a frantic twisting and wriggling, so violent that the entire tail flicks into the air again and again, before it slowly dies its isolated death. The base of the tail is swollen and the ultimate zone of possible fracture begins at the posterior end of this swollen area, about 10 mm. behind the anus. On each side of the vent is a double tooth or spine-like scale. The regenerated tails lack much of the pre- hensile and coiling ability of the original and can never lie as flatly and as closely applied to any surface. Eye : The eye of this gecko is magnificent. (PI. V, Fig. 15). It is very large, adapted for nocturnal sight, perfectly round and with the lids reduced to tiny folds. The nic- titating membrane, which is crumpled into a useless fold in our own eye, is here a convex “watch crystal” of hard, transparent tissue within which we can see the eye ball rotate freely. But though lids are absent, yet the eye is protected from excessive light by the great mobility of the pupil. In day- light there is visible only a narrow, vertical slit with opposing nicks which when pressed together leave four very small diamond- shaped openings, admitting sufficient light for the detection of danger and the per- formance of all necessary activities. The rest of the eye, the iris, in breeding males, is a splendid ball of glowing coppery gold, shot and zigzagged with veins of rich warm brown. In most females and less developed males, the iris is bright silver. Tongue: The tongue is moderately pro- trusible and mobile. When some sudden alarm has passed the gecko will often “lick its chops” six or eight times, the tongue curling up and over the lips. In males in full breeding condition (No. 2923, Kartabo, Plate 759) the rounded tip of the tongue is a conspicuous bluish-violet. Vague hints of an elaborate courtship which have come to notice make it possible that these lingual colors may play a part. As in other geckos, the tongue functions as eye-lids and from time to time is passed over the eyes, clean- ing them of any dust or other detritus. Male, length 135 mm. (Kartabo, ' KOH No. 2034, July 5, 1920) : Hyoicl: Characterized by the pronounced curve of the ceratohyal and the absence of any hint of a second branchial arch. The basihyals send forward a long, slen- der lingual process, the glossohyal. The pos- terior bifurcations of the basihyals are wide spread, but short and not very thick. A tripointed core of dense cartilage marks 1944] Beebe: Gekhonidae of British Guiana and Venezuela 157 Text-pig. 6. Thecadactylus rapicaudus. Hyoid. X5. the juncture. The hyoid and first branchial arches arise from the tips of the basihyals, the former at right angles, and the latter continuing the arms of the basihyal forks. The hypohyals are wholly cartilaginous and bend sharply out and backward, end- ing in a hyaline tip. At right angles pos- teriorly to this transparent tip, arise the ceratohyals, long, slender and cartilagin- ous, and bent outward and upward into the form of an inverted half heart. The distal extremity is widened and flattened into a shoe of considerable size. The first bran- chial ai’ch continues the line of direction (obliquely backward) of the basihyal bi- furcations. This arch consists of the cera- tobranchial, a very long, slender, outcurv- ing rod with a strong core of bone. At the proximal point of juncture there is a tiny nodule of bone, representing the hemi- spherical cap of Sphaerodactylus and Neusticurus; probably the remains of a hypobranchial. The ceratobranchial term- inates in a separate, slender, pointed seg- ment of hyaline cartilage representing the remains of the first epibranchial. Measure- ments: Basihyal length 1.7 mm., width of forks 2.2, hypohyal 2.5, ceratohyal length (measured straight between tips) 4.5, ceratobranchial length 8 mm. Checked with a second specimen, No. 3081 from Caripito, no difference is seen except that the ceratohyals are consider- ably less curved, more nearly parallel with the ceratobranchials. In this specimen, too, the end of the epihyal makes a narrow loop and connects directly with the area near the auditory apparatus. The more curved character of the ceratohyals in the Kartabo lizard may well be a mechanical contraction owing to loss of otic connection and a subsequent lessening of longitudinal strain. Feet and Progression : (PI. VI, Figs. 16 and 17). The movements when these geckos are undisturbed are very slow and deliberate, recalling the ponderous prog- ress of an elephant, the very antithesis of the nervous, waving, shaking little hands of a Cnemidophorus lizard. The raising and lowering of the feet is an interesting proceeding. At each step the whole foot is pressed down firmly, but without any sud- den or abrupt movement, and there is nothing to indicate that the sub-plantar, vertical plates are developing most efficient vacuums. The loosening of this grip is amazing and unexpected, one which gives the same shock as the abnormal, boneless- appearing feats of an accomplished acro- batic contortionist. Starting with the very tips, all five toes slowly curl and roll up and back, until they are curved flat against the wrist or ankle, transforming the penta- dactyl foot into a small, compact round ball whose surface is composed of backward- curved lamellae. The limb advances, is pressed down again and all five toes uncurl and are placed firmly on bark, wood or glass, as the case may be. This perform- ance is completely lost to the eye when the gecko moves rapidly. We can illustrate what takes place by assuming that the back of our hands are the palms, and then walking them along a table top, with fin- gers extended and curled tightly in turn. Before I ever saw this in life I wondered why these lizards so often died with their digits curled far backward. Judging by the similarly flexed toes of preserved Hemi- dactylus mabouia, I should expect to find the same method of devacuuation. Voice: Now and then in the night we hear the call of this gecko in the laboratory at Kartabo, a high, rapidly reiterated chick-chick-chick or chack-chack-chack, fif- teen or twenty times repeated, much like the note of some insect. It seems possible but not probable that besides being a sex- ual call this may be mimetic, alluring in- sects which may have somewhat similar notes. I should guess that the sound is mechanical, produced by some action of tongue and palate. Food: These geckos will eat almost any insect they can catch. Several individuals learned to come every evening to our lab- oratory tables at Kalacoon and take small moths from the tip of our forceps. I once fed two of them some immature vile- smelling hemiptera and while the insects were swallowed, it was with reluctance and with subsequent licking of lips and rub- bling of the sides of the mouth. The insects were very decidedly distasteful. 158 Zoologica: New York Zoological Society [XXIX: 14 At Kartabo I once heal'd the clatter of an empty cigarette tin and the racket keep- ing up I went to investigate. The tin ap- peared to be jiggling about under its own steam. Beneath, I found a large gecko and a winged roach of largest size in a rough- and-tumble fight, which kept up after I had exposed them. The lizard had the insect by the head, but the great flapping wings prevented further ingestion for some time. The tin had been disturbed and fallen over the combatants but had not interefered with the row. The stomach contents of four Kartabo geckos were as follows: (1), termite worker, beetle, red mite and moth fly; (2), many small insects and several spid- ers; (3), mollusk shell, small beetle, 3 moths, 2 flies, 2 flying ants; (4), 3 small wood roaches and 4 moths. Fighting and Courtship : Whenever a full-grown gecko meets a small one of his own species, the latter shows his nervous- ness by a tremulous waving of the tail and any continued threat of approach results in headlong flight. Three separate times I saw adult geckos meet each other by acci- dent in the laboratory, when there ensued much vigorous tail waving and licking of lips on the part of one, while in the other these actions were absent or were kept in low gear. On one occasion, one of the two chack-chacked three times in quick succes- sion. Either because of my presence or the illumination from my lamp or flash, noth- ing further resulted. Enemies : I can tell of only one enemy, and that of my own contriving. On April 6, 1919, I placed a Thecadactylus within reach of a pet cebus monkey. It was seized and when the tail fell off, the lizard was dropped and the monkey concentrated on the wriggling, dancing tail, which he grasped and ate. Ultimately he also caught and devoured the more quiescent owner, which had failed to make the most of his opportunity and get beyond the radius of the tethering chain. Breeding-. Female No. 3282, Kartabo, August 25, 1922, laid an egg in her viva- rium. It was a very broad oval, 13.4 by 11.4 mm. and weighed 1 gram. The shell was hard and white originally, but when found it was completely covered with a fine mosaic of debris, bits of quartz, cast gecko skin and insect remains, giving it the ap- pearance of a finely mottled, gray-brown egg. The debris was sunk deep into the shell structure and firmly fixed, as if the surface had been soft when the egg was laid, perhaps also mucilaginous, and the egg had been rolled about until thoroughly encrusted. References. Beebe, W. 1919. The higher vertebrates of British Gui- ana, Amphibia, Reptilia and Mam- malia. Zoologica, Vol. II, No. 7, pp. 205-227. 1925. Studies of a tropical jungle; one- quarter of a square mile of jungle at Kartabo, British Guiana. Zoologica, Vol. VI, No. 1, 1925, pp. 1-193. Beebe, W., G. L. Hartley & P. G. Howes. 1917. Tropical wild life in British Guiana: Part I, by William Beebe, pp. 23-290. Published by the N. Y. Zoological Society. Boettger, O. 1894. A preliminary list of the reptiles and batrachians of the Island of Trinidad (by R. R. Mole and F. W. Urich) with descriptions of two new species (by O. Boettger) .Jour. Trinidad Field Nat. Club, Vol. II, p. 80. Boulenger, G. A. 1887. On a new Geckoid Lizard from British Guiana. Proc. Zool. Soc. London, 1887, p. 154. Burt, C. E., & M. D. Burt. 1931. South American Lizards in the collec- tion of the American Museum of Nat- ural History. Bull. Amer. Mus. Nat. Hist., Vol. LXI, Art. VII, pp. 227-395. 1933. A preliminary check list of the Lizards of South America. Trans. Acad. Sci. St. Louis, Vol. XXVIII, No. 1, pp. 1-104. Dumeril, A. M. C., & G. Bibron. 1836. Erpetologie general ou histoire natu- relle complete des reptiles. Vol III, p. 415. Gray, J. E. 1831. A synopsis of the species of the Class Reptilia. (In “The Animal Kingdom, by the Baron Cuvier, with Additional Descriptions,” etc., by Edward Grif- fith. Vol. 9: The Class Reptilia, ar- ranged by the Baron Cuvier, with spe- cific descriptions by E. Griffith and E. Pidgeon). At end of volume, p. 51. Guichenot, A. 1855. Reptiles (In F. de Castelnau’s “Ani- maux nouveaux ou rares recueillis pen- dant l’expedition dans les parties cen- trales de l’Amerique du Sud, de Rio a Lima”, etc.), p. 13. Gunther, A. 1859. Second list of cold-blooded Vertebrata collected by Mr. Fraser in the Andes of Western Ecuador. Proc. Zool. Soc. London, 1859, p. 410. Houttyn. 1782. Zeeuwsch. Genoot. Wet. Vlissingen, Vol. IX, p. 232. 1944] Beebe: Gekkonidae of British Guiana and Venezuela 159 Lichtenstein, M. H. C. 1856. Nomenclator reptilium et amphibiorum Musei Zoologici Berolinensis, p. 6. Moreau de Jonnes. 1818. Bull. Soc. Philom., 1818, p. 138. Noble, G. K. 1923. New lizards from the Tropical Re- search Station, British Guiana. Zoo- logica, Vol. Ill, p. 301. Van Lidth de Jeude, T. W. 1904. Note II. Reptiles and Batrachians from Surinam. Notes Leyden Museum. Vol. 25, 1904, p. 87. 160 Zoologica: New York Zoological Society EXPLANATION OF THE PLATES Plate I. Fig. 1. Gonatodes albogularis. Head and shoul- der. X 3.5. Fig. 2. Gonatodes albogularis. Natural size. Fig. 3. Gonatodes annularis. Head and fore body. X 5. Fig. Plate II. 4. Gonatodes annulay'is. Four eggs in a bamboo stub. Plate III. Fig. 5. Gonatodes beebei. Head. X 2.7. Fig. 6. Gonatodes booni. Natural size. Fig. 7. Gonatodes booni. Head and fore body. X 5.6. Plate IV. Fig. 8. Gonatodes caudiscutatus. Head and fore body. X 8.3. Fig. 9. Sphaerodactylus molei. Natural size. Fig. 10. Sphaerodactylus molei. Head enlarged, dorsal view. X 2.5. Fig. 11. Sphaerodactylus molei. Head enlarged, lateral view. X 6. Fig. 12. Fig. 13. Fig. 14. Fig. 15. Fig. 16. Fig. 17. Plate V. Thecadactylus rapicaudus. Day colora- tion. Thecadactylus rapicaudus. Night col- oration. Thecadactylus rapicaudus. Natural size ; renewed tail. Thecadactylus rapicaudus. Head and fore body. X 2.6. Plate VI. Thecadactylus rapicaudus. Sole of foot. X 7.6. Thecadactylus verted. X 2.2. rapicaudus. Toes re- BEEBE. PLATE I. FIG. 1. FIG. 2. FIG. 3. FIELD NOTES ON THE LIZARDS OF KARTABO, BRITISH GUIANA, AND CARIPITO, VENEZUELA. BEEBE. PLATE II. FIG. 4. FIELD NOTES ON THE LIZARDS OF KARTABO. BRITISH GUIANA. AND CARIPITO. VENEZUELA. BEEBE. PLATE III. FIELD NOTES ON THE LIZARDS OF KARTABO. BRITISH GUIANA. AND CARIPITO, VENEZUELA. BEEBE. PLATE IV. FIG. 11. FIELD NOTES ON THE LIZARDS OF KART ABO. BRITISH GUIANA. AND CARIPITO, VENEZUELA. BEEBE. PLATE V. FIG. 15. FIELD NOTES ON THE LIZARDS OF KARTABO BRITISH GUIANA, AND CARIPITO, VENEZUELA. BEEBE. PLATE VI. FIG. 16. FIG. 17. FIELD NOTES ON THE LIZARDS OF KARTABO, BRITISH GUIANA. AND CARIPITO, VENEZUELA. Crane: Color Changes of Fiddler Crabs in the Field 161 15. On the Color Changes of Fiddler Crabs (Genus Uca ) in the Field.1 Jocelyn Crane. Research Zoologist, Department of Tropical Research , New York Zoological Society. [This is a contribution resulting from a trip made in the spring of 1944 to Venezuela, Colombia and Ecuador, under the auspices of the New York Zoological Society and the Com- mittee for Inter-American Artistic and In- tellectual Relations. Invaluable assistance was received from Dr. William Beebe, Director of the Department of Tropical Research, the Creole Petroleum Corporation, Mr. and Mrs. William H. Phelps and Mrs. Sherman P. Haight.] Contents. I. Introduction 161 A. Brachyuran Chromatophores 161 B. Hormonal Activators of Chromatophores in Uca 162 II. Color Changes of Uca in the Field 162 A. Diurnal Darkening 162 B. Submergence Darkening 163 C. Darkening on Capture 163 D. Display Brightening 163 III. Development of Display Color 164 A. Sequence of Display Color Development . . 164 B. Display White as an Evolutionary Trend 164 C. Example of Color Development: U. princeps 165 D. Sexual Dimorphism 166 E. Geographic Color Variation . . .' 166 F. Ecologic Factors 167 IV. Summary and Conclusions 167 V. References 168 I. Introduction. During the past few years a number of investigators have done a great deal of excellent experimental work on the role of secretions from the crustacean eyestalk sinus gland and central nervous organs in chromatophore activation. Fiddler crabs have been frequently employed with great success as the subjects of the experiments. None of this research, however, by its very nature could be done on animals living in natural surroundings, and all was performed on three western Atlantic species, Uca pug- nax, U. minax and U. pugilator, none of which shows the high degrees of color de- velopment found in some tropical forms during display. Our own previous observa- tions (1941, 1943.1, 1943.2) indicated both the importance of the natural habitat in bringing about the crab’s most extreme color changes and the great diversity of color development within the genus. There- fore it seems worthwhile to present certain more recent observations, summarize the 1 Contribution No. 695, Department of Tropical Research, New York Zoological Society. field color studies, and correlate these as much as possible with the results of the laboratory investigators. By this means it is hoped that directions will be suggested for future research, preferably in ideal com- binations of experimental endocrinology with observation in the field. My thanks go to Dr. F. A. Brown, Jr., for his helpful comments via personal letter on some of his unpublished observations. Kleinholz (1942) and Brown (1944) have published the most recent and comprehen- sive surveys of our present knowledge of crustacean hormones, including the mech- anism of chromatophore control. For de- tailed discussion of the subject and exten- sive bibliographies, reference should be made to these papers. For the purposes of this discussion, the following brief resume may be made of the aspects immediately related to the present subject. A. Brachyuran Chromatophores: It is known that in brachyuran crabs most ex- ternal pigment is located in monochromatic chromatophores which are found chiefly in the epidermis. They have permanent branch- ing processes, so that the pigment may be either concentrated in the chromatophore’s center, in which case it is scarcely or not at all perceived in the macroscopic colora- tion of the crab, or it may be widely dis- persed into the processes. Four pigments occur in crab chromatophores, all of which are found in Uca: black or brownish-black (probably a melanin; hereafter called simply “black”), red, yellow and white (perhaps guanin). A fifth crustacean pigment, a blue, is the only one found outside of the chro- matophores; it has not been investigated in Uca, although blue is of frequent occurrence in the genus. All of the striking specific color differ- ences noted when fiddler crabs are studied in the field may be explained by the assump- tion that each species contains these five pigments in varying proportions, distribu- tion and states of dispersal. Research, how- ever, has scarcely begun on this question: Carlson (1936, p. 67), working with pugil- ator, found red chromatophores fewest of all 162 Zoologica: New York Zoological Society [XXIX: 15 kinds, but still numbering half as many as the abundant black; Abramowitz (1937) found, on the other hand, that in pugnax red is usually lacking. A connection between localized white chromatophores and breed- ing condition has been demonstrated in fe- male shrimps ( Leander ) (Knowles & Callen, 1940), and similar seasonal influences will almost certainly be found to apply to some pigments in Uca. Doubtless there will prove to be differences of specific rank in amount or potency of hormones affecting the dis- persal of pigments, of importance equal to or greater than the actual number of chrom- atophores or amount of pigment within them. At present, however, nothing what- ever is known of these aspects. B. Hormonal Activators of Chromato- phores in Uca: Two sources of chromato- phorotropic hormones are at present known to exist in fiddler crabs, the sinus gland of the eyestalk and the central nervous organs. The hormonal principles in the sinus gland, numbering at least two, are responsible for the dispersal of the black, red and perhaps the white chromatophoral pigment, and the concentration of the yellow and perhaps the white. One or more principles from the cen- tral nervous organs bring about the disper- sal of both the black and the white, as in the case of the sinus gland, according to an unpublished paper by Brown & Cunningham (Brown, 1944, p. 131, and personal com- munication). The known effects in TJca may be expressed in Table 1. The influence of hormones on the blue, probably extra- chromatophoric, pigment is omitted because of lack of data. II. Color Changes of Uca in the Field. In Table II is given a list, compiled from field observations on more than 20 species, of the apparent motivation of color changes as they occur in nature. All except those on diurnal rhythm, which have been well estab- lished by previous observers, are the result of our own field work carried out chiefly in the American tropics ; they apply to a lesser extent, however, to the three northern Atlan- tic species mentioned above. Here only dark- ening and lightening are indicated, regard- less of the chromatophoric principles in- volved. Table II. Color Change in Uca in the Field. Time, Place or Condition Crab Darkens Crab Lightens Day X — Night — X Submergence in burrow X — Capture and holding X — Display — X These color changes will be discussed in order. Since individual Uca show little or no adjustment to background color, that adaptation, so characteristic of many crus- taceans, is not considered in the present paper. A. Diurnal Darkening. It is well estab- lished that normal fiddlers are somewhat darker during the day than at night, and that this rhythm continues to a certain ex- tent regardless of whether the crabs have been blinded, their sinus glands removed or both (see especially Abramowitz & Abram- owitz, 1938, and Brown, 1944, p. 131). As Table I. Chromatophore Control in Uca. Origin of Hormone (s) Effect on Chromatophores Sinus Gland Central Nervous Organs . Key : D — Disperses C — Concentrates The function of these control systems in nature has scarcely begun to be explored. Brown considers it almost certain (1944, p. 131) that the diurnal rhythm of fiddler crabs (see below), which continues to a cer- tain extent regardless of light intensity or loss of eyes, is controlled by secretions from the central nervous organs. No function has yet been assigned, however, to the color changes controlled by sinus gland principles. In the following pages possible connections between endocrine functions and natural color change are suggested, and some aspects of the development of color in the group dis- cussed. Black Red Yellow White D D C D and/or C D ? ? D mentioned above, the diurnal rhythm is probably controlled by principles from the central nervous organs, which during the day-time disperse the black pigment. As there is no evidence that the white is dis- persed at night, the nocturnal lightening appears to be due to the simple concentra- tion of the black, particularly on the legs and ventral surface. It is much less intense than the dazzling white assumed by some species in display, which often extends over the entire body and, more rarely, the ap- pendages as well. As has long been known, Uca is altogether diurnal; at night it never feeds, digs or waves, and rarely even 1944] Crane: Color Changes of Fiddler Crabs in the Field 163 emerges from its burrow, although it may sit in the mouth during nocturnal low tides. A day-time darkening is of conceivable practical benefit in making the crab incon- spicuous in its usually dark muddy-sand or muddy habitat. B. Submergence Darkening. A fiddler crab in light display coloration (see below) is darkened slowly by prolonged submersion in its damp burrow. Hence the principle re- sponsible for this type of darkening — whether originating in the sinus gland, the central nervous organs or elsewhere — obvi- ously inhibits the effect of the display- brightening mechanism. This darkening appears more extensive than could be ex- plained merely by the saturation of the cuticle and its resulting translucence, which would permit body organs to show through. Also, it is quite different from the superfi- cial effect caused by clinging mud or sand. Since the display-brightening mechanism inhibits the diurnal rhythm presumably motivated by the central nervous organs, it is likely that the darkening principle of the sinus gland will be found to be respon- sible for the dispei’sal in burrow submer- sion. Any benefit to the crab is hard to imagine; it seems more likely that this ef- fect will prove to be a psychological by- product connected with some other process — perhaps to activity of the gland in some branch of metabolism concerned with respi- ration or water-intake. In fact, this darken- ing is perhaps a slight disadvantage, if, as seems almost cei’tain, display-brightness has recognition, advertising and/or stimulating value; a darkened crab must lose precious low tide display time in regaining bright- ness. No darkening effect is seen after brief descents, caused by momentary alarms, occasional digging activities, or the need for moistening the gills. It is extrem’ only after prolonged digging and after submerg- ence during high tide. C. Darkening on Capture. A similar dark- ening effect is seen when a crab in display coloration is seized and held in the hand, only in this case the darkening is usually much more rapid than when the crab merely descends into its burrow. The time required for the change is highly variable in species and individuals. So far, it has appeared most rapid in U. latimanus, one of the species in which display-blanching is most highly developed. The incidental effect of adrenalin on some vertebrate chromatophores during emergency may be kept in mind for possible parallelisms among fiddler crabs, especially in view of the similarity already suspected (Brown, 1944b, p. 13B) of crustacean cen- tral nervous organ extracts to adrenalin or acetylcholine. If a Uca nervous organ prin- ciple is responsible, however, for “emerg- ency” darkening of display coloration, it is probably different from the diurnal-darken- ing principle, since the latter is inhibited by the display-brightening substance. It may be remarked here that we have not so far succeeded in inducing either dis- play-coloration or waving in any fiddlers with pronounced display colors (e.g. U. sten- odactyla, stylifera, saltitanta, etc.) kept in captivity, although these experiments have not yet been carried out systematically. D. Display Brightening . The lightening, often to dazzling whiteness, of fiddlers in the breeding season and, during that season, of displaying individuals on bright days at low tide, has been repeatedly observed by us in both temperate and tropical species in the western hemisphere. For unknown reasons, the effect appears to reach its highest devel- opment in the tropical eastern Pacific. Its genera] characteristics have been previously discussed (Crane, 1941, pp. 154 ff). In re- gard to the chromatophoral hormonal as- pect, which has yet to be investigated ex- perimentally, the following remarks may be made: 1. One or more principles must be active which inhibit the normal effect of the cen- tral nervous organ principle which is appar- ently responsible for diurnal darkening as well, perhaps, as the dark-dispersing princi- ple of the sinus gland, the function of which is so far unknown. 2. The same or other principles must be responsible for the maximum dispersal of the yellow and white pigments, which are chiefly concerned in the brightening, as well as those bringing about the occasional sub- sequent concentration of the yellow (see p. 165), and special display coloration such as bright red chelae, purple ambulatories and bright blue eyestalks. 3. Since this maximum brightening oc- curs in the display season, which is inti- mately connected with breeding in at least a majority of species, it is likely that sex hor- mones, produced in the gonads or elsewhere, will be found to be responsible. 4. At least one immediate source may prove to be a dispersing-white principle of the sinus gland. The existence of this prin- ciple is recorded by Brown (1944, p. 40) as an unpublished observation, although Abramowitz (1937) has considered it as a concentrating-white agent. 5. Although no sex hormones have yet been proved to exist in Crustacea, the white chromatophores appearing in egg-bearing segments of female Leander should be kept in mind in attempting to explain the dis- play colors in Uca. Knowles & Callen (1940) found that these chromatophores failed to develop in Leander which had been parasit- ized or X-ray castrated, but think that little 164 Zoologica: New York Zoological Society [XXIX: 15 evidence was thus given for the existence of a female sex hormone. Their preferred, alternative conclusion is that these white chromatophores are a by-product of the in- creased metabolic rate of the breeding sea- son. Some such relatively simple explanation as the latter may prove to be true in Uca: The same authors remark on the part known to be played by guanin accumulations in the nuptial coloration of many vertebrates. An- other phenomenon, reported by McVay (1942) working with Cambarus, may prove to have a bearing on the problem: she re- ports a seasonal change in the concentration of the white chromatophore-contracting principle from the brain of females, and a striking difference between males and fe- males in the total amount present. It is likely that the seasonal behavior of white pigment in the three crustaceans — shrimp, crayfish and fiddler crab — will prove to be related. Should a basic execretory origin be proven, an adaptive use of the resultant white pig- ment as a recognition and stimulatory de- vice in Uca would not of course be necessar- ily eliminated. Some further aspects of display-brighten- ing are considered in the following section. III. Development of Display Color. A. Sequence of Display Color Develop- ment. In individuals as well as in the appar- ent trends of evolution within the genus, a certain sequence of chromatophoral display color development is discernible. This runs from black to red to yellow to white. When each of these colors appears dominant, ob- viously the pigment within the remaining chromatophores is concentrated. The full course of this black-white change is seldom run, and one or more phases are usually sup- pressed, but in various stages it can be traced in both groups of species and in in- dividuals within a species. Also, save for some exceptions in the development of color on major chelipeds, the sequence never changes: for example, a white carapace never precedes a yellow, nor yellow legs red. Combination phases of orange, pink and cream in various stages of development oc- cur frequently, however, and must be brought about by the simultaneous dispersal of red and yellow, red and white, and yellow and white pigment, respectively. The fifth pigment, blue, which has been shown in other crustaceans to be extra-chro- matophoric, is relatively uncommon and be- haves irregularly. It is almost always con- fined to local areas, such as eyestalks, mouth- parts or frontal regions. It may be said hers only that it is a true display color, develop- ing at the expense of diurnal black, and in- hibited both by burrow submergence and by capture. It seems always to develop be- fore the general dispersal of white, and usu- ally persists even when white has succeeded the remaining pigments. Often it has a strongly iridescent appearance, and may range in shade from purple and violet to turquoise and green, doubtless through simultaneous expansion of red or yellow chromatophores, respectively. The behavior of the chromatophoric pig- ments will now be considered from other angles B. Display White as an Evolutionary Trend. In a previous paper (1941, p. 156) it was pointed out that dazzling white is ex- ceedingly prevalent in the display coloration of the end-forms in each of the three groups of Uca in which display had at that time been observed. This white is assumed sea- sonally by the species, and daily by the indi- viduals. Typical examples of this extreme whitening are stylifera in Group 1, salti- tanta in Group 4 and terpsichores in Group 5. Since then, I have had opportunities of noting display coloration in a number of Group 2 fiddlers (examples: pugnax, mor- dax) as well as in other examples of the re- maining groups, and have found no reason to alter the conclusion that the development of display-white is a general evolutionary trend throughout the genus. It is most com- mon in the widely separated Groups 1 and 5, which contain the most specialized spe- cies, moderately so in Group 4, and rarest in Group 2; it is also well developed in 6, the offshoot proposed for the aberrant panamensis ( l.c ., p. 166). Occurrence in Group 3 has not yet been studied. Group 2 continues to appear to be the most primitive western hemisphere group, and, as just noted, white is less highly developed here than in any of the others. An exception is g alapag ensis , a close Pacific relative of pugnax ; when seen in display, in Ecuador, 1944, the few waving examples were pure white; all the rest, both males and females, ranged from grayish to bright yellow and cream. Depending on the sequence and extent of lightening of carapace in display coloration, species in which this phenomenon has been carefully observed may be divided into five divisions, as follows: (a) . Dark: No appreciable yellowing or whitening of carapace; display colors con- fined chiefly to appendages. Group 1 : mara- coani, insignis; Group 4: oerstedi, spini- carpa, inaequalis , batuenta, cumulanta. (b) . Carapace changes from dark to mud- dy yellow or grayish-white: Group 2: pug- nax, mordax, minax; Group 4: festae. (c) . Carapace changes from dark to cream: Group 5 (primitive offshoot), pugi- lator. 1944] Crane: Color Changes of Fiddler Crabs in the Field 165 (d) . Carapace changes from dark to pink to white: Group 5, stenodactyla (but bright blue maintained anteriorly). (e) . Carapace changes from dark to yel- low to cream to white: Group 1, stylifera, princeps; Group 2, galapagensis. (f) . Carapace changes from dark to cream to white : Group 4, saltitanta; Group 5, beebei, deichmanni, terpsichores, latima- nus; Group 6, panamensis. Several points must be emphasized here : First : the above represents maximum known color changes for each species. Sometimes, as in the case of princeps and beebei (see p. 166), they are not attained in every popu- lation. Sometimes, as in pugndx, dark dis- playing- crabs are the rule, and lighter phases the exception. Contrariwise, individ- ual crabs will be found displaying, or even actually mating, in colors not nearly as bright as those characterizing their immedi- ate neighbors. Finally, (f) differs from (e) only in having the white pigment start diffu- sion simultaneously with, instead of subse- quent to, the yellow. Dispersal of red and yellow pigment often persists simultaneously with black, as may be seen by the persistence of red joints in U. minax apparently at all seasons, and of the frequent occurrence of reddish-or-yel- lowish-brown carapaces, as well as of ochra- ceous tinges on the chelipeds, even when some species are in their dark phases. Pure yellow and cream phases are exceedingly transitory, however, and require special comment. In at least the three species listed under (e) above, the adults, both males and females, of whole populations coming into the display season go for days, perhaps weeks, with daily changes to the bright yel- low, or at most cream-yellow phase, but no continuation into white — that is, no sub- sequent concentration of yellow pigment accompanied by full dispersal of white. Later, individuals of princeps and stylifera have been seen to skip the yellow altogether in the daily change. Whether entire popula- tions at the height of the breeding season may skip the yellow is not yet known. How- ever, in the species listed under (f) above, which are preponderantly of the specialized Group 5, the creamy daily phase is some- times so short as to be practically non- existent, or it may be confined to some yellow speckles on the carapace, giving an over-all cream appearance. To account for these phenomena, four separate chromatophoric actions must be explained, even when the frequent involve- ment of red and blue pigment is omitted: (1), The normal diurnal expansion of the black pigment is inhibited; (2), the yellow pigment is dispersed; (3), the yellow pig- ment is concentrated; (4), the white pig- ment is dispersed. Whether a separate hor- mone is responsible for each mechanism is of course completely unknown. Equally un- known is whether these chromatophoral changes are merely by-products of physio- logical processes concerned in reproduction, w'hether they are a true adaptive result of display evolution, with recognition and/or stimulatory functions, or whether tempo- rary intermediate phases, such as the yellow, are recapitulations of ancestral conditions. All three possibilities will probably be found to be true in part, and to vary with the species. Brown and Wulff’s (1941, p. 344) observations on the behavior of yellow pig- ment in Crago should be kept in mind: the yellow in eyestalkless animals was first maximally concentrated and then rapidly dispersed by sinus gland extracts. C. Example of Color Development : U. princeps. The large cheliped and often the ambulatories may run a course of color development independent of that of the cara- pace. For example, a majority of even the dullest fiddlers have the tips of the chelae completely white, and a great deal of red or orange at least on the major manus, while purple shades are common on the ambula- tories. This red or orange sometimes per- sists even in species which otherwise change the carapace to complete white (example: stylifera). Again, the dispersed pigment of the cheliped may not only persist but may be white in the young, and reverse the usual sequence by developing orange and red secondarily, as shown below. An example of ontological color develop- ment will be given in detail, as it occurs in U. princeps, which passes through a dis- tinct yellow phase. The observations were all made at Puerto Bolivar, Ecuador, late in April. Apparently the display season was just beginning. Stage I. up to ca. 10 mm. Carapace above and below greenish-brown; manus and che- lae brilliant white, inside and out; ambula- tories rich plumbaceous anteriorly; rest of cheliped and other legs muddy brown; eye-, stalks green or greenish-yellow. One excep- tion, among hundreds of similar size, showed precocious coloration, being entirely white. Stage 11. ca. 11 to 18 mm. Like above, ex- cept for yellow developing on lower part of manus. This changes to dull orange, then blazing scarlet orange and spreads all over manus, inside of carpus and merus and throughout pollex except tip. Stage III. ca. 19 mm. to maximum {ca. 30 mm.). Cheliped remains bright; carapace and legs lighten to dull orange. Stage IV. Like above, except dull orange carapace and legs brighten to yellow, then cream. A few males in this stage fought and 166 Zoologica: New York Zoological Society [XXIX: 15 displayed. This was the most abundant stage among adults, which obviously were not in full display season. Stage V. Crab completely dazzling white except for bright orange pollex, or pollex and lower manus. All the males in this stage were displaying. D. Sexual Dimorphism : As in many other groups of animals, there are species of Uca (e.g. saltitanta, stenodactyla, lati- manus ) where the female is always very dull and the male exceptionally brilliant in dis- play coloration, while in closely related spe- cies the females are as bright as the males (except, of course, for absence of the often gaudily colored major cheliped). The latter group includes some species where white development reaches its height. The most striking examples of this found to date are the females of princeps and terpsichores, all observed in Ecuador. In others, such as gala- pagensis and pugilator, the female attains a cream only slightly deeper than the white or cream of the male. It may be noted that in species such as stenodactyla, where the fe- male is very dull and the male especially bright, the dimorphism in size is also ex- treme; in those where the color difference is slight, the size difference also is relatively small. Here is another possible example of linked hormonal effects. E. Geographic Color Variation: A most interesting observation made in Puerto Bolivar, Ecuador, gives evidence of marked geographic variation in the display colors of two species. Study of displaying princeps (Group 1) and beebei (Group 5) in Panama (February and March, 1941) during the dry season (1941, pp. 170, 193) showed that these species, although waving strongly, and with ovigerous specimens present, showed practically no white except on the fingers: instead, the display coloration con- sisted of gray, rose, orange and purple in princeps (which were in Stage II., p. 165), and gray, green, purple and ochre in beebei. Close relatives of each, on the other hand, stylifera and terpsichores, respectively, when in full display coloration had every- thing except the appendages dazzling white. In observations made in April, 1944, at Puerto Bolivar, Ecuador, displaying individ- uals of both princeps and beebei were found to be almost completely dazzling white in color, although individuals of the darker and intermediate shades were present, as usual. In princeps there was also, as de- scribed above, an intermediate yellow phase. In beebei, a very few individuals in the darker phase characteristic of Panama were displaying; in princeps, none. It was near the beginning of the dry season in Ecuador, so that the comparison with the Panamanian observations was ideal. In morphological characters the indi- viduals proved typical of their species in every detail, and their displays were in- distinguishable. More evidence that the vari- ation is actually geographic, and not a result of the observations of purely season differ- ences, lies in the fact that immediately after the Ecuadorian study, the same Panamanian mud-flats at La Boca, Canal Zone, and Bella- vista, Panama City, were visited during the second week in May, after the start of the rains. Both species were displaying, and both were in the identical dark coloration found three years previously during the dry, without a single white individual pres- ent, even on the clearest, sunniest days at full low tide. A point of special interest concerns the size difference between displaying males of princeps in the two regions. In 1941 (p. 70) it was remarked that in the Panamanian colonies the largest males seen, all of which were displaying strongly, measured around 15 mm. in length, although the maximum measurements of specimens from other localities were around 25 mm. In addition, it has been determined that in these small males the abdominal appendage is slightly pre-adult in form. The same conditions pre- vailed there in 1944 in the colony at Puerto Bolivar; however, the only males displaying were large individuals measuring around 25 to 30 mm., with abdominal appendages of mature development, and all of these were in the white, or at least the creamy-yellow phase. As far as could be determined, gen- eral ecologic factors were exceedingly simi- lar; in both the Panamanian and Puerto Bolivar populations, the salinity (high tide, near full moon : unfortunately only one sam- ple was taken from each) was about 38 parts per thousand, or about 75 per cent, of aver- age open ocean concentrations; also, the general facies of the mud-flats, including high Pacific tides, and climatic conditions with pronounced wet and dry seasons, were much alike. There are at least four possible explana- tions, which are not necessarily mutually exclusive. (1) . In the Panamanian habitat some unknown environmental factor may promote early maturation of the sex glands before maximum size and pigment development are attained. (2) . Contrariwise, a factor may inhibit development of the last two characteristics. (3) . In Puerto Bolivar, some factor may encourage the production of white and yel- low pigment, or of the hormone (s) govern- ing their dispersal, or both. Since the white is confined to displaying forms, however, connection with gonad activity still seems essential. Also, this alternative would not 1944] Crane: Color Changes of Fiddler Crabs in the Field 167 be adequate to explain the small size and pre-adult form of abdominal appendages in Panamanian princeps. It would however be sufficient to cover the development of the white in Ecuadorian beebei, which show no size or appendage difference from Panama- nian specimens. (4). A fourth explanation, which would apply equally well to both species, is that in Puerto Bolivar species development, through mutation and selection, has progressed be- yond the stage found at Panama, and has reached the normal summit of Uca color evolution, which tends toward white display dress in all groups. Because of the lack of morphological differences and the presence of Panamanian color phases in both princeps and beebei at Puerto Bolivar among the white individuals, there seems to be no adequate basis at pres- ent for the erection of subspecies or other subdivisions. The same two species must obviously be studied in other localities and at all seasons before the “normal” forms can be identified. F. Ecologic Factors: Too little is known of precise ecological factors to draw any conclusions at present. It may only be stated here that so far as known crabs with the most brilliant coloring, including develop- ment of display white, live only on tropical shores with more than half the salt concen- tration of open ocean water and a high tidal range. In the western hemisphere, these conditions are met in the tropical eastern Pacific, where the most active spe- cies, and the greatest numbers of species also occur. Second, bright sun is needed to bring about maximum dispersal of white pigment, just as it is necessary for maxi- mum display activity: although there are always individual exceptions, a given popu- lation is invariably notably darker and far less active on cloudy days than on clear. IV. Summary and Conclusions. Four types of normal color changes in Uca have been described, namely diurnal darkening, submergence darkening, dark- ening on capture and display brightening. None except the first has been studied microscopically or investigated experi- mentally from an endocrinal viewpoint. It is already fairly certain that diurnal darkening is brought about by a secretion from the central nervous organs; it is sug- gested in the present paper that the slow- working submergence darkening may be motivated by the dark-dispersing element of the sinus gland, while the relatively rapid darkening on capture may be controlled by the latter, or by another element in the nervous organs. The display brightening principle (s) inhibit the diurnal darkening mechanism as well as the normal sinus gland darkening, and are in turn inhibited by those of the submergence and capture darkening. Display brightening includes, in addition to widely occurring bright chelipeds and ambulatories in males, a general trend throughout the genus toward the develop- ment of complete dazzling whiteness in both sexes during the display season. Fully white crabs occur in widely separated species. In ontological seasonal and daily changes, a strict sequence of display color development is apparent, which runs from black to red to yellow to white, each phase representing dis- persal of one of the four chromatophoric pigments. Any of the phases except the first may be suppressed, and combinations are frequent, giving orange, pink and cream ef- fects. Yellow and cream phases are the most transitory. In most but not all species where bright display color is well developed, the carapaces of the females are decidedly darker than those of the males. Two cases of geographic color variation are described, where displaying males of two unrelated species are dazzling white in Ecuador and much darker in Panama, under apparently similar ecological conditions. In one species, the dark Panamanian examples, although giving all signs of actual breeding, were in addition small with subadult abdominal appendages. Ecologic factors for development of dis- play color have not yet been specifically studied. Bright sun, however, is essential for maximum daily development of display color. So far, the brightest and most active of western hemisphere species, as well as the greatest local concentration of species, have been observed in the tropical eastern Pacific on shores with high tidal ranges and with a salinity concentration of more than half that of open ocean water. The above data has been put on record with the chief hope of suggesting lines for future endocrine and psychological research in this group, preferably with a combina- tion of field and laboratory methods. The adaptive functions of display color are still only suspected: it is not even known whether fiddler crabs have color vision, much less whether the brightening of the male crab in display has recognition and/or stimulat- ing value for the female, and serves for territorial limitation and challenge to other males, whether it is merely a metabolic by- product, or whether recapitulation of an- cestral color is involved in cases of transi- tory phases. Judging from the behavior of the females and rival males, however, it seems certain that at least the brightness (irrespective of color) of the waving chelae, 168 Zoologica: New York Zoological Society which adds so much to the conspicuousness of the display in even the dullest species, must be of actual value. References. Abramowitz, A. A. 1937. The comparative physiology of pig- mentary responses in the Crustacea. Jour. Exp. Zool., 76: 407-422. Abramowitz, A. A., & R. K. Abramowitz. 1938. On the specificity and related proper- ties of the crustacean chromotophoro- tropic hormone. 'Biol. Bull., 74, 278- 296. Brown, F. A. Jr. 1944. Hormones in the Crustacea: Their Sources and Activities. Quart. Rev. Biol., Vol. 19, No. 1 (pp. 32-46), and No. 2 (pp. 118-143). Brown, F. A., Jr., & V. J. Wulff. 1941. Chromatophore types in Crago and the endocrine control. J. Cell, and Comp. Physiol., Vol. 18, No. 3, pp. 339-353. Carlson, S. P. 1936 (1937). Color changes in Brachyura crustaceans, especially in Uca pugila- tor. Kungl. fysiogr. Sallsk. i Lund For- handl. 6:63-80. Crane, J. 1941. Eastern Pacific Expeditions of the New York Zoological Society. XXVI. Crabs of the Genus Uca from the West Coast of Central America. Zoologica, Vol. XXVI, pp. 145-207. 1943.1. Crabs of the Genus Uca from Vene- zuela. Zoologica, Vol. XXVIII, pp. 33-34. 1943.2. Display, Breeding and Relationships of Fiddler Crabs (Brachyura, Genus Uca ) in the Northeastern United States. Zoologica. Vol. XXVIII, pp. 217-223. Kleinholz, L. H. 1942. Hormones in Crustacea. Biol. Rev., 17 : 91-119. Knowles, F. G. W., & H. G. Callan. 1940. A change in the chromatophore pattern of Crustacea at sexual maturity. Jour. Exp. Biol., 17, 262-266. McVay, J. A. 1942. Physiological experiments upon neuro- secretion, with special reference to Lumbricus and Cambarus. Doctorate Thesis. Northwestern University, Ev- anston, 111. NEW YORK ZOOLOGICAL SOCIETY General Office: 630 Fifth Avenue, New York 20, N. Y. Publication Office: The Zoological Park, New York 60, N. Y. Fairfield Osborn, President Alfred Ely, First Vice-president Laurance S. Rockefeller, Second Vice-president Harold J. O’Connell, Secretary Cornelius R. Agnew, Treasurer John Tee-Van, Executive Secretary Jean Delacour, Technical Adviser William Bridges, Editor and Curator of Publications Zoological Park Lee S. Crandall, General Curator & Curator of Birds Leonard J. Goss, Veterinarian Claude W. Leister, Associate, Mammals John Tee-Van, Associate, Reptiles Grace Davall, Assistant to General Curator Aquarium Christopher W. Coates, Curator and Aquarist Ross F. Nigrelli, Pathologist Myron Gordon, Assistant Curator C. M. Breder, Jr., Research Associate in Ichthyology G. M. Smith, Research Associate in Pathology Homer W. Smith, Research Associate in Physiology Department of Tropical Research William Beebe, Director Jocelyn Crane, Research Zoologist Henry Fleming, Entomologist George Swanson, Staff Artist OFFICERS SCIENTIFIC STAFF General Claude W. Leister, Education William K. Gregory, Associate Gloria Hollister, Associate John Tee-Van, Associate Mary VanderPyl, Associate Editorial Committee Fairfield Osborn, Chairman William Beebe William Bridges Christopher W. Coates Lee S. Crandall Jean Delacour Claude W. Leister John Tee-Van ZOOLOGICA SCIENTIFIC CONTRIBUTIONS of the NEW YORK ZOOLOGICAL SOCIETY VOLUME XXIX Part 4 Numbers 16-19 Published by the Society The Zoological Park, New York December 30, 1944 CONTENTS Page 16. Anatomy of the South American Woolly Monkey ( Lagothrix ) . Part 1. The Forelimb. By Donald Ford Robertson, M.D. (Plates I-V) 169 17. Some Venezuelan Membracidae. By W. D. Funkhouser 193 18. Field Notes on the Lizards of Kartabo, British Guiana, and Cari- pito, Venezuela. Part 2. Iguanidae. By William Beebe. (Plates I-VI ; Text-figures 1-17) 195 19. Materials for the Study of the Life History of Tarpon atlanticus. By C. M. Breder, Jr. (Text-figures 1-9) 217 Index to Volume XXIX 253 Robertson: Anatomy of the Woolly Monkey 169 16. Anatomy of the South American Woolly Monkey ( Lagothrix ). Part 1. The Forelimb. Donald Ford Robertson, M. D. (Plates I-V). Introduction. In presenting this description of the gross anatomy of the forelimb of Lagothrix, the South American woolly monkey, it should be explained that subsequent papers are planned to cover the entire body. These will consist of appropriate regional divisions, de- signed, so far as possible, to permit ready integration. Rather than employing the standard, sys- tematic method, I have chosen the regional or “dissector’s manual” type of description. In my opinion this is the only logical way to present the material in a form that is readily usable for practical purposes. Func- tional interpretations will be omitted ex- cept in those parts where the anatomical details revealed in the dissection seem to lend themselves particularly to the empha- sis of certain points, or where the functional aspect appears to require an especial anno- tation underlining the conclusions I have reached. The anatomical material on which this description is based consists of one adult female, one young female, and one male infant. The first two are L. lagotricha, the third L. cana. I am indebted to Dr. L. J. Goss, veterinarian of the New York Zoologi- cal Park, for his great courtesy and gen- erosity in placing this material at my dis- posal. Without his help and encouragement much of this work would not have been pos- sible. A search of the literature has revealed oniy the most fragmentary data on the genus Lagothrix. I am indebted to Dr. R. J. Terry, Pro- fessor of Anatomy, Washington University, School of Medicine, and Dr. M. J. Guthrie, Professor of Zoology, University of Mis- souri, for their kindness in reading and correcting the manuscript. External Characteristics. The upper extremity in Lagothrix, as in all primates, is a remarkably differentiated member capable of a great variety of move- ments and serving for locomotion, and in particular, prehension which is character- istic of primates. By grasping objects and drawing them near for close visual inspec- tion, and for stereognostic appreciation, the primate is peculiarly favored in a rnjnute and accurate orientation to its environment. The upper extremity of Lagothrix acts simply as a forelimb when the animal is walking on the ground. The hand, with its char 'Cteristic position of flexion into a balled fist, carries the weight on the well- developed hypothenar eminence which pre- sents a proximal extension. The flexed fing- ers may bear some weight as would the toes for very brief periods in man, but the fing- ers are never extended while the animal is walking on a flat surface, and any weight- bearing by the fingers is on the dorsal surface of the distal phalanges of the flexed digits. This is of interest as it may repre- sent a stage in the development of the digi- tal posture of the anthropoid hand. Lagothrix is not usually thought to be a brachiator, but some authors have men- tioned observations of the occasional use of this method of progression. I have fre- quently observed genuine brachiation, al- though it must be admitted that it is modi- fied by the constant use of the prehensile tail for temporary support as well as maintenance of balance. This is in contrast to Atsles which frequently brachiates with- out employing its tail for support, and of course to che anthropoid apes which lack a tail. The forelimb is very important in climbing and in supporting its share of the weight in the four-legged type of progres- sion erect or inverted. The hand presents its most specialized prehensile activities, however, when the position of the animal, such as sitting on its haunches, standing erect on the hinder extremities, or hanging supported by them and the tail, frees the arms for this pur- pose. The grasp, as discussed elsewhere, is of two general kinds— one a simple hap- hazard and hasty grasp, the other more elaborate and directed toward exact ap- proach to the object. 170 Zoologica: New York Zoological Society [XXIX: 16 Integument. The skin of the forelimb is thickly cov- ered with the thick rabbit fur-like hair characteristic of this genus for which the name was given. As a general rule, all species or types show a tendency to a predominance of darker hair from the el- bows distally. The hair is dense on the extensor surfaces. It is thinner and sparse on the flexor surfaces, presenting bare spaces in the axillae about the nipple, in the supraclavicular region and in the ante- cubital fossa. Elsewhere it is thick and dense. Adults of all types present along the anterior border of the limb a thick fringe of long hair, often of great length. On viewing the animal anteriorly, this is seen to be the upper arm of a chiasmal growth that runs down the mid-ventral region and extends along the upper and lower extremities at the cephalic borders. The skin is everywhere freely movable. It is thin and less elastic on the flexor sur- faces and thicker and more turgid on the extensor aspect. The integument of the palm displays the characteristic ridges and grooves, as- sembled in configurational systems, com- prising the patterns of the volar pads and the intermediate and surrounding areas in which there is no definitive arrangement. The variability and diversity of pattern configuration and distribution is such that, like man, no two individuals present iden- tical arrangements. In some twelve pairs of hands I have thus far examined, no repetitions have been found. The generalized plan of volar pads, as exemplified by the chiridia of the lower mammals, has been materially modified, although remnants of each of the general classifications described by Whipple (4) may be found, varying from one individual to another. The apical pads on the tips of the fingers are comparatively constant and show a monotonous similarity of concen- tric rings, varying to some degree in the minutiae. Pads on the middle and proxi- mal phalanx show a somewhat greater degree of diversity and variation although they too present little of the specialized diversifications of the palmar pads proper. If one can accept the thesis that the elab- oration of the patterns is correlated in some degree with the role of these areas in tactile sensibility and hence with per- ceptual accuracy, it is suggested that the palmar surface of the proximal phalanges and the opposing surface of the palm cons- titute the chief areas of perception in this monkey. The interdigital pads in the hands examined are occasionally absent, or much reduced (one specimen), but in all the others there were at least two and in some cases three. These consist of numbers I ard II and occasionally III (Whipple’s generalized schema). In both hands of one specimen there are intermediate pads with patterns of concentric rings. These are sufficiently well organized and discrete to be considered as probably independent de- velopments and not due to segregation by the superimposed creases, of areas former- ly included with the interdigital pads. The thenar and hypothenar areas present well- developed pads, although in none of the specimens I have examined do these pads present significant arrangement of ridges and sulci into concentric rings or whorl- like patterns. They consist, then, simply of open fields, but contrary to the state- ment of Bychowska, (2) I do not find, except in one specimen, that the lines are directed proximo-distally ; on the contrary the direction in all cases is from side to side. This might be expected in view of the importance of these areas, especially the large hypothenar pad, in weight bear- ing when the animal walks on all fours. The transverse direction naturally increas- es the frictional resistance, also thought to be an important function of these ridges by some observers. I do not find the areas on the palmar surface proper in which epidermic warts rather than ridges are found by Whipple. In all my specimens the intervening areas between patterns are covered by skin bearing well-developed ridges and at the margins of the palmar area, the ridges simply fade out into non-sculptured skin with no transition in the form of wart-like structures. The hand of Lagothrix has been well described by Pocock (5) as zygodactylous. That is, the space between digits two and three is much wider than that between one and two, and the grasp, especially of small elongated objects, is between digits two and three. The thumb should not be dis- regarded in this connection, however, since, while it moves with the index and the two digits constitute what I call the “thumb unit,” it is capable of a consider- able degree of independent motility, es- pecially when the index is fixed. There appear to be two possible explanations for this development of a thumb unit con- sisting of digits one and two; first, that this zygodactylous separation is directed by the same forces that have determined the development or persistence of the cloven hoof of digitigrade quadrupeds; or, second, that this functional division simply represents an individual specialization emphasizing the importance of the index in the absence of independent functional potentialities of the pollex which is in- capable of opposition, and is here simply a satellite of the index. It must be con- sidered also that a combination of factors is operative here since in these monkeys 1944] Robertson: Anatomy of the Woolly Monkey 171 the hand serves a dual function, that of support in the suspended position or in walking on all fours, and also serves to grasp movable objects of various sizes. In the one case strength and stability of structure are essential, while in the other delicacy of motion with graded ap- proaches and refinements of manipulative power are required. At this stage I am not prepared to discuss the importance of the respective factors. Subcutaneous Tissues and Cutaneous Nerves — Fasciae. On removing the skin the loose areolar tissue forming the superficial fascia is revealed. There is rarely an appreciable amount of fatty tissue, hence care must be exercised in removing the skin to pre- serve the superficial sensory nerves. Ex- cept for the area of their terminal distri- bution these run beneath the deep fascia. The deep fascia is not so well defined as it is over the forearm structures. This is especially true around the shoulder region and the upper part of the arm. About the elbow and from there distally, however, the deep fascia forms a well-defined sheet of considerable strength. The Nn. supraclav tculares will be de- scribed in detail with the cervical plexus. Two of these emerge in the region of the acromion to supply the skin of the supra- clavicular fossa and the superior aspect of the shoulder. The anterior aspect of the shoulder over the deltoid muscle is sup- plied by a branch of the N. dorsalis, scapu- lae that emerges between the deltoid and pectoralis major in the groove containing the cephalic vein. The medial aspect of the arm is supplied by the medial brachial cutaneous nerve whose rami perforate the deep fascia serially. The branches of the medial antibrachial cutaneous nerves spread out in fan-shaped manner over a triangular area on the medial aspect of the forearm, the most distal continuing almost to the wrist. The intermediate an- tibrachial cutaneous nerves may be more conveniently described later. The lateral brachial branches of the N. radialis emerge in the lower half of the arm in the region of the lateral intermuscular septum. These branches supply the lateral aspect of the lower third of the arm and the el- bow region. The lateral antibrachial cu- taneous nerve consists of three branches supplying an elongated triangular area of the lateral and dorsal aspect of the fore- arm. The posterior aspect of the shoulder and arm are supplied by cutaneous branch- es of the N. axillaris which emerge from the lower border of the spinous division of the M. deltoideus. The superficial veins of the hand and forearm form two main trunks, the cephal- ic and basilic, which are formed on the lateral and medial aspects of the wrist by the union of variable veins draining the fingers and dorsum of the hand. There is no dorsal arch and these venous channels from the hand arise from the first and second and from the fifth digits only. The cephalic vein passes upward along the dorsolateral aspect of the forearm and the arm passing medially over the superior border of the M. pectoralis major and be- hind its clavicular origin to join the Vv. brachiales. The basilic vein passes upward over the medial and volar aspect of the arm to the antecubital fossa where it pierces the fascia to join the Vv. brachia- les at this point. Anterior Thoracic Region. The fascia of the anterior thoracic re- gion is irregular in thickness, being less restrictive over the main area of the pec- toral region. It becomes better defined near the lateral margin of the M. pectoralis major at its lower border. This fascia forms a definite diaphragm between the M. latissimus dorsi, the M. pectoralis major and the fascia of the arm in the axilla. On each side of the midline the fascia is pierced serially by the anterior cutaneous branches of the intercostal nerves, and in the mid-axillary line by their lateral divisions. The fascia is attached firmly to the clavical superiorly and along a median raphe the entire length of the sternum. The M. pectoralis major arises, as in other monkeys, from a median raphe with interlacing fibers crossing the midline from the entire length of the manubrium and corpus sterni, the sternoclavicular joint and the medial third of the clavicle. The upper margin of the lateral third of the pectoralis major lies beneath the lower margin of the deltoid. On separating the deltoid from the pectoralis major the ori- gin of the latter from the inner third of the clavicle and the course of these superi- or fibers which insert at the lowest point on the humerus are better revealed. When the pectoralis major is severed from its sternal origin and reflected, the inferior portion is found to arise in two sheets, the superficial from the sternum and median abdominal raphe, the deep from the costal cartilage of the sixth rib. These lower fibers take a progressively deeper position as they proceed to the insertion at the highest point on the lateral lip of the bicipital groove of the humerus. When the pectoralis major is turned upward the anterior thoracic nerves and an artery are seen entering its deep surface. This neuro- vascular bundle passes between the pec- toralis minor and the pectoralis abdominis. A small accessory pectoral muscle which 172 Zoologica : New York Zoological Society [XXIX: 16 may be called the M. pectoralis accessorius in the absence of other identification, has been found in these specimens. It arises variably from the deep fascia of the ab- domen or from the sixth or seventh rib lateral to the lower border of the pec- toralis major, and passes behind the major in the midclavicular line to insert with it by a fine aponeurotic tendon on the under surface of the pectoralis tendon. Its func- tional significance is obscure. A fine branch of the anterior thoracic nerve passes downward to it from the rami supply- ing the pectoralis major. 1 The M. pectoralis minor arises super- ficially from the costal cartilages of the third, fourth and fifth ribs, the individual parts being separated by perforating branches of the intercostal arteries and nerves. On reflecting this superficial sheet a deep origin is revealed. This is formed by separate slips of aponeurotic structure from the third, fourth and fifth costal cartilages at a point lateral to the origins of the superficial sheet. There is a third and still deeper origin from the costal cartilage of the second rib. These three sheets are easily separable up to their point of fusion, just short of the inser- tion by an aponeurotic tendon which the M. pectoralis minor shares with the pec- toralis abdominis, covering and attaching to the lesser tuberosity of the humerus. The pectoralis minor is supplied by its own artery and nerve from the anterior thoracic and thoracoacromial group. The M. pectoralis abdominis arises by both a deep and a superficial head. The former lies directly under the latter as they arise from the seventh rib and from the sheath of the rectus abdominis. At mid distance from their origin these two heads diverge, the superficial being in- serted more distally than the deep, which occupies an intermediate position between the former and the pectoralis minor, on a thin common tendon of insertion. This covers the lesser tuberosity of the humerus, lying superficial to the insertion of the sub- scapularis. When the superior and deep head of the pectoralis minor are reflected, the origin of the subciavius muscle on the costal cartilage of the first rib is revealed. Lateral to the origin of the subciavius is found the upper head of a segmented muscle arising from the first, second, third and fourth ribs in the midclavicular line, each slip giving rise to a thin, flat apo- neurosis superimposed one on the other to fuse with the origins of the pectoralis minor on the costal cartilage of the fifth x j/^s muscle may be the sole persistent representation of the M. panniculus carnosus. Its insertion with the pectoralis major speaks for this explanation (see Hartman & Straus, p. 103), although origin suggests that it is a part of the pectoralis complex. rib. Beneath its termination the upper- most origin of the rectus abdominis from the fourth costal cartilage can be seen. This muscle is only revealed in its full extent in the dissection of the thoracic wall and will receive further consideration in that section. Separation of the anterior margin of the deltoid from the pectoralis major re- veals that, except for about a centimeter of its length, the insertion of the pectoralis major is covered by the deltoid. The two muscles diverge near the clavicle and a bare area of this bone in its middle third is revealed. The M. deltoideus has a lengthy linear origin, and according to the various points may be divided into clavicular, acromial and scapular (spinous) parts. The clavic- ular origin is from the lateral third of this bone, separated by a distinct interval from the clavicular origin of the pectoralis major. This hiatus of the muscles leaves the bare area described above. The clavic- ular fibers sweep downward over the in- sertion of the pectoralis major, with the deeper fibers forming an aponeurotic ten- don which fuses in part with the tendon of the pectoralis major, The acromial origin occupies the full curve of this prominence, and these fibers proceed straight downward to insert on the lowest point of the deltoid crest of the humerus. The scapular origin consists of a thin triangular sheet with a concave lower margin arising in part from the lateral two-thirds of the inferior margin of the spine of the scapula and in part by a thin membranous aponeurosis from the lower third of the vertebral margin of the scap- ula. These lower fibers sweep laterally and converge downward to join the acro- mial division, inserting posteriorly on the humerus. The deitoid is supplied by the N. axillaiis which is spread over its deep surface in fan shaped manner. The nerve is accompanied by the terminal muscular branches of the posterior circumflex humeral artery. if the ae.cOici is severed near its origin and reflected downward, the insertion of the pectoralis major is found to fuse in part with the insertion of the anterior fibers of the deltoid. The superior portion of this conjoint tendon forms a tunnel through which the tendon of the long head of the biceps passes from its intra- articular origin to insert on the supra- glenoid tubercle. 'the axiiiary fossa is pyramidal in shape with the base presenting laterally. The pectoral muscles together with the com- mon origin of the coracobrachialis and the short head of the biceps form the an- terior covering of the fossa. Posteriorly it is bounded by the M. subscapularis, 1944] Robertson: Anatomy of the Woolly Monkey 173 j medially by the chest wall, and laterally by the axillary fascia. On removing the axillary fascia and the pectoral muscles the axillary neurovascular bundle is re- vealed. The first third of the axillary artery is bounded below by the axillary vein and posteriorly by the combined cords of the brachial plexus. The middle third of the artery lies behind the cords of the plexus and continues in this position in its distal third. The axillary vein lies first below the artery and nerves and then passes behind the nerves to occupy a posi- tion medial to the axillary artery just before it becomes the brachial artery. The first branch of the axillary artery in the axilla is the A. thoracoacromialis. This dvides into pectoral branches to the Mm. pectoralis major, minor and abdominis, and gives deltoid and acromial branches to the clavicular and acromial origins of the deltoid. Proximal to the origin of thora- coacromial artery, the medial cord of the brachial plexus gives off muscular branch- es to the pectoral group and to the sub- clavius. From the thoracoacromial trunk three arteries constitute its terminal branches; these are the A. thoracalis suprema, the A. thoracalis lateralis to the serratus anterior, and the A. thoracodor- salis supplying the latissimus dorsi. The latter is accompanied by the N. thoraco- dorsalis. The long thoracic nerve passes to the serratus anterior somewhat pos- terior to the region of the muscle to which the A. thoracalis lateralis is supplied. Note : To display the brachial plexus in its entire extent it is necessary to disar- ticulate the clavicle from its sternal at- tachment and reflect certain structures having attachments in this region. Since this necessitates disturbing these parts, they will be described here. The M. sternocleidomastoideus separates into two distinct bellies in the lower part of the neck. The lateral one inserts on the upper and posterior surface of the medial third of the clavicle. The medial head passes anterior to the sternoclavic- ular joint and inserts on the superior and anterior aspect of the manubrium sterni posterior to the upper fibers of origin of the pectoralis major. When the lateral head of the sternocleidomastoid is reflected from its clavicular insertion the sternoclavicular joint is revealed. The articulatio sternoclavicularis is a typical diarthrosis with an articular disc of fibro-cartilage. The capsule is loose and permits comparatively free movement. The ligaments consist of the anterior and pos- terior sternoclavicular ligaments passing in an oblique manner laterally from the sternum. The posterior is several times stouter than the anterior ligament. There is also a costoclavicular ligament between the clavicle just distal to the joint and the superior margin of the first costal car- tilage. By severing the ligaments and releasing the joint, and by severing the M. suh- clavius just beyond its stout origin on the anterior surface of the first costal car- tilage and superior surface of the first rib, the clavicle may be displaced and the entire brachial plexus is revealed. Passing over the superior belly of the M. sterno- costalis is the axillary vein and above it is the axillary artery, lying here on the insertion of the M. scalenus anterior. Here the axillary also comes into relation with the upper surface of the arch of the first rib. Prominent in this region is the V. jugularis externa passing downward to join the V. jugularis interna. The N. phrenicus crosses over the subclavian ar- tery to enter the mediastinum just lateral to the A. cervicalis ascendens. This latter divides into the A. transversas scapulae and continues on as the ascending cer- vical. The transverse scapular artery gives a branch to the lateral head of the M. sternocleidomastoideus, and then passes posteriorly to the superior border of the scapula where it divides into superficial and deep branches. Its further course will be described with that region. The M. subclavius is a very stout muscle in Lagothrix. It passes upward and lateral- ly from its origin on the costal cartilage of the first rib, behind the clavicle, to insert on the postero-inferior surface of that bone near the attachment of the acromio-clavicular ligament. A second part of the muscle diverges and passing pos- teriorly inserts on the superior border of the scapula near the base of the coracoid process. The Brachial Plexus. The fifth, sixth, seventh, and eighth cervi- cal and the first thoracic nerves fonn the bra- chial plexus. These divide into anterior and posterior divisions, which in turn form the lateral, medial and posterior cords by re- union. Union of the anterior divisions of the fifth and sixth nerves forms the lateral cord, while the medial cord is formed by the union of the anterior divisions of the seventh and eighth nerves, and the pos- terior cord by the union of the posterior divisions of the fifth, sixth, seventh and eighth nerves. The fourth cervical nerve contributes to the brachial plexus solely by the branch which joins the rami of the fifth and sixth nerves to form the N. phrenicus. The first branch of the lateral cord is the N. suprascapularis which consists of the major part of the fibers of the fifth cei’vical nerve. This passes laterally and posteriorly over the superior margin of 174 Zoologica : New York Zoological Society the scapula medial to the insertion of the M. subclavius and beneath the M. supra- spinatus. Next the lateral cord gives off in suc- cession the ramus of the fifth cervical root to the N. phrenicus, and a ramus which joins with a branch of one from the medial cord to form the anterior thor- acic nerve to the pectoralis major; the second branch of this ramus from the medial cord passes to the pectoralis minor. Soon the lateral cord terminates in the N. musculo cutaneus and in the lateral head of the N. medianus. In some specimens this lateral head may receive a contribu- tion from the medial cord before the junction of the medial head itself. The medial cord may consist of a well- fused mass from the anterior divisions of the seventh and eighth nerves, or these two may combine only at a point some- what proximal to the formation of the medial head of the N. medianus. In the latter case it is clear that the contribution to the anterior thoracic nerves comes from the eighth cervical and that the latter also gives off the medial brachial cutaneous, the medial antibrachial cutaneous, and the intermediate antibrachial cutaneous nerves. The terminal branches of the medial cord are the medial head of the N. medianus and the N. ulnaris. The posterior cord like the median cord may form a group of loosely joined plexi- form branches with only a small part of its extent forming a true cord. It must be admitted that the separation effected by dissection is often artificial, but it serves to define more accurately the representa- tion of the cervical segments in the final common pathway. The first branch of the posterior cord is the N. thoracalis lateral- is which arises from the sixth cervical nerve near its spinal exit, and passes in- teriorly through some of the heads of origin of the M. scalenus anterior to emerge near the anterior border of this muscle and pass obliquely downward and posteriorly over the M. serratus anterior, terminating in its individual heads by segmental branches. The posterior cord gives off a series of subscapular branches to the muscle of that name. Those supplying the more posterior heads come from the fifth cervical, the next from the sixth, while those nearest the anterior border of the scapula are clearly derived from the seventh cervical nerve. The N. axillaris is formed by the pos- terior divisions of the fifth, sixth, and seventh cervical nerves. The remaining branches of the various contributions en- ter into the formation of the N. radialis. From the medial (eighth cervical) contri- bution, the N. thoracodorsalis to the M. [XXIX: 16 latissimus dorsi passes downward and latei’ally to distribute itself over the inner surface of the muscle. Near its exit from the axilla, the N. radialis gives off the muscular branch to the epitrochlearis and a ramus which joins the ulnar nerve and passes distally in its sheath to the mid portion of the arm where it passes to the deep-lying medial head of the triceps. On the right side the fifth cervical nerve contributes to the phrenic nerve with a ramus as large as that forming its source from the fourth cervical nerve. The ramus from the fifth receives a slender branch from the cervical sympathetic trunk and in addition gives off the motor branch to the M. subclavius. It then curves an- teriorly over the M. scalenus anterior to join the phrenic trunk. On the left side the phrenic nerve re- ceives contributions from the fifth and sixth cervical nerves as well as from the fourth. The nerve to the subclavius like- wise arises on the left side from the fifth cervical root of the phrenic nerve. Returning to structures about the shoulder joint, anteriorly the common origins of the Mm. coracobrachialis and biceps brachii, caput breve, superimposed on each other, form a stout tendinous band passing obliquely into the medial aspect of the arm. When this common tendon is severed and reflected, the N. musculocu- taneus is found along the posterior surface, giving branches to the M. coracobrachialis in this region. From the deep surface of the coracoid process beneath the M. cora- cobrachialis, a thin, flat muscle arises and passes laterally to insert on the antero- medial surface of the neck of the humerus just above the insertion of the latissimus dorsi on the medial lip of the bicipital groove. This muscle is supplied by a slender branch of the anterior humeral cii’cumflex artery, which curves laterally below its in- sertion, and it is innervated by a branch of the musculocutaneus nerve. Hartman & Straus (7) call this muscle the M. coraco- brachialis profunda in the rhesus monkey but its separate and distinct course in my opinion makes it worthy of distinctive des- ignation. I have called it the M. coracohu- meralis. In the region about the shoulder joint the combined tendon of the latissimus dorsi and dorso-epitrochlear muscles lies distally and deep to the coracobrachialis. These two muscles join at an acute angle in the pos- terior part of the axilla to form the broad, flat tendon which inserts on the superior part of the medial lip of the bicipital groove anterior to the insertion of the teres major. The latter occupies a position posterior to the former along the entire length of the medial lip. Between these tendons there is a large bursa. 1944] Robertson: Anatomy of the Woolly Monkey 175 In the bicipital groove the prominent rounded tendon of the long head of the biceps is seen passing interiorly to pierce the capsule of the shoulder joint from its origin on the supraglenoid tubercle. When the short head of the biceps is reflected with the coracobrachialis the origin of the anterior humeral circumflex artery is re- vealed arising from the axillary artery near the inferior border of the coracobrachialis. It passes beneath this muscle and the mus- culocutaneus nerve to give branches to the distal part of the pectoralis major and to the M. cor acohumer alls. A recurrent branch passes to the subdeltoid area of the capsule of the shoulder joint and anastomoses with branches of the posterior humeral circum- flex and acromial branches of the thora- coacromial arteries. The pectoral group of muscles is evident- ly quite variable and inconstant. In a full grown female (L. lagotricha) the pectoralis major arises from the inner third of the clavicle, from the sternum down to the end of the gladiolus and from the seventh costal cartilage. There are no fibers extending from the sheath of the rectus abdominis. The accessory pectoral muscle arises by two heads from the sixth and seventh costal car- tilages, the superficial somewhat lateral to the deep head. The two combine near their mid portions and become narrow and rounded. Beyond this point the muscle be- comes flattened and fan-shaped. It passes behind the lower border of the pectoralis major and the fan-shaped insertion joins with that of the latter near its superior part. The pectoralis minor arises from in- ferior, deep and superficial heads. The first takes origin from the sixth costal cartilage. The latter two from the third, fourth, fifth and sixth costal cartilages. All of these joint to insert in a common tendon on the lesser tuberosity. In this specimen there is no muscle which can properly be identified as the pectoralis abdominis. Posterior Thoracic Region. The M. atlantoscapularis super ficialis runs obliquely from the atlas to insert on the lateral extremity of the spine of the scapula near the base of the acromion and covers the acromial insertion of the tra- pezius. The M. trapezius has a broad transverse origin from a horizontal line on the oc- cipital bone and from the spinous processes of cervical vertebrae by direct, fleshy bun- dles. At the level of the lower cervical ver- tebrae the muscle bellies retreat laterally, giving place to a triangular aponeurosis which extends down to about the fourth thoracic vertebra. From that point to the end it arises again directly from the verte- bral spines and interspinous ligaments. The cervical portion is inserted on the superior surface of the outer third of the clavicle, on the acromion, and the entire length of the scapular spine. The lower portion in- serts only on the inferior margin of the medial third of the scapular spine. Inferior to the trapezius the posterior attachments of the deltoid are observed to consist of the posterior part of the acromial section and the spinal division which have been described. On reflecting the trapezius from the ver- t bral origins the M. rhomboideus capitis is revealed. This muscle originates by two heads. The medial head arises from the oc- cipital bone below the origin of the trapezius and from the nuchal ligament, while the lateral head arises by a fan-shaped origin from the occipital bone beneath the origin of the trapezius. These two heads merge to form a wedge-shaped muscle which in- serts on the upper third of the vertebral margin of the scapula near the superior angle. The M. rhomboideus cervicis arises from the spines of the cervical vertebrae and is inserted along the vertebral margin of the scapula just inferior to the insertion of the rhomboideus capitis. The M. rhomboideus dorsi arises from the first four or five thoracic spines and is in- serted along the vertebral margin of the scapula down to the inferior angle. The M. atlantoscapularis posterior which is revealed on a deeper plane when the trapezius is reflected laterally, arises from the transverse process of the atlas and passes downward and posteriorly to insert on the posterior superior scapular angle, lateral to the insertion of the rhomboideus capitis. When the fat, areolar tissue and the deep chain of cervical lymph nodes are removed from beneath the trapezius, the A. trans- verse, scapulae is observed passing posterior- ly and laterally. After crossing the superior margin of the scapula the artery gives off a long muscular branch to the trapezius and a superficial and deep branch to the M. supraspinatus. There is also a contribution to the periarticular anastomotic rete over the shoulder joint. The reflection of the trapezius reveals the origin and main body of the M. supi~aspi- natus which takes origin from and occupies the entire supraspinous fossa. Crossing its surface is the superficial branch of the transverse scapular artery which gives off muscular branches here and continues down- ward to anastomose with the terminal bran- ches of the posterior circumflex humeral artery. The deep branch of this artery passes beneath the supraspinatus, emerging at its lower border beneath the coracoclavic- ular ligament to contribute its acromial branches to the rete acromiale. From there 176 Zoologica: New York Zoological Society [XXIX: 13 the artery continues on as the A. transversa scapulae profunda beneath the root of the acromion, where it enters the infraspinous fossa to supply the M. infraspinatus and the M. teres minor. It is accompanied by the N. dorsalis scapulae which supplies the Mm. supraspinatus and infraspinatus. The M. teres minor is supplied by the N. axillaris. This ramus enters the muscle near its in- sertion. The M. supraspinatus forms a well- defined tendon which inserts at the highest point on the greater tuberosity of the hu- merus. In the infraspinous fossa lie the bellies of the M. infraspinatus and the M. teres minor. Passing over them superficially is the apo- neurotic (origin of the long head of the M. triceps) from the spine of the scapula. Medial to the long head of the triceps, the terminal muscular branches of the posterior humeral circumflex artery emerge from be- tween the teres major and minor, ending in the teres major, minor, and infraspinatus. Lateral to the long head of the triceps, the deltoid branches of the posterior circum- flex artery emerge with the axillary nerve. By reflecting the latissimus dorsi, after severing it near its insertion, the inferior angle of the scapula is revea.ed. Three mus- cles converge here: The teres major arises from this angle and the entire length of the axillary margin of the scapula. The lower bellies of the M. serratus anterior in- sert on the inferior angle. The lower fibers of the M. rhomboideus dorsi from the mid- thoracic spines pass superiorly to insert near the angle. The major part of the infraspinous fossa is occupied by the M. infraspinatus but it shares this area with the M. teres minor. The latter is flattened from above downward near its origin, but becomes flattened from before backward so that the fibers having the deepest origin are inserted at a higher point on the greater tuberosity. This inser- tion is directly below that of the M. infra- spinatus. When the rhomboid muscles are severed from their origins and reflected, the des- cending branch of the transverse scapular artery is found passing downward on the deep surfaces of the rhomboideus capitis and cervicis. The artery continues its course interiorly and contributes to the anastomo- ses along the vertebral margin of the scap- ula. The M. splenius capitis arises from the external occipital protuberance and nuchal ligament to the first or second thoracic spine. Overlying the lower part of this origin is the thin aponeurotic sheet of origin of the serratus posterior superior. Fleshy muscle bundles of the latter, three or four in num- ber, appear beyond the margin of the epaxial musculature and insert on the correspond- ing ribs. The insertion of the splenius capitis is on the superior nuchal line lateral- ly as far as the mastoid process. There is a prominent representation of the levator scapulae complex which arises from the transverse processes of the cervical vertebrae and inserts on the superior mar- gin of the scapula to the superior angle, occupying a space between the insertion of the M. atlantoscapularis posterior and the upper insertion of the M. serratus anterior and M. rhomboideus capitis. Medial and Anterior Aspect of the Arm. As the deep fascia is removed, the branch- es of the medial brachial cutaneous and medial antibrachial cutaneous nerves are revealed. The latter continues downward in the groove between the biceps and dorso- epitrochlearis to emerge near the antecubital fossa. These have been observed in the axil- lary fossa arising by a common trunk with the anterior thoracic nerves from the medial cord of the plexus. The intermediate anti- brachial cutaneous nerve, also a branch of the medial cord, runs along the medial bor- der of the short head of the biceps, crosses under the lacertus fibrosus and continues down the middle of the volar surface of the forearm to the wrist. When the short head of the biceps is separated from the dorso-epitrochlearis and reflected laterally, the neurovascular bundle of the arm is exposed. The A. brachialis occupies the following positions with respect to the median nerve which may be taken as a point of reference: The artery and vein lie lateral to the median nerve in the upper third of the arm. At its middle third the artery passes behind the nerve. Just before this occurs the A. profunda brachii arises and passes posteriorly in company with the N. radialis. The latter has occupied a posi- tion posterior to the artery up to this point. The A. nutricia humeri arises from the A. brachii in its upper third just below the A. profunda brachii. The brachial artery in the lower third of the arm lies medial to the median nerve and posterior to the bra- chial vein. In the arm the brachial artery gives off muscular branches to the long and short heads of the biceps, to the brachialis, the A. profunda brachii, and terminates in the antecubital fossa by dividing into the A. radialis and A. ulnaris. Curving laterally beneath the short head of the biceps, and becoming then for the first time clearly separated from it, the cora- cobrachialis is found to insert on the an- terior and medial aspect of the humerus between the origins of the brachialis and the medial head of the triceps. The coraco- brachialis is supplied by the N. musculo- cutaneus which curves around it laterally or pierces the insertion, giving off muscular branches to both heads of the biceps at this level and continuing medially and distally 1944] Robertson: Anatomy of the Woolly Monkey 177 to the middle third of the arm where it is joined by a branch of the median nerve. This combined trunk supplies the lower part of the M. brachialis and continues to the lateral side of the long head of the biceps where it pierces the deep fascia of the lower third of the arm to supply the skin in this region and in the antecubital fossa. The median nerve likewise gives off a stout branch in the middle third of the arm which passes laterally beneath the bi- ceps to reach the antecubital fossa. This branch continues down the volar surface of the forearm to the wrist as the anterior antibrachial nerve — it usually branches into two parallel rami in the forearm. On reflecting the brachial artery and vein and the median nerve laterally the en- tire course of the ulnar nerve to the elbow is exposed. This lies first medial to the vein, then behind it, lying beside the radial nerve to the point where the latter passes pos- teriorly with the A. profunda brachii. From that point it lies on the anterior border of the medial head of the triceps, parallel to the tendon of the dorso-epitrochlearis and passes posteriorly behind the medial epicon- dyle. The ulnar nerve gives off no branches in the arm. Lateral and Dorsal Aspect of the Arm. The muscles presenting on the dorsal and lateral aspect of the arm are the brachialis and triceps. The origin of the former encir- cles the insertion of the deltoid superiorly and occupies the lateral and anterior aspect of the humerus deep to the biceps. From the curving lower margin of the spinal portion of the deltoid the long and the upper part of the lateral heads emerge and soon fuse. From between the lateral head of the triceps and the brachialis the lateral cutaneous branches of the radial nerve emerge to supply the skin in this region. From be- tween the lateral head of the triceps and the brachialis below, in the supracondylar re- gion additional lateral brachial cutaneous rami appear. Still lower the lateral and dor- sal antibrachial cutaneous nerves spread out fanwise over the lateral and dor- sal aspect of the upper third of the fore- arm. The cephalic vein passes superiorly over the brachialis to attain the groove be- tween the deltoid and pectoralis major. The M. triceps may now be described in its entirety. The spinous and infraglenoid origin of the long head have been noted before ; the long head fuses with the lateral head on a level with the lowest point of insertion of the deltoid. The lateral head arises from a point inferior to the greater tuberosity of the humerus and on a level with the attachment of the capsule of the shoulder joint. This origin is essentially on the posterior surface of the humerus, by a narrow strip of aponeurotic tendon that is fused with the tendon of insertion of the deltoid. Distally, the origin moves laterally and occupies the middle third of the lateral border of the humerus. It shares the lateral intermuscular septum here with the M. bra- chialis. The medial head in Lagothrix is essentially a separate muscle. Its origin be- gins on the same level as that of the lateral head, but medial to it as far as the medial lip of the bicipital groove. From this point distally it occupies an extensive area on the postero-medial aspect of the humerus almost to the olecranon fossa. It forms a stout tendon which encircles the olecranon and inserts separately and deep to the tendon of the long and lateral heads. The lower portion of this medial tendon is split by a branch of the A. ulnaris collateralis (three specimens). Some of the lower muscle bun- dles arising from the medial supracondylar ridge insert still more deeply on the capsule of the joint posteriorly. The Mm. anconeus lateralis and medialis are essentially alike in representing further muscle bundles of the triceps complex. Each arises from the corresponding epicondyle and from the capsule of the elbow joint. They insert on the lateral and medial bor- ders of the ulna respectively. These mus- cles form the superficial covering of the tunnel through which the Aa. recurrens id- naris and radialis pass to join the anasto- motic rete about the elbow joint. The triceps complex is supplied by the N. radialis and the A. profunda brachii. The two bear important relationships to the triceps. At the junction of the upper and middle thirds of the arm, the artery and nerve pass posteriorly between the long and medial heads of the triceps. The radial nerve has already given off rami to the M. dorsopitrochlearis and to the long and lateral heads. In the intermuscular canal the A. profunda brachii gives muscular branches to the triceps and to the dorso- epitrochlearis; somewhat lower, the A. col- lateralis radialis arises. This latter emerges from the lower border of the lateral head of the triceps and passes distally to join the anastomotic rete about the elbow joint. In the middle third of the arm the radial nerve and profunda brachii lie in the radial groove of the humerus in direct contact with the bone and covered here by the conjoint bellies of the long and lateral heads of the triceps. In the lower third of the arm, the N. radialis approaches the surface in the angle formed by the divergence of the later- al head of the triceps and the M. brachialis. For a distance of a centimeter or so the N. radialis is exposed in this lateral area, lying parallel to the border of the brachialis. It then plunges deeply once more beneath the origin of the M. brachioradialis to enter the forearm. The lateral antibrachial cuta- 178 Zuologica: New York Zoological Society [XXIX: 16 neons and dorsal antibrachial cutaneous nerves are given off in this area. The insertion of the M. dorsoepitroch- learis is on the medial aspect of the base of the olecranon process. In addition to ex- tending the forearm, it is clear that this muscle will assist the biceps in supinating the forearm. The Forearm. On removing the integument, the forearm is seen to be sheathed in a stout fascial sheath having strong attachments to the bony prominences of the humeral epicondy- les, the olecranon and the borders of the subcutaneous surface of the ulna. There are strong thickenings forming dorsal and volar transverse carpal ligaments. Just distal to the antecubital fossa, on the medial aspect of the forearm, the prominent lacertus fibrosus of the biceps muscle can be traced across the flexor group of muscles almost to the medial border of the ulna, where it fuses with the deep fascia. A short distance proximal to it the medial anti- brachial cutaneous nerves pierce the deep fascia and are distributed to the skin over the dorsum of the forearm distal to the olecranon, and to the medial aspect of the forearm for the proximal two-thirds of its length. Attempts to remove the fascial sheath of the muscles of the forearm reveal the importance of this coat as an origin for some of these and the comparative rigidity of the compartments in which the individual muscles are contained. This is especially true of the flexor group which can be sepa- rated only with difficulty from each other since neighboring ones take origin from common intermuscular septa. When the deep fascia is removed the mus- cles of the forearm are seen to be divisible into three general groups. First, those which originate on the medial epicondyle of the humerus consisting of the flexors, and held in place by the fascial sheath which is firm- ly attached to the medial aspect of the sub- cutaneous surface of the ulna. The second consists of the extensor group of muscles which arise from the lateral epicondyle. Their fascial sheath has its strongest at- tachment to the lateral aspect of the sub- cutaneous border of the ulna. The third group consists of a single muscle, the bra- chioradialis, which has a broad, flat, fleshy origin from the humerus on the upper part of the lateral epicondylar ridge and in part medially from the tendon of insertion of the M. brachialis. In the tunnel between these two heads of origin the radial nerve passes into the forearm to join the radial artery and vein. If the biceps is reflected laterally, the median nerve, and the brachial artery and vein medial to it, are exposed lying on the brachialis tendon. Just proximal to the bi- furcation the brachial artery gives off the A. collateralis ulnaris. This is a slender branch passing over the medial head of the triceps. It joins the ulnar nerve and passes behind the medial epicondyle to anastomose with the A. recurrens ulnaris. Immediately after its origin the A. radialis passes later- ally across the median nerve and the tendon of the M. biceps, joining the radial nerve behind the M. brachioradialis. At this point it gives off the A. recurrens radialis which turns back at once and passes lateral to the tendon of the biceps and of the brachialis to join the anastomosis about the elbow. The A. radialis also gives muscular branches here to the M. supinator and to the brachio- radialis before continuing distally in the forearm. In the region of the bifurcation of the brachial artery in the arm, the A. ulnaris gives off the A. collateralis ulnaris inferior which passes medially under the tendon of the medial head of the triceps, giving off. as it passes, a slender branch to the ad- jacent origins of the M. pronator teres. If the M. brachialis is rotated laterally just above the point of formation of its tendon, a bare area of the humerus may be ob- served; there are no muscular attachments in this region. Here also the median nerve gives off muscular branches to the M. pro- nator teres before passing deeply in the angle formed by the insertion of this mus- cle on the radius. The tendon of insertion of the M. brachialis is partially revealed pass- ing obliquely downward in the antecubital fossa. It is seen to consist of very stout fibers forming a thin knife-like edge with the sharp edge presenting distally and fol- lowing its oblique course posteriorly to in- sert on the coronoid process of the ulna. The similarly flattened tendon of the biceps lies in the same sagittal plane, passing to its insertion on the tuberositas radii. The M. pronator teres forms the medial boundary of the antecubital fossa. It arises by stout fleshy fibers from the medial epi- condyle of the humerus. Its fibers sweep laterally and the upper ones insert on the deep or medial surface of the radius while the lower ones, which arise in part from the medial epicondyle and in part from the in- termuscular septum between this muscle and the origin of the M. flexor carpi radialis, form a broad, flat aponeurotic tendon which inserts on the lateral border of the curve of the middle third of the radius. On re- flecting the origin from the medial epicon- dyle a stout tendon is revealed lying on the deep surface of the muscle. Extending from a medial position this tendon passes through the mid portion of the muscle to insert on the radius and form a bridge beneath which the ulnar artery and median nerve pass. It should be noted that the lower third of the 1944] Robertson: Anatomy of the Woolly Monkey 179 medial border of this muscle presents a thin, free margin. When the pronator teres is reflected, the first part of the ulnar artery and median nerve are revealed. The artery soon joins the nerve and the two proceed distally to- gether in the forearm, deep to the flexor group and parallel to the ulna. In this region there are muscular branches of the median nerve to the flexor muscles. The M. flexor carpi radialis originates from the medial epicondyle and from the intermuscular septa which separate it from the pronator teres laterally and the flexor digitorum sublimis medially. It is a thin rounded muscle which becomes flattened near the wrist and passes laterally to insert on structures of the wrist region to be described with that part. It is innervated by the median nerve. The M. flexor digitorum sublimis occupies the next medial compartment of the flexor group. It has a narrow fleshy origin from the medial epicondyle and a broad flat aponeurotic tendon of origin from the ulna. This latter extends from a region about a centimeter from the olecranon to within an equal distance from the radio-ulnar joint. It is richly supplied by the median nerve and the ulnar artery gives off its muscular branch near the middle of the belly. In the oval hiatus created by the humeral and ul- nar origins of this muscle, the muscular branches of the ulnar nerve to the flexor carpi ulnaris pass to that muscle. When the flexor digitorum sublimis is reflected the humeral origin of the flexor digitorum profundus is observed to be a fleshy bundle which shares a common tendon with the humeral head of the sublimis. The humeral head of the flexor digitorum pro- fundus becomes flattened distally and ter- minates in a flat tendon which emerges with the principal tendon of its deeper origins. Relation of Nerves and Arteries to These Structures. The median nerve lies deep to the humeral head of the flexor digitorum sublimis for most of its length in the forearm, passing laterally in the lower third to occupy a median position beneath the tendon of the flexor carpi radialis. In this region it gives off superficial cutaneous branches to the volar surface of the wrist and proximal part of the palm of the hand. There are fine branches of the median nerve which pass in recurrent fashion about the ulnar artery at the point where it gives off muscular branches to the flexor digito- rum profundus. These run backward prox- imally to the elbow joint in the region where the ulnar nerve enters the forearm. At the point where these recurrent artic- ular branches turn proximally, the ulnar artery gives off the A. interossea communis. This may arise as a single trunk dividing at once into the Aa. interossea volaris and dorsalis, or these may arise separately from the ulnar artery. Beyond this point the ul- nar artery proceeds medially to join the ul- nar nerve passing beneath the humeral head of the flexor digitorum profundus. From this point on the artery courses distally to the wrist with the nerve, deep to the ulnar origin of the flexor digitorum sublimis and gives off prominent muscular branches to this muscle at the point of junction with the ulnar nerve. The M. flexor carpi ulnaris is the last superficial muscle of the flexor group. This takes origin from the most posterior part of the medial epicondyle of the humerus and from the major part of the medial border of the subcutaneous surface of the ulna. Its tendon of insertion passes volarwai'd to insert on structures of the wrist which will be described with that region. The M. palmaris longus is a thin and variable muscle, sometimes absent or fused with the flexor carpi ulnaris. It arises from the medial epicondyle and lies superficial to the other flexors. It soon gives rise to a thin, flat tendon that passes over the transverse carpal ligament to insert on the palmar aponeurosis. These two muscles are supplied by the N. medianus. The Extensor Group. Arising from the lateral supracondylar ridge of the humerus, just below the origin of the brachioradialis, is the thin and nar- row M. extensor carpi radialis longus. In the middle of the forearm this muscle gives rise to a narrow tendon which pursues a distal course parallel to the next muscle. The M. extensor carpi radialis brevis takes origin by a few fibers from the su- pracondylar ridge, but in the main from the lateral epicondyle and in part from the intermuscular septum which it shares with the next muscle. While the brevis is a stouter muscle than the longus, its tendon likewise begins in the middle of the fore- arm. However, bipenniform fibers continue to join this tendon almost to the wrist. It accompanies the longus into the wrist. Both of these muscles are innervated by the radial nerve and supplied by the radial artery. The M. extensor (digitorum communis arises by a stout, flat aponeurosis from the lateral epicondyle just posterior to the pre- ceding muscle. It is a narrow muscle sharing the intermuscular septum with the extensor carpi radialis brevis and continues as one muscle almost to the wrist. Just above the dorsal carpal ligament, however, two bellies diverge and give rise to closely compressed bundles of tendons which enter separate compartments of the ligament. The nerve supply is furnished by the N. radialis. The M. extensor carpi ulnaris arises from 180 Zoologica: New York Zoological Society [XXIX: 16 the posterior aspect of the lateral humeral epicondyle and for a short distance its proximal portion takes origin from the sheet of the deep fascia that binds it to the lateral border of the ulna. This thin narrow muscle runs parallel to the others and the flat ten- don passes somewhat medially over the head of the ulna to the medial side of the wrist. When the origins of the last two muscles are separated, the dorsal interosseus artery is revealed emerging from between the M. supinator and the ulnar head of the ab- ductor carpi radialis. This gives muscular branches to the M. extensor digitorum lon- gus and the M. extensor carpi radialis. Both of the latter muscles are innervated by branches of the N. radialis which pierce the ulnar head of the M. supinator to reach the dorsum of the forearm. These branches are from the ramus profundus of the radial nerve. Reflection of the origins of the extensor muscles reveals the extensive origin of the M. supinator. This arises by a wide, fleshy and aponeurotic origin from the lateral epi- condyle of the humerus and surrounds the radio-humeral articulation on all but its dor- sal aspect. It arises in part, also, by a deep head from the upper third of the ulna. The more medial fibers arising from the epi- condyle pass in a sagittal direction distally to insert on the medial aspect of the radius. It is likely that these fibers act to flex the forearm. On the mediolateral aspect of the muscle there is a stout tendinous band which passes almost uninterruptedly to the radial insertion. It is beneath this that the ramus profundus of the radial nerve passes to the dorsum. The M. supinator is innervated by the radial nerve and supplied by the radial artery. The deep muscles of the dorsum of the forearm are the abductor carpi radialis and a muscle complex to be described presently. The former arises by a long narrow head from the lateral aspect of the ulna, from a point on a level with the head of the radius for the full extent of the upper third of the ulna. From this point on it verges laterally, taking origin in part from the interosseus membrane and in the middle third of the ulna by an aponeurotic sheet. Fleshy fibers again take origin from the lower third of the ulna almost to its head. The radial head begins narrowly along the border of the in- sertion of the M. supinator and continues distally over the middle third of the radius. Beyond the M. supinator, this margin of the muscle is free. The two heads converge on a central tendon which moves laterally to occupy the lateral border of the muscle where it crosses over the Mm. extensores carpi radialis longus and brevis to enter a compartment of the dorsal ligament of the wrist. Beyond this the tendon passes in its own serous sheath divided into two which insert on the Ossa multangulum ma- jus and minus. The tendon is crossed in this region by the radial nerve and artery. The last deep muscle is a complex extensor which arises from the middle third of the ulna along its lateral border lying deep to the extensor carpi ulnaris. Near the wrist this gives rise to two tendons; the lateral passes in a separate compartment in an oblique direction laterally beneath the com- parment containing the M. extensor digito- rum communis. The medial tendon and the tendon bundle of the latter muscle pass through the same compartment. These are the Mm. extensor indicis proprius and ex- tensor pollicis. Both of these deep muscles are innervated by the ramus profundus of the N. radialis and supplied by the A. in- terossea dorsalis. Extensor Tendons on the Dorsum of the wrist. The relation of the tendons of the exten- sor muscles in their respective synovial sheaths as they pass through the dorsal (radio-ulnar) ligament are as follows from the lateral to the medial side: The tendon of the abductor carpi radialis divides into two distinct tendons proximal to the point of entrance into the carpal sheath. The more lateral of these inserts proximal to the medial one. Its fibers origi- nate entirely from the radius, and while there are mutual insertions of the fibers of the radial and ulnar heads on the tendons of the other, these may be considered as two separate muscles. The tendons enter the common sheath obliquely and course in a groove on the dorsum of the ulnar head. Both insert on carpal bones, the medial and longer tendon curving over to the volar sur- face of the wrist where it inserts on the Os multangulum minus. The proximal tendon inserts on the Os multangulum majus. The M. abductor carpi radialis corres- ponds to the Mm. abductor pollicis brevis and longus in man. I have chosen to call it the carpal abductor, since it has its principal insertions on the carpus and it is evident that its most important action is to abduct the hand rather than the pollex or the first metacarpal alone. Even though there may be fibers of insertion on the base of the first metacarpal, the strength of the carpal unit insures the movement of the entire mass rather than that of the comparatively weak pollex. It also extends the wrist when the latter is fixed by the extensors and the M. extensor carpi ulnaris. Beneath the tendons of the abductor carpi radialis, the tendons of the extensor carpi radialis longus and brevis pass in their own sheath. As they emerge on the dorsum of the hand they separate from one another, the longus inserting on the carpo-metacarpal joint and on the base of the second meta- 1944] Robertson: Anatomy of the Woolly Monkey 181 carpal bone. The brevis inserts correspond- ingly on the third metacarpal bone and the carpo-metacarpal joint. Crossing over the insertion of the M. ex- tensor carpi radialis brevis, two tendons are observed to diverge near the middle of the second metacarpal bone; the lateral one is the tendon of the M. extensor pollicis, the medial one is the tendon of the M. extensor indicis proprius. Just before the latter pass- es to the index it sends an extension into the medial aspect of the tendon of the ex- tensor pollicis. On tracing these tendons backward to the proper carpal canal, they are found to arise from the single deep muscle in the forearm arising from the mid- dle third of the ulna. This lies parallel to the extensor muscle that with it occupies the deepest compartment of the dorsal canal. This latter muscle joins the tendon of the extensor digitorum communis. The tendons of the extensor digitorum communis occupy the centra] compartment and these are joined on the deep surface by the tendon passing to the index and pollex. This latter divides into three, the lateral joining the under surface of the tendon to the index, the other two proceeding to the third and fourth fingers respectively. The tendon of the extensor digitorum communis itself divides immediately into five divisions which are distributed as follows: The first passes to the index alone after giving off the connecting band to the tendon of the third finger. The second division passes exclusively to the middle finger. The third straddles the third and fourth, giving tendons to each. The fourth gives a con- tribution to the fifth in the metacarpal re- gion and then passes directly to the fourth finger. The fifth straddles the fourth and fifth digits, giving tendons to each. The medial part of the extensor digitor- um communis may be considered as a sep- arate muscle although it has a common origin with the communis. However, its belly may be separated almost the entire length of the forearm and it passes through a separate, more medially placed compart- ment in the wrist. In the carpal region the tendon divides to give a lateral tendon to the fourth finger which passes deep to the tendon of the extensor digitorum communis and a medial one which may be considered as constituting the tendon of the M. exten- sor digiti quinti. The most medial compartment of the dorsum of the wrist is occupied by the ten- don of the extensor carpi ulnaris. This courses medially in an oblique direction in its groove on the head of the ulna to pass volarward and inserts on the volar surface of the base of the fifth metacarpal. Each common extensor tendon spreads out in a flat sheet over the metacarpophalan- geal joint. This sheet is continued over the proximal phalanx and extends volarward with well defined margins on each side about half the thickness of the digit. This sheet inserts laterally on the base of the second phalanx. The central dorsal portion con- tinues as a long triangular extension pass- ing over the dorsum of the second phalanx ; its base inserts on the base of the distal phalanx. The Dorsal Carpal Ligame?it. The dorsal (transverse) carpal ligament is exceptionally stout and provides firm restraint for the extensor tendons. There are three major and distinctly separate parts. The first extends from a ridge on the medial side of the base of the radius over the extensor tendons, the head of the ulna, and inserts on the apex of the Os trique- trum below the Os pisiforme. The second arises from a point distal to the origin of the first, beneath the tendon of the extensor carpi radialis brevis. This ligament extends medially parallel to the first; its insertion is broader than its origin. The proximal attachment is to the head of the ulna and distally it curves about the extensor tendons to form the medial wall of their restraining canal, inserting on the carpus. The third originates on the base of the radius lateral to the tendon of the extensor carpi radialis longus and inserts in part on the lip of the groove in which the tendon of the abductor carpi radialis is confined and continues distally to insert on the Os multangulum majus. A fourth component covers the groove in which the tendons of the Mm. extensor carpi radialis longus and brevis lie. The divisions of the extensor tendons of the hand have important functional implica- tions. The existence of a separate, common extensor of the pollex and index helps to explain the observation that these two digits act as a joint unit. Further, its ulnar origin and the lateral course of its tendons serve to abduct this thumb unit of the hand from the remainder of the digits. The common extensor of the digits arises from a region that is medial with respect to the origin of this muscle and yet the tendons of both are fixed in the carpal canals at essentially the same point. Thus, when the muscles contract they are exerting their effective pull in es- sentially opposite directions. The only hint of functional independence of the remaining digits is that when sep- arated by rough dissection, the muscular bellies of the individual parts are found to originate from progressively dorsal posi- tions on the lateral epicondyle, those of the fourth and fifth fingers become fleshy at a point considei’ably below the bellies of the bundles going to the second and third fin- gers. The common origins from intermus- cular septa and the firm sheath in which 182 Zoologica: New York Zoological Society [XXIX: 16 they are held makes the functional inter- pretation of these details difficult. The Mm. extensor carpi radialis longus and brevis not only extend the hand, but when the latter is in a position of radial deviation, they tend to adduct it in an ulnar direction and bring it into the median axis. The same is true when the hand is in the position of ulnar deviation. When the principal extensors of the digits are elevated from their canal in the wrist, it is clear that the depth of this passage with the firm bony walls formed by the grooves in the base of the radius and the head of the ulna and the strong ligamentous roof provide fulcra which direct the effec- tive pull of the muscles. Examination of the origins and insertions of the respective muscles makes clear the combined extend- ing and separating action. The extensor digiti quinti takes origin lateral to the median axis of the wrist, but its insertion is well to the medial side of the axis; its tendon passing through the canal is firmly held in the median line and hence abduction of the fifth occurs with extension, when the exclusive motion of this digit effected by the extensor is considered. The origin and insertion of the extensor pollicis and indicis are the direct opposite of the foregoing, but the essential movement is of the same nature, combined extension and abduction. The interlacing fibers between the tendons to the third and fourth fingers indicate in my opinion a relative interdependence be- tween the motion of extension of these fin- gers produced by their extensors. That is, extension of one without simultaneous ex- tension of the others would appear to be impossible anatomically. While the axis of effective pull of the extensor indicis proprius points to an independent motion, in as- sociation with that of the pollex, the equally strong tendinous contribution from the ex- tensor digitorum communis suggests a be- ginning transition to an association of the movements of this finger with that of the others opposed to the thumb. The modifi- cation of these considerations by the action of the intrinsic muscles of the hands will be discussed elsewhere. When the extensors are reflected, a long- branch of the ramus profundus of the N. radialis is revealed on the deep surface of the M. extensor pollicis et indicis and may be traced to the dorsum of the wrist, the hand and into the wrist joint itself. The terminal branch of the A. interossea volaris emerges from the lower margin of the M. abductor carpi and anastomoses with the superficial branch of the radial artery and other contributors to the Rete carpi dorsalis. The superficial branch of the A. radialis which curves dorsally over the base of the radius to gain the dorsum of the hand, passes superficial to the tendons of the M. extensores carpi radialis longus and brevis and turns medially forming the Arcus carpi dorsalis. It continues medially deep to all the extensor tendons and anastomoses with the ramus dorsalis of the A. ulnaris. On the dorsum it sends anastomotic rami proxim- ally to the Rete carpi dorsalis and at each interspace including the first, gives off Aa. metacarpi dorsales. Each of these gives off a digital artery to the corresponding side of the digits and perforating branches which pass between the heads of the metacarpals to anastomose with the volar metacarpal arteries. Superficial Structures of the Palm. The skin of the palm is densely attached to the palmar fascia and bound to the first and fifth metacarpal bones and to the volar carpal ligament at the wrist. When this is removed the strong palmar aponeurosis is revealed occupying a central position over the carpus and metacarpal areas. There are extensions of variable strength and promi- nence reaching the bases of the digits. Arising from the ulnar margin is the M. palmaris brevis which passes medially to insert on the skin of the medial border of the palm. It is innervated by the ulnar nerve. The palm is well padded with lobules of fat which are held in firm compartments of fibrous tissue. These pads of fibro-adipose tissue emerge between the four principal ex- tensions of the palmar fascia running to the bases of the second, third, fourth and fifth digits. The pads receive a rich nerve supply from the metacarpal divisions of the nerves of the palm. When the palmar fascia is severed in the middle of the palm and reflected distally it is seen to be tightly bound to the medial side of he palm by stout projections that insert on the lateral border of the fifth me- tacarpal and by deep extensions it is firmly bound to the metacarpo-phalangeal joints. Proximally the palmar fascia is fused with the transverse carpal ligament which forms the volar roof of the carpal flexor canal. Medially, the ulnar artery emerges to give rise to the Arcus volaris super ficialis. This curves laterally over the palm and is visible between the digitations of the pal- mar aponeurosis. It gives off Aa. metacarpi • volar es, each of which terminates in the cor- responding interspace to give rise to the Aa. digitales volares propriae. Emerging with the ulnar artery is the N . ulnaris which gives off terminal branches to the lateral side of the fifth digit, the medial side of the fourth digit and after joining with a stout branch of the N\ medianus, supplies the lateral side of the fourth digit and the medial side of the third digit. It should be noted that the N. ulnaris passes in a superficial compartment over the trans- verse carpal ligament, while the N. media- 1944] Robertson: Anatomy of the Woolly Monkey 183 nus passes through deep to this ligament in company with the flexor tendons. The median nerve supplies in the hand, the lateral side of the thumb by an independent branch, the medial side of the thumb and the lateral side of the index, the medial side of the index and the lateral side of the middle finger and, in conjunction with the ulnar, the medial side of the middle and lateral side of the ring finger. The super- ficial branch of the N. ulnaris passes be- neath the u’nar artery and the Arcus pal- maris super ficialis to the lateral side of the fifth finger and medial side of the fourth. On reflection of the palmar fascia in the proximal part of the palm, the relations of the N. medianus, N. ulnaris and the A. ul- naris are revealed. The superficial branch of the median nerve passes medially over the tendon of the M. flexor digitorum sublimis under which it has coursed up to this point. It is joined, beneath the inser- tion of the M. flexor carpi ulnaris on the Os pisiforme by the ramus palmaris of the ulnar nerve and the ulnar artery. These enter a compartment beneath the overhan- ging Os pisiforme. Here there are plexiform branches of the median nerve that pass to the ulnar nerve. The ulnar nerve divides to give off a superficial branch which passes to the medial side of the fifth finger and a deep branch which passes dorsally between the M. abductor digiti quinti and the M. flexor digiti quinti brevis. The floor is formed by the stout transver- se carpal ligament. This has exceedingly strong attachments to the carpus and forms the roof of the deep carpal canal through which the tendons of the flexor muscles pass. If the transverse carpal ligament is div- ided longitudinally, the carpal canal and its contents are revealed. The canal is smooth- ly lined and its thickness and rigidity clear- ly act as a restraint to the flexor tendons. Lying on the tendons superficially the N . medianus is loosely bound to them by loose areolar tissue. This emerges in the palm below the distal border of the transverse carpal ligament and divides to form three metacarpal branches passing to the first, second and third interspaces and corres- ponding sides of the thumb and fingers. A fourth branch gives off the muscular supply to the M. flexor pollicis brevis and continues distally to supply the lateral aspect, of the thumb. The most medial branch is joined by a branch of the superficial ramus of the N. medianus. It is probable that these plexi- form intercommunications have no great functional significance aside from the ob- vious overlapping representation of impor- tant sensory areas. The M. flexor digitorum sublimis tendons in the wrist lie superficial to those of the M. flexor digitorum profundus, but in the carpal canal they lie medial to the latter. The flexor sublimis may be reviewed now that it is revealed in its entire length. The four tendons to which it gives rise may be fol- lowed proximally into four distinct bellies, but of these latter only that whose tendon passes to the index remains comparatively distinct in its entirety. In the canal the tendon to the index lies deep to the other three and arises from that part of the mus- cle lying most deeply and constituting the part taking origin from the ulna. However, the pointed upper extremity of the belly joins with the others to take origin from the medial epicondyle of the humerus. The strong flat tendon forms on the lateral aspect of the muscle, the lower fibers of which take origin directly from the ulna. It would seem that these tend to draw the index toward the midline or median axis of the hand when flexing the digit. The remaining tendons and the bellies from which they derive are as follows: The division for the fifth digit is the most medial and superficial. It takes origin from the strong aponeurotic covering of the muscle which fixes it to the epicondyle. This apo- neurosis sweeps medially to attach to the ulna and the lower fibers of this belly arise from it. The intermediate belly takes origin solely from the humeral epicondyle and in its upper part lies superficial to the division going to the middle finger. The latter is a typical bipennate muscle deriving its fibers of origin chiefly from the epicondyle, deep to the last muscle, but in addition receives a sheaf of fibers from the common apo- neurotic tendon of this muscle attaching to the ulna. Its fibers pass deep to those of the index division. The tendon comes to lie lateral to those to the fourth and fifth finger. These tendons as they lie in the carpal canal are simply held together by loose areolar tissue. There is apparently no synovial sheath. When the flexor sublimis is elevated the common tendon of the M. flexor digitorum. profundus is found to lie in part lateral and in part deep to that of the sublimis. The two separate and distinct bellies of this muscle as seen in the forearm, the one arising from the entire length of the radius and by a deeply lying humeral head, the other arising from the entire length of the ulna, are easily separable from one another. The rami of the N. medianus enter the bellies at a point above their juncture in the upper third of the forearm. In the carpal canal, however, the tendons of these two join by a broad union to form a single flat band which immediately divides into five individual tendons. When the tendons are severed in the wrist and displaced distally, the deep surface presents an in- teresting complexity. Medially the tendon of the digiti quinti 184 Zoologica: New York Zoological Society [XXIX: 16 may be traced directly from the lateral por- tion of the ulnar division of the muscle, and this may be traced as a separate muscle belly in the forearm lying medial to the others. However, this tendon sends a prom- inent contribution to the tendon of the fourth finger. The latter is formed of the remaining bundles from the ulnar division and from a stout contribution of the radial muscle. The tendon of the middle finger proceeds directly from the radial muscle, and the fibers of this tendon and that of the index are in part seen to curve pos- teriorly from the anterior part of the mus- cle. Thus, the tendon of the pollex is par- tially sheated in its first portion. It should be pointed out here that there is no tendon of the flexor digitorum sublimis to the thumb. This interlacing structure of the tendons no doubt serves to reinforce the combined grip of the individual fingers. The flexor profundus may be thought of as a muscle chiefly employed for climbing and hanging suspended, actions which involve all the fingers at once and require none of the re- finements of independent motility. The tendons in the carpal canal are not surrounded by a sheath, and the Vaginae propriae of the deep tendons do not begin until the point at which they are joined by the tendons of the M. flexor sublimis; both tendons enter a common sheath at this point. The fibers passing to the tendon of the index curve from the antero-lateral as- pect of the common tendon, around the ten- don of the pollex. The lowest muscle fibers arising from the radius curve medially and posteriorly to be continued as the tendon of the index. Thus, the tendon of the thumb arises from that portion of the muscle lying on the anterior surface toward the mid-line. The tendon of the third finger is derived from that portion of the muscle lying most medially and the fibers may be traced chiefly to the humeral head. The tendon of the fourth finger receives a con- tribution from the tendon of the fifth, which joins the remaining fibers of the ulnar muscle, and the stout tendon which passes to it from the radial muscle on the deep surface of the tendon mass. The four Mm. lumbricales consist of thin spindle-shaped muscles taking origin from the bifurcation of the tendon of the ex- tensor digitorum profundus and for a short distance from the adjacent sides of these. They lie in the space between these tendons until they reach the metacarpo-phalangeal joints where they enter a canal or sheath formed by extensions of the fascial slips from the'digitations of the palmar fascia where it becomes firmly attached to the joint capsule. Between the heads of the metacarpals there are well-defined intercapitular liga- ments consisting of flat sheets blending with the metacarpo-phalangeal joint cap- sules. These sheets form slings in which the lumbrical muscles rest, separating them from the deeper intrinsic muscles. The ; lumbricales pass in separate fibrous com- partments over the antero-lateral aspect of the metarcarpo-phalangeal joints to insert in each case on the volar border of the extension of the broad tendon of the M. I extensor digitorum communis. A single ex- ception is the fourth lumbrical which has been found to insert in one case on the lateral aspect of the proximal phalanx of the fifth digit rather than on the adjacent extensor tendon. With the fingers flexed on the palms the lumbricals serve to extend the second phalanx on which this portion of the extensor tendon inserts. The first lumbrical inserting on the index, by reason of its origin well to the medial side of the axis of this finger, will produce some degree of abduction also. In the middle of the palm the tendons of the flexor digitorum sublimis join the ten- dons of the flexor digitorum profundus and at this point the sheath of the latter opens and extends deeply to include the two ten- dons which from that point to their inser- tions occupy a common sheath. Over the metacarpo-phalangeal joint the sheath pre- sents a circular thickened band which strongly confines the tendons; beyond this the capsine again becomes thin and mem- branous. Similar annular ligaments are found at each interphalangeal joint. At points between these there are additional annular ligaments near the middle of Hie proximal and middle phaianges. These cor- respond to the ngamenta cruciata in man, buc in Lagothrix they are simpie circulate bands wmch are firmiy attached to the borders ol tne groove in which the tendons lie. Near the middle of the proximal phal- ; anx the overlying tenoon ox the flexor suo- ] iimis divides to permit the tendon of the flexor protundus to emerge and continue us central course to insert on the base ox the distal phalanx, it also is divided and the two parts diverge slightly at the point of insertion. The two parts of the flexor sub- limis occupy positions on each side of the deep tendon. At the middle of the second phalanx the tendon of the profundus spreads out, partially covering the insertion of the sublimis on the margins of the volar groove of the phalanx in which the tendons lie. The flexor tendons possess well-developed vincula passing to their under surfaces fi’om the hollow of the sheath. There is one from the middle of the proximal phalanx to the sublimis tendon at the point of divi- sion and a second just proximal to the first interphalangeal joint. A third joins the ten- don of the profundus proximal to the second interphalangeal joint. 1944] Robertson: Anatomy ..of, the Woolly Monkey 185 The tendon of the pollex deriving from radial division of the flexor digitorum pro- fundus passes unaccompanied to the thumb. There are the usual annular ligaments over the joints and a single very stout annular ligament binding the tendon firmly to the distal half of the proximal phalanx. Here, too, the tendon shows the usual division of the deep tendons and inserts similarly on the distal phalanx, but at a point some- what nearer the distal tip. Removal of the flexor tendons reveals the first layer of the intrinsic muscles of the hand forming the floor of the palm. The most prominent of these is the fan-shaped M. adductor pollicis. This muscle is variable in extent, but occupies roughly half of this region of the palm. It is triangular in shape and consists of four heads of incon- stant shape, although the origins and gen- eral configuration are readily identified in each case thus far examined. The proximal head takes origin from the carpo-metacarpal ligament at the base of the third metacarpal and courses distally on the upper portion of the common tendon of the muscle. This head also takes origin from a curved tendon whose concavity is directed medially, extend- ing from the carpo-metacarpal ligament to terminate in the joint capsule of the fifth metacarpo-phalangeal joint. The second head takes origin from the convex border of the curved tendon. The third and fourth arise variably from this tendon or from separate tendons, the third from the base of the fifth finger and fourth from the re- gion of the fourth metacarpo-phalangeal joint. The fibers of each converge on the common tendon which is inserted on the medial aspect of the base of the proximal phalanx of the thumb and in part upon the extensor tendon. It is innervated by the ramus profundus of the ulnar nerve as this crosses the palm; the nerve enters the deep surface of the muscle. When the M. adductor pollicis is reflected, a bicapitate muscle is revealed arising from the upper portion of the curved central ten- don from which the M. adductor pollicis also originates and from the slip that takes origin from the base of the fourth finger, converges on a common tendon which in- serts on the medial border of the extensor tendon of the index. This is the M. adductor indicis proprius. In some cases it inserts largely on the lateral aspect of the proximal phalanx. In some specimens this muscle is as well developed as the adductor pollicis and consists of stout fleshy bellies arising from the entire length of the central tendon from which the adductor pollicis takes origin. In such cases it continues down as a prominent muscle bundle to insert on the medial aspect of the proximal phalanx. This duplication of adductor muscles is strong evidence in support ofi the combined action of the thumb and index as a “pollicial unit.” The M. abductor indicis takes origin from almost the entire length of the first and second metacarpals and at its proximal tip from the carpus. The fibers converge on a narrow tendon which passes over the lateral aspect of the metacarpo-phalangeal joint. In this region the tendon passes through a fibrous sheath passing over it from the extensor tendon. The abductor indicis is inserted on the strong circular band which surrounds the base of the phalanx and is continued on the volar surface as the an- nular articular ligament of the flexor ten- don. The abductor indicis receives the ter- minal branch of the ramus profundus N. ulnaris. The remaining intrinsic muscles of the palm present a degree of complexity which does not lend itself readily to sys- tematic description. There are three muscul- ar heads taking origin on the same plane as the abductor pollicis and indicis from a central tendon extending from the lower border of the deep transverse carpal liga- ment. The most medial of these passes to the lateral aspect of the fifth digit where it is joined by a deep head arising from the carpus and from the lateral aspect of the fifth metacarpal. This combined muscle forms a narrow tendon which passes over the lateral aspect of the fifth metacarpo- phalangeal joint and is inserted on the ex- tensor tendon. It is innervated by the ramus profundus of the N. ulnaris. A small slip of the superficial head passes deep to insert on the lateral aspect of the fifth metacarpal just proximal to the metacarpo-phalangeal joint. The second head arising from this common tendon passes to the medial side of the fourth metacarpal where it is joined by a deeper head also arising from the carpus and from the medial aspect of the volar sur- face of the fourth metacarpal. The tendon formed by these two passes over the antero- medial aspect of the fourth metacarpal joint and inserts on the extensor tendon. In the fourth interspace the dorsal in- terosseus muscle, arising from the carpus, the lateral aspect of the fifth metacarpal and the medial aspect of the fourth converges on a stout tendon which passes over the medial aspect of the fourth metacarpo-phalangeal joint and inserts on the base of the proximal phalanx of the third digit. The third and medial superficial head arising from the central tendon joins a deeper head arising from the carpus and the lateral aspect of the fourth metacarpal. These converge on a common tendon which inserts on the extensor tendon proximal to the insertion of the third lumbrical. On the medial and lateral aspects of the middle finger there are symmetrical muscles consisting of superficial heads arising from 186 Zoologica : New York Zoological Society [XXIX: 16 the carpus and the medial and lateral aspects of the third metacarpal and deep heads which are the dorsal interossei arising from the lateral and medial aspect of the third and fourth and from the second and third metacarpals, respectively. The heads con- verge on tendons which pass on the cor- responding sides of the metacarpo-phalan- geal joint to insert on the extensor tendons. The second metacarpal presents on its medial volar aspect a superficial head cor- responding to those described above, origin- ating from the carpus and the medial aspect of the bone and giving rise to a tendon which inserts on the extensor tendon. On the lateral aspect of the second metacarpal there is a similar superficial head with its origin on the carpus and the lateral side of the volar aspect of the bone. This inserts on the volar surface of the proximal phalanx anterior to the insertion of the abductor indicis. All of these muscles are innervated by the ramus profundus of the N. ulnaris. This nerve curves laterally beneath the combined aponeurotic tendon of the three superficial heads of the medial group and to the volar surface of the heads of the second, third and fourth metacarpals it gives off a fine branch which terminates in a fibro-fatty pad here or is distributed to the joint. It is impossible to trace these to their ter- mination exactly but it is assumed that they subserve proprioceptive perception. Intrinsic Muscles of the Fifth Digit. The M. abductor digiti quinti consists of a fleshy belly constituting the principal mus- cle mass of the hypothenar eminence. It takes origin from the volar transverse car- pal ligament and from the pisiform bone. The slender rounded tendon begins halfway in its length and inserts on the base of the proximal phalanx of the fifth digit. Its functional activity is improved by the pas- sage of this tendon through a firm fibrous ligament which is a continuation of the annular articular ligament of the flexor ten- dons and by a fibrous sheet extending dor- sally from the border of the insertion of the flexor digiti quinti brevis into the extensor tendon. It is innervated by the ramus pro- fundus of the N. ulnaris. The M. flexor digiti quinti brevis arises by an aponeurotic tendon from the trans- verse carpal ligament. Its origin provides the lateral boundary of the tunnel through which the ramus profundus of the ulnar nerve passes. This muscle pursues a some- what oblique course from its laterally placed origin to its insertion on the medio.ateral aspect of the base of the first phalanx of the fifth digit and on the extensor tendon. It is innervated by the ramus profundus of the N . ulnaris. The M. opponens digiti quinti lies deep to the flexor brevis and takes origin from the carpus over the fifth carpo-metacarpal joint. Its fibers insert on the lower half of the volar surface of the fifth metacarpal. The functional significance of the firm fibrous sheath which encloses the muscle is not clear. Intrinsic Muscles of the Pollex. The M. abductor pollicis brevis takes orig- in from the carpus just proximal to the carpo-metacarpal joint and passes obliquely and laterally to insert on the lateral aspect of the base of the first phalanx. Its tendon is held in place by a stout fibrous sheath extending between the extensor tendon and the insertion of the flexor pollicis brevis. It receives its nerve supply from the N. medianus. The M. flexor pollicis brevis, which is in- nervated by the N. medianus, takes origin from the volar transverse carpal ligament and passes obliquely to insert in part on the lateral aspect of the proximal phalanx and in part on the extensor tendon. Its in- sertion partially encircles the tendon of the abductor pollicis brevis. The M. opponens pollicis is a variable muscle, not so well defined as the opponens digiti quinti. It takes origin from the volar transverse carpal ligament deep to the flexor brevis and inserts in some cases on the volar surface of the first metacarpal and some- times in common with the flexor, but slight- ly proximal to it on the head of the meta- carpal over the joint. Innervation is from the N. medianus. The Arcus volaris profundus is formed from a terminal branch of the ulnar artery which accompanies the ramus profundus of the N. ulnaris beneath the superficial layer of muscles and courses beneath their origins from the medial to the lateral side of the palm supplying fine branches to the intrinsic muscles and numerous contributions to the rete volaris carpi. There are also fine meta- carpal branches passing to the interspaces where they anastomose with both dorsal and volar metacarpal branches from the super- ficial volar arch and the dorsal arch. The M. pronator quadratus may be reveal- ed by separating the flexor bellies and re- flecting them to either side. It is a quadri- lateral sheet of muscle fibers with its origin on the volar surface of the ulna at a point somewhat higher than its insertion on the radius. Thus, the fibers sweep distally and laterally to the volar surface of the distal extremity of the radius across the interos- seus membrane. When the muscle is lifted up it is found to arise also from this struc- ture. The insertion on the radius is over a broad area of the volar surface from the lateral to the medial margin. This extensive insertion would seem to signify the impor- tance of this muscle in the initiation as Robertson: Anatomy of the Woolly Monkey 187 1944] well as the completion of pronation. It is supplied by the volar interosseous artery and the volar interosseous branch of the N. medianus. These latter enter its deep surface and the artery itself terminates in fine anastomotic branches which pass in the medial dorsal compartment beneath the ex- tensor tendons to join the dorsal carpal rete. The Shoulder Joint. The shoulder joint ( articulatio humeri) is a diarthrosis of the ball and socket type. 1 As in other primates, including man, it : depends on the strong muscles surrounding I it and the stout tendons of their attach- | ments for its strength. The capsule of the joint itself is relatively loose and is only thickened at those points where the mus- cular tendons blend with it. It surrounds the glenoid cavity of the scapula and is at- tached on an oblique line around the head of tne humerus. On the antero-medial aspect there is an opening to give exit to the tendon of the long head of the biceps which is [j intra-articular up to that point from its origin. This tendon passes over the head of j the humerus from its origin on the supra- g.enoid tubercle of the scapula. The second muscie entering into intimate relations with the joint capsule is the M. sybscapularis. The tendon of this muscle sweeps anterior- ly about the neck of the humerus and its upper fibrous border blends with the cap- sule of the joint. This tendon forms a thickened ligament presenting a free mar- gin within the joint cavity. There is a notch in the edge of the glenoid cavity into which the subscapular tendon fits. The portion of the subscapular tendon forming this ligament can be traced to the scapula and { . idenc.y limits external rotation of the humeius, since the musc.e hoers are reduced in numoer in this part of the muscle. Ihe Lig amentum, acromio-claviculare has a strong attachment to the anterior tip of the acromion, beneath the acromio-clavicular joint. It curves superiorly to attach to the under surface of the Ciavicle just proximal to the acromio-clavicular joint, forming the posterior retaining ligament of this arti- culation. The Ligamentum coraco-claviculare ex- tends from a point on the inferior surface of the clavicle just a short distance beyond the L*. acromio-claviculare and curves medially to the base of the coracoid process. There is a strong band of fibers which curves upward from the acromion to the clavicle and down to the coracoid process, uniting the foregoing, and while it appears to be a separate iigament, the separation is undoubtedly artificial and this complex may be considered as a unit. This combined ligament constitutes an arch through which the tendon of the supra- spinatus passes to its insertion into the greater tuberosity. (The deltoid branch of the dorsal scapular nerve passes under this arch also). Posteriorly the joint is covered by the tendons of the supraspinatus, the infraspinatus, and the teres minor. Deep to the first two there are bursae separating the tendons from the joint capsule. The Articulatio acromio-claviculare is a- simple diarthrosis with a stout capsular ligament and an articular cartilaginous disc. The tip of the clavicle articulates with the anterior aspect of the tip of the acromion. The Articulatio Cubiti and Adjacent Deep Structures. It is necessary to remove all the muscles about this joint in order to reveal its struc- tures completely, and as this is done the relation of these to the joint may be re- viewed. Anteriorly the broad tendon of the M. brachialis is found to cross from the lowest point of origin on the anterior aspect of the humerus some distance above the eoronoid fossa to its oblique longitudinal insertion on the antero-medial aspect of the ulna. This insertion is about a centimeter in length and begins at a point just beyond the processus coronoideus. The Articulatio Cubiti includes the humero-ulnar, the radio-humeral and the proximal radio-ulnar articulations. The cap- sula articularis is relatively more restrictive than the capsule of the shoulder joint and presents a number of thickenings and ac- cessory ligaments which both strengthen and limit the movements of which the joint is capable. There are irregular thickenings between the antero-laterai aspect of the humerus running obliquely to the eoronoid process of the ulna. The anterior aspect of the radio-humeral joint is likewise thicken- ed. The latter is exceptionally stout and consists of exceedingly dense fibrous tissue. It is firm.y attached to the ulna at each end of the shallow semi-lunar articular facet. The ug amenta coaaierales uinaris and radialis consist of strong fanshaped flat ligaments extending from the inferior aspects of the medial and lateral condyles respectively to the ulna and annular liga- ment of the radius. The strongest and most prominent ac- cessory ligament is the chorda obliqua. This stout ligament extends between the anterior aspect of the medial apicondyle and the posterior surface of the radial tuberosity proximal to the insertion of the M. biceps. Just beyond its attachment to the epicondyle it senas an oblique and medially directed band which passes over the medial aspect of the base of the eoronoid process to the ulna. These two limit the extent of supina- tion. In any position, except with the arm in extreme medial rotation, complete supi- nation of the wrist is prevented by this band. The Ligamentum annulare radii is a stout 188 Zoological New York Zoological Society [XXIX: 16 thickening of the capsule of the radio-ulnar joint. It surrounds the neck of the radius presenting a strong attachment to the base of the coronoid process passing beneath the chorda obliqua. When the elbow joint is opened anteriorly a very stout, partly intra-articular ligament is found extending fan-like from a broad base in the coronoid fossa. This converges on a narrow rounded band which fuses at right angles with the Ligamentum annulare radii. It also sends a slender band of in- sertion on the lateral aspect of the proces- sus coronoideus of the ulna. The lateral border of this ligament presents a free margin within the joint, but its medial fibers are fused with the part of the capsule over the medial condyle. This ligament has no counterpart in the human being. It definite- ly limits extension and by the fibers coming from the anterior aspect of the medial epicondyle limits supination as well. When the joint is fully opened, a thicken- ed fold of the posterior part of the capsule forms a ligament somewhat like the last described. This has a similar concave free margin within the joint. The fold curves from the lateral epicondyle of the humerus medially and downward over the posterior aspect of the articular surface of the ulna. As it crosses over this to insert in the medial border of the incisura semilunaris, the medial border of it likewise becomes free and this narrow band of fibrous tissue covered by synovial membrane lies between the articular surfaces of the trochlea of the humerus and the incisura semilunaris of the ulna. Between the head of the radius and the lateral condyle of the humerus, laterally there is a pad of fibro-fatty tissue covered by the synovial membrane, constituting a fold or plica which seems to complete the incisura radialis. These folds are present in all of the joints thus far described wherever the irregularities of the bones leave interstices. The insei'tion of the tendon of the M. biceps brachii is on the dorsal aspect of the tuberosity of the radius. When this tendon is reflected medially a bursa between the tendon and the bone is disclosed. With the forearm fully supinated it is seen that the plane of this insertion lies anterior to that of the brachialis of the ulna, although the lines of insertion are parallel and in pronation they lie in the same antero-pos- terior plane. Lateral to the biceps tendon, covering the entire volar surface of the radio-humeral joint and the neck of the radius, is the M. supinator. The medial side of this muscle partially encircles the biceps tendon ; a nar- row bundle of fibers extends downward to insert at the base of the insertion of the chorda obliqua. The principal insertion of the supinator is wrapped about the. lateral volar and dorsal aspect of the upper third of the radius. It meets the insertion of the pronator teres at an acute angle. The Membrana interossea is a stout fibrous sheet extending between the respec- tive Cristae interosseae of the radius and ulna. The proximal attachment is at a higher point on the radius, just below the radial tuberosity and this free border ex- tends downward obliquely to the ulna. The fibers present an interlacing structure be- yond this point. It serves as an additional surface of origin for the deeper volar and dorsal muscles and also limits supination. The Wrist Joint. The wrist is a relatively loose condyloid joint permitting free motion dorso-ventrally and from side to side. Before describing the joint proper, it is convenient to review the terminations of long tendons inserted about the joint. The tendon of the M. flexor carpi radialis may be examined first. This tendon lies on the volar surface of the wrist, entering its own proper sheath and passing through a compartment in the superficial part of the transverse carpal volar ligament. It is re- tained in this region laterally by the un- cinate process of the Os multangulum majus. It may give off a slip here to insert in the central portion of the transverse volar carpal ligament and the palmar fascia. (Note: the palmaris longus may be ab- sent). The remainder of the tendon plunges deeply to terminate about the first, second and third carpo-metacarpal joints. Its in- sertions here are not constant. In some cases there are slight slips to the base of the first metacarpal, but the principal ter- mination is on the base of the second meta- carpal. Functionally, it must be recognized that the rigidity of the second, third, and fourth carpo-metacarpal joints renders this insertion essentially that of a flexor of the carpus and of the central metacarpals with the latter. The tendon of the M. flexor carpi ulnaris inserts on the Os pisiforme which rests on the volar surface of the Os triquetrum. The tendon is continued downward distally to insert on the base of the fifth metacarpal. Here it lies medial to the hamulus ossis hamatum and lies in a groove on the lateral surface of this process, (c.f. infra: Liga- ments of the Os pisiforme). While these flexor tendons insert on op- posite sides of the carpus, the pull of the muscles is not only in the direction of flexion, but also of ulnar deviation of the hand and assists in pronation. The pronated position of the hand is most effective in sustaining flexion of the fingers by aligning the strong ulnar origins with the point of chief pull in the carpal canal. 1944] Robertson: Anatomy of the Woolly Monkey 189 The transverse carpal ligament may be considered as consisting of superficial and deep parts which are fused on the medial and lateral aspects of the carpus to form the flexor tunnel, or as a circular collar 1 resting on the volar aspect of the carpus and held in place by strong attachments to the triquetrum and pisiform medially, the navicular and multangulum majus laterally, and the capitate centrally. It fuses proximal- ly with the capsule of the radio-carpal joint. There is a stout radio-capitate ligament extending from the volar surface of the base of the radius to the volar surface of the capitate bone. When the wrist joint is opened from the dorsal surface a number of important retaining ligaments are re- vealed. The first of these is the first portion of the dorsal carpal ligament which ex- tends obliquely from the radius to the dor- sum of the triquetrum and its continuation ji in the triquetro-lunate ligament which together form the canal through which pass the tendons of the extensor digitorum com- munis. When the radio-carpal joint is opened the adjacent articular surfaces of the navicular and lunate are exposed. The con- i cave articular surface of the radius facing ji them presents about its rim a fibro-cartil- aginous labrum. From the edges of this the articular capsule extends to the carpus. The labrum is continuous with the fibro-cartil- aginous disc between the base of the radius and head of the ulna and with an inter- articular ligament consisting of two parts. The dorsal portion extends from the radio- ulnar articulation to the triquetro-lunate articulation. The volar portion of this in- erarticular ligament is found to consist of the deeper fibers of the radio-capitate liga- ment. The articulatio ulno-carpalis is be- tween the head of the ulna and the trique- trum and pisiform. The ulnar surface is convex, fitting into the depressions of the articular surfaces of the triquetrum and pisiform. This is the reverse of the radio- carpal joint where the concave surface is presented by the radius. The volar ligament extends from the radio-ulnar fibrocartilage to the pisiform and the volar surface of the triquetrum. Further exposure of the joint reveals the volar radio-carpal ligament. This is a fanshaped ligament extending from the entire width of the radius on the volar aspect of the margin of the articular surface converging on the volar aspect of the lunate. On the volar aspect of the pisiform there are three stout ligaments diverging distally as follows: The proximal one extends from | the pisiform to the Os multangulum majus; the intermediate one extends from the pisi- form to the proximal aspect of the base of the hamulus of the Os hamatum; and the distal one extends to the volar surface of the base of the fifth metacarpal, just beyond the carpo-metacarpal joint. When the carpal joints are further ex- posed from the dorsal surface, the hamate and capitatum are found to fit closely to each other and form a rounded proximal surface which articulates with correspon- dingly concave articular facets of the proximal row. The sharp distal edge of the Os centrale forms the lateral lip of this cup and intervenes between the lateral aspect of the Os lunatum and the medial aspect of the Os multangulum minus and the Os multangulum majus which lies some- what volarward with respect to the lesser multangulum. The articulations of the metacarpals and the carpal bones are as follows: The first metacarpal articulates exclusively with the Os multangulum majus. The second artic- ulates with the Os multangulum majus and minus. The third articulates with the Os capitatum. The fourth with a very nar- row facet on the capitatum and a broad area of the Os hamatum. The fifth artic- ulates with the hamatum alone. The Osi centrale is disc shaped and articulates with and rests in a notch on the distal surface of the navicular bone and by its tight interos- seus capsular ligaments forms essentially an integral part of the navicular. When fully isolated the carpus is seen to be curved from side to side with the concavity on the volar surface. The volar rims are formed on the radial side by the Os navicularis proximally and the Os multangulum majus distally and on the ulnar side by the proximal Os pisi- forme, which lies wholly on the volar aspect of the carpus, and the distally situated hamulus of the Os hamatum. On the volar surface the stout triquetro-lunate ligament forms the basis of the transverse carpal ligament. This extends upward to bind the pisiform firmly in this unit and continues laterally to form the naviculo-lunate liga- ment. The distal row of carpal bones is similarly connected by strong volar interosseus liga- ments but these are somewhat less distinct than those of the proximal row. Examina- tion of the bases of the metacarpal bones shows them to articulate very closely with one another, and the interlocking articula- tions between adjacent metacarpals indi- cates the relative rigidity of this area. The sole exception is the first metacarpal which has no articulation with the second, of the kind noted between the others, and lies in a distinctly volar plane with respect to the base of the second. There is, however, a tubercle on the volar aspect of the base of the second metacarpal and into the groove between this and the base proper, the first metacarpal fits. The connection here, as between the bases of the others on both dorsal and volar aspects, is by stout limiting ligaments. The Scapula is a thin, flat bone, trian- 190 Zoologica : New York Zoological Society [XXIX: 16 gular in shape, with a broader thickened inferior axillary margin which forms an acute angle with the medial or vertebral margin. The shorter superior margin joins the vertebral margin almost at right angles, although this posterior superior angle is rounded. The bone is concave on its deep surface which is applied to the posterior thoracic wall. This concave surface presents three ridges running from the vertebral margin and converging on the neck or lateral angle. The dorsal convex surface bears in its upper part the perpendicular Spina scapulae which begins some distance from the vertebral margin, sloping upward to its crest and terminating in the acromion process which is directed anteriorly and superiorly overhanging the glenoid fossa. At the neck the bone is thickened and this lateral angle bears on its summit the shal- low, ovoid articular surface for the humerus. Arising from the superior border, near the neck, is the coracoid process. This is thick at the base and directed anteriorly at an obtuse angle, with respect to the superior border. There is a deep notch in the superior border near the base of the coracoid process. In some specimens the posterior tip of the coracoid process, to which is attached the Ligamentum transversum scapulae superius, may be partially ossified, almost closing the incisura to form a foramen. The lateral aspect of the coracoid process projects laterally and downward to support the strong origins of the caput breve M. biceps and the M. coracobrachialis and M. coraco- humeralis ( coracobrachialis profunda) . The Spina scapulae divides the posterior surface into the deep supraspinous fossa and the more shallow but broader infra- spinous fossa. The inferior margin of the latter is formed by a prominent ridge of bone which in part overhangs the inferior border. The overhang is more definite as it approaches the lateral angle. The crest of the spine presents an S-shaped curve, the posterior limb directed superiorly near the vertebral border and the anterior or lateral limb directed interiorly. In the concavity of the lateral curve is the depres- sion for the origin of the spinous division of the M. deltoideus. The inferior ridge gives a broad area for the origin of the long head of the M. triceps and for the Mm. teres major and minor. This may be described as the Crista infraglenoi- dcdis. The Cavitas glenoidalis is on a flattened ovoid prominence with its plane at right angles to the body of the bone, occupying the lateral angle. The very shallow depres- sion has regular margins and is pointed superiorly, being narrowed anteriorly by a groove in which the tendon of the sub- scapular muscle rests. Above the pointed tip is a facet, the Tuberositas supraglenoidalis for the attachment of the caput longum of the M. bicevs. The clavicle (clavicula) is a short, s'ender, cylindrical bone with an S-shaped curve. It is divisible into a shaft and two extremities. The clavicle occupies an oblique position on the anterior aspect of the thorax, forming an angle of about 45° with the mid-sagittal line. The shaft is bent convexly forward for almost two-thirds of its length and the con- vexity backwards is sharper and more acute than the medial curve. The extremitas ster- nalis is broadened and presents a convexity on its anterior aspect which constitutes the roughened surface for the clavicular at- tachment of the strong sterno-clavicular ligament. The facies articularis sternalis is covered with cartilage and consists of a nar- row border directed medially. This lies on the curve of the spoon-shaped sternal ex- tremity. The extremitas acromialis is some- what broadened and occupies roughly one- fifth of the length of the bone. It is flattened from above downwards and the under sur- face presents a rough concave area for the attachment of the coraco-clavicular liga- ment. The extreme tip of the bone which articulates with the acromion is blunt, and, as noted in the discussion of that joint, occupies a very small and insignificant area. On the convexity of the lateral curve there is a small roughened area to receive the insertion of the M. subclavius. The Humerus is a long, cylindrical bone articulating with the scapula above and the radius and ulna below. It presents a shaft or corpus and two extremities. The corpus is distinctly cylindrical above and flattened antero-posteriorly below. The upper extremity is broadened and supports the caput humeri which is hemi- spherical in shape, the base joining the extremity at an acute angle. The articular surface of the head comprises an area some- what greater than twice that of the glenoid cavity of the scapula. The caput humeri is separated from the shaft by a constricted area, the collum anatomicum humeri. On opposite sides this bears the larger lateral tubercidum majus and the smaller, medially placed tubercidum minus. Between the two there is a deep groove, the sidcus inter- tubercularis. Below, this depression becomes shallower, but in the upper part it is bor- dered by two crests of bone. The crista tuberculi mujoris may project or overhang the sulcus in its upper part. This crest may be traced distally to the lower limit of the deltoid surface, in the middle third of the corpus. The crista tuberculi minofis, which is a broad flattened ridge extending downward from the base of the tuberculum minus, disappears for a short distance to rise again with a sharper crest for the attachment of the Mm. latissimus dor si and teres major. Robertson: Anatomy of the Woolly Monkey 191 1944] The Tuberculum minus in adult specimens 1 presents a broad flat facet on its medially facing apex for the stout insertion of the M. sub scapular is. The Tuberculum majus presents three oval facets on its superior surface near the junction with the collum anatomicum; the two more lateral ones touching and partially surrounding the medial one. These serve for the insertions of the Mm., supraspinatus, infraspinatus , and teres minor from above downward. It should be noted that the crista tuber- culi majoris crowns the summit of an antero- posterior curve of the corpus humeri in its upper third. The apex of this curve, which continues through the greater length of the humerus, lies at the lower point of the in- sertion of the M. deltoideus, in the middle third of the shaft. This is a double curve b01’ng from medial to lateral in the middle third of the humerus as well as antero- posterior throughout. The broad flat area extending downward from the tuberculum majus serves for the insertion of the deltoid. The lower extremity of the humerus, con- sisting largely of the lower third, is flatten- ed from before backward and is much broad- er than the shaft or the upper extremity. From each side there are rough projections of some size, the medial being considerably more prominent than the lateral. These are the respective epicondyles which give at- tachment for the ligaments of the joints and for the extensor and flexor groups of muscles of the forearm. Between these lie the articular surfaces of the inferior ex- tremity. This surface bears an oblique re- lationship to the shaft of the humerus. The medial of these two surfaces is the roller-shaped trochlea which articulates with the incisura semilunaris of the ulna. The lateral surface is the rounded capitulum with its prominent convexity presenting largely anteriorly. The trochlea, on the con- trary, extends around for a considerable area posteriorly. Above the junction of the trochlea and capitulum anteriorly there is a very shallow depression, the fossa coronoidea. Lateral to the latter is the deep rounded fossa radialis. The relative depth and development of these is the reverse of conditions in man. This in turn may be related to the greater par- ticipation of the radius in the elbow joint. Posteriorly, extending from epicondyle to epicondyle, the fossa olecrani occupies the breadth of the inferior extremity. The Ulna is a long, cylindrical bone with a very large, thick, proximal or upper extremity and a long, slender shaft ter- minating in the small lower extremity or head. The upper extremity articulates with the humerus by means of the deep saddle- shaped incisura semilunaris which presents a high proximal crest from the olecranon. and the equally well developed processus coronoideus distally. On the lateral aspect of the latter, facing anteriorly and obliquely laterally, is the shallow depression of the incisura radialis. The coronoid process shows medially a flaring surface which gives added area for the articulation with the trochlea humeri. This flare is definitely more medial than that of the olecranon, which is on a line with the incisura radialis. This latter alignment accounts for the fact that the fossa radialis and the fossa olecrani lie opposite each other on the humerus. Distally at the base of the coronoid proc- ess there is a crest of some size which slopes posteriorly to end on the facies medialis of the ulna. This serves for the insertion of the M. brachialis. The olecranon is a roughened stout process of the upper extremity distinctly separated by an ap- preciable interval from the crest forming the superior lip of the incisura semilunaris. Its proximal surface is irregularly flattened for the insertion of the tendons of the triceps complex. Medially there is a separate facet for the insertion of the M. dorso- epitrochlearis. The upper extremity of the ulna is flattened from side to side and from the coronoid process to the junction with the middle third is three to four times as wide as the lower part of the shaft. This region is constricted also between the an- terior border and the posterior subcutaneous border. The shaft of the ulna presents a distinct curve with the convexity directed posterior- ly. In the lower third, the distal extremity becomes cylindrical and exhibits a promi- nent crest for the origin of the M. pronator quadratus. Below this the bone widens somewhat and terminates in two projections of about equal size, the proximal capitulum which articulates by a synostosis with the radius, and the longer distal processus styloideus which articulates directly with the concavity of the Os triquetrum and to a slight extent with the Os pisiforme. The foramen nutricium is found proximally on the medial or lateral surface of the flattened area some distance beyond the coronoid process. It is directed proximally. The Radius is a long, slender, cylindrical bone with a prominent well-developed head bearing a shallow circular depression, the fovea capituli radii, for its articulation with capitulum humeri. Medially there is a semilunar convex area for the incisura radialis of the ulna — this is the circum- ferentia articularis. Just below the capit- ulum the bone is narrowed to form the collum radii and close below this is a round- ed projection directed medially, the pos- terior rim of which serves for the insertion of the flattened tendon of the M. biceps. The anterior bulge of this tuberositas radii serves as a fulcrum increasing the mechan- 192 Zoologica: A lew York Zoological Society ical efficiency of this insertion. Distal to the tuberosity is the beginning of a laterally directed curve in the shaft which has its convexity in the middle third. To the sum- mit of this curved area the M. pronator teres has its broad insertion. The volar surface of the radius is obliquely flattened, sloping posteriorly to the interosseous crest. The broad lower extremity is flattened in a dorso-ventral plane; the anterior surface serving for the insertion of the M. pronator quadratics. The articular surface of the lower extremity faces toward the medial or ulnar side of the arm and bears thus an oblique relation to the shaft of the bone. The processus styloideus is a flat projection constituting the lateral border of this ar- ticular surface. Medially there is a shallow depression for articulation with the ulna, but as described in the section on the wrist joint this articulation is not by true articu- lar surfaces. Dorsally the distal extremity presents a prominent crest which is broad at its base and separates two deep grooves. The medial one is for the tendons of the M. extensor digitorum communis. The lateral groove serves to retain the strong tendons of the M. abductor carpi radialis and the Mm. extensores carpi radialis brevis and longus. The carpal bones will not be considered individually; they have been discussed in connection with the wrist joint. The point of chief interest in the phalan- ges is the striking curve of the proximal phalanges of digits two, three and four, especially. This curve has its convexity dor- sally and the volar surface presents a deep groove into which the flexor tendons are tightly bound by the annular ligaments. It is suggested that this great curve increases the mechanical efficiency of the fingers in forming a rigid hook once flexion of the metacarpo-phalangeal joints passes beyond the 90° angle. Bibliography. Authorities consulted but not in each case quoted. 1. Ogilby, W. Proc. Zool. Soc. Lond., 1836, 25-28. 2. Bychowska, M., Folia Morph. Warszawa, 1930, 2, 60-121. (Abst. in French, 116-121). 3. Miller, R. A., Am. J. Phys. Anthrop., 1932, 17, 1-56. 4. Whipple, L L., Z. Morph. Anthrop., 1904, 7, 261-368. 5. Pocock, R. I., Proc. Zool. Soc. Lond., 1920, (1) 91-113; 1925 (1), 27-47. 6. Ashley-Montagu, F. M., Am. J. Phys. Anthrop., 1931, 15, 291-314. 7. Hartman & Strauss, The Anatomy of the Rhesus Monkey, The Williams and Wilkins Co., 1933. EXPLANATION OF THE PLATES. Plate I. Fig. 1. Lagothrix. Dissection of shoulder re- gion. Plate II. Fig. 2. Lagothrix. Medial aspect of arm and volar surface of forearm. Plate III. Fig. 3. Lagothrix. Flexor tendons in carpal canal. Fig. 4. Lagothrix. Extensor tendons. Plate IV. Fig. 5. Lagothrix. Lateral view of metacarpo- phalangeal joint. Fig. 6. Lagothrix. Superficial muscles of the palm. Fig. 7. Lagothrix. Extensor tendons and their insertions on the phalanges. Fig. 8. Lagothrix. Flexor tendons and their insertions on the phalanges. Fig. 9. Lagothrix. M. flexor digitorum pro- fundus tendons. Volar aspect. Plate V. Fig. 10. Lagothrix. Deep muscles of the palm (volar interossei, intrinsic muscles of thumb, index and fifth digit) . Note: The relation of the flexor carpi radialis tendon to its insertion is not clearly shown here, x and y are the os navicularis and os multangulum majus to which the sheath of this ten- don is bound. It does not insert on them. ROBERTSON. PLATE I. M. deltoideu: Common origin: M. coracobrachialis & M. biceps brachii capul breve Clavicula Common insertion: MM. pectoralis minor & abdominis N. radio I is A. subclovia V. subclovia N. axillaris f Insertion a*. M. pectoralis major M. biceps brachii, caput longum musculo cutaneus 'v- ; •/> 'N. ulnaris N. medianus A. brachialis V. brachiali M. dorso-epitrochlearis Ms. pectoralis abdominis ANATOMY OF THE SOUTH AMERICAN WOOLLY MONKEY (LAGOTHRIX). PARTI. THE FORELIMB. 0 RTSON. PLATE II. M. flexor carpi ulnaris A. brachialis ^ A. ulnaris brachialis Tpndo m. biceps brochii — licertus fibrosus brachioradialis A. radialis N. radialis M. pronator teres insertion M. flexor digitorum sublimis M. flexor digitorum profundus N. medianus A. collateralis ulnaris, N. ulnaris _ M. pronator teres M. flexor carpi radialis M. palmaris longus N \ FIG. 2. ANATOMY OF THE SOUTH AMERICAN WOOLLY MONKEY (LAGOTHRIX). PARTI. THE FORELIMB. ROBERTSON. E PLATE III. i j ANATOMY OF THE SOUTH AMERICAN WOOLLY MONKEY (LAGOTHRIX). PARTI. THE FORELIMB. ROBERTSON. PLATE IV. /' M. interosseus (volar) FIG. 5. Lig. capitulorum transversa Vagina and tendines MM. flexores digitorum sublimis et profundus M. lumbricale I Tendo m. flexor Jj/fl digitorum sublimis Lig. annulares (interarticular) Lig. annulare-*’ (articular) Insertion tendon M. flex, dig. profundus M. adductor d i g i t i q u i n t i "Ramus profundus N\ uinaris Volar carpal ligament **Os pisiforme • N. uinaris FIG. 7. Radial aivision FIG. 8. i Ulnar division • J> 9 / FIG. 6. FIG. 9. ANATOMY OF THE SOUTH AMERICAN WOOLLY MONKEY (LAGOTHRIX). PARTI. THE FORELIMB. \ M. adductor indicis proprius ROBERTSON. PLATE V. FIG. 10 \ M. flexor digiti quinti brevis Tendo m. adductor poll \ M. interossei dorsalis 0) M. abductor p o II i c i s M. flexor pollicis brevis M. abductor indicis MM. interossei volares x carpal canal Lig. radio-carpeum volore ANATOMY OF THE SOUTH AMERICAN WOOLLY MONKEY (LAGOTHRIX). PARTI. THE FORELIMB. Funkhouser : Some Venezuelan Membracidae 193 17. Some Venezuelan Membracidae.1 W. D. Funkhouser University [This is a contribution from the Forty-third or Venezuelan Expedition of the Department of Tropical Research of the New York Zoological Society made under the direction of Dr. William Beebe. The expedition was sponsored by grants from the Committee for Inter-American Ar- tistic and Intellectual Relations and from four trustees of the Zoological Society, George C. Clark, Childs Frick, Laurance S. Rockefeller and Herbert L. Satterlee, and by invaluable assistance from the Standard Oil Companies of New Jersey and Venezuela.] Through the courtesy of Dr. William Beebe, the author has been permitted to examine a small collection of Membracidae from Caripito, Venezuela. The species rep- resented in this collection should be re- corded in the literature of the family since fourteen of the eighteen species have never before been reported from that country. The records of Venezuelan Membracidae are very meager and Dr. Beebe’s material is valuable in that it furnishes additional in- formation regarding the geographical dis- tribution of the membracids. The species represented in the collection, together with brief statements of their known general distribution, are as follows: Membracis foliata Linnaeus. Described from Surinam. Common throughout Central America and north- ern South America. Membracis c-album Fairmaire. May be a variety of M. foliata. De- scribed from Brazil. Has about the same range as the preceding. Membracis lefebvrai Fairmaire. Described from French Guiana. Has been reported from Brazil and Mexico but from no other countries. This is the first record from Venezuela. Membracis arcuata DeGeer. Rather common throughout northern South America but this is the first ac- tual record from Venezuela. 1 Contribution No. 698, Department of Tropical Research, New York Zoological Society. / Kentucky. Membracis tectigera Olivier. One of the most abundant species of the genus in tropical America. Is found from Brazil to Mexico. Enchophyllum quinquemaculatum Fair- maire. A rather rare insect which has previ- ously been reported only from Brazil. This is the first record from Venezuela. Enchenopa albidorsa Fairmaire. Abundant in Brazil, Colombia, British Guiana and occasionally found in Ar- gentina but not previously known from Venezuela. Enchenopa lanceolata Stoll. Very common over most of Central and South America but not previously re- corded from Venezuela. Enchenopa serratipes Buckton. Formerly considered a variety of E. albidorsa but now known to be distinct. A new record of Venezuela. Enchenopa ignidorsum Walker. Previously known from Mexico, Pana- ma, Ecuador and Dutch Guiana. Appar- ently not common anywhere. Another new distribution record. Campylenchia hastata Fabricius. Described from Peru with a general distribution from Brazil to Mexico. Very common in the Canal Zone area but not previously recorded in the liter- ature from Venezuela. Spongophorus guerini Fairmaire. Another new record from Venezuela but its presence in this country is to be expected since it is quite common in most of northern South America. Heteronotus vulnerans Germar. A very rare species of the subfamily Darninae which has previously been known only from Brazil and is seldom 194 Zoological New York Zoological Society [XXIX: 17: 1944] seen in collections. Dr. Beebe has two specimens. A new record for Venezuela. Heteronotus strigosa Butler. Another rather uncommon form also recorded in the literature of the family only from Brazil. Another new distribu- tion record. Darnis latior Fowler. Described from Panama and reported from British Guiana by Haviland and Funkhouser. This is the first record from Venezuela. Ceresa vitulus minor Fowler. Very common over most of northern South America and extends up through Central America to Mexico and south- ern United States. Cyphonia fuscata Buckton. Described from Brazil and known to inhabit Peru and Argentina. This is the first record from Venezuela. Stegaspis laevipennis Fairmaire. Known in the literature from Brazil, British Guiana, Peru and Colombia. Not previously known from Venezuela. Beebe: Iguanidae of British Guiana and Venezuela 195 Field Notes on the Lizards of Kartabo, British Guiana, and Caripito, Venezuela. Part 2. Iguanidae.1 William Beeee. Director, Department of Tropical Research, New York Zoological Society. (Plates I-VI ; Text-figures 1-17). [This contribution is a result of various ex- peditions of the Department of Tropical Re- search of the Dew xork Zoological Society to British Guiana and to Venezuela, all made under the direction of Dr. William Beebe. The Guiana expeditions were made during the years 1909, 1916, 191 <, 1919, 1920, 1921, 1922, 1924 anti 19^.6, and the Venezuelan trips in 1908 and 1942. The latter was sponsored by grants from the committee for Inter-American Artistic and Intellectual Relations and from four trustees of the Zoological Society, George C. Clark, Childs Frick, Laurance S. Rockefeller and Her- bert L Satterlee, and by invaluable assistance from the Standard Oil Companies of New Jer- sey and Venezuela.] Contents. Page Family Iguanidae 196 Ecology of the Genera 196 Anolis aeneus 196 Anolis biporcatus 197 Anolis chrysolepis 197 Anolis fusco-auratus 199 Anolis nitens 200 Anolis punctatus 200 Anolis sagrei 201 Anolis schiedii 201 Iguana iguana iguana 201 Plica plica 204 Pb'ca umbra 207 Polichrus marmoratus marmoratus 209 Tropidurus torquatus hispidus 212 Uranoscodon super ciliosa 213 Urocentron azureum 215 References 215 Introduction. This is a second caper on the lizards of Kartabo, British Guiana, and Caripito, Ven- ezuela.2 In the year 1909, and from 1916 to 1928, eight expeditions were sent out from this department to British Guiana, and in 1908 and again in 1942 field work was carried on in Venezuela. Throughout the course of these expedi- tions many field notes, color plates and photographs were made of tropical verte- brates, and the object of this present series 1 Contribution No. 699, Department of Tropical Research, New York Zoological Society. 2 Part 1, Gekkonidae, Zoologica, Vol. 29: pp. 145-160. is to assemble and publish some of these notes and illustrative material. Any change or alteration of the original notes is piaced between brackets. The chief value of these data is that they are concerned with living or freshly killed specimens. The observations in Guiana were made in one-quarter of a square mile of jungle at Kartabo, and those in Venezuela at or close to Caripito, which is only 528 kil- ometers northwest of Kartabo. In addition to numerous technical papers in Zoologica and several popular volumes, there have been published the following general ecological summaries: Zoologica: (Kartabo) Vol. II, No. 7, 1919, pp. 205-227; Vol. VI, No. 1, 1925, pp. 1-193. (Caripito) Vol. XXVIII, No. 9, 1943, pp. 53-59. Also see “Tropical Wild Life in British Guiana” by Beebe, Hartley and Howes, published by the New York Zoological Society, 1917, pp. 1-504. My hearty thanks go to Dr. Charles M. Bogert of the American Museum of Natural History and Dr. Karl P. Schmidt of the Chicago Natural History Museum for iden- tifications and for bringing up to date my often out-worn names of many years ago. My original field numbers and other data have been appended to descriptions, breed- ing and other notes. These specimens are either in the collections of the Department of Tropical Research or in those of the American Museum. Whenever the term total length is used, a perfect, unregener- ated tail is understood. Many of the figures in the plates are black and white reproduc- tions of original colored paintings, so only the pattern is preserved. The following are from paintings by Isabel Cooper, Plate- figures 1, 2, 3, 7, 9, 10, 17 and 18. Text- figures 3, 7, 11, 12 and 16. From drawings by George Swanson. Plate-figures 11 and 19. Text-figures 1, 4, 5, 8, 9, 13, 15 and 17. From photographs by T. V. Smolucha, Plate- figures 4, 5, 6, 8, 14, 15 and 16. 196 Zoologica: Neiv York Zoological Society [XXIX: 18 Family Iguanidae. Ecology of the Genera : Seven genera and fifteen species of this family of lizards were found at Kartabo, British Guiana, only four of which were recorded from Cari- pito, Venezuela. A longer period of field work at the latter locality would doubtless reveal many of the others. Eight of the fifteen species belong to the genus Anolis, which is not surprising when we recall that about one hundred species of these little lizards range from North Carolina to Brazil. Only two other Anolis, lentiginosus and longicrus, have been recorded from “The Guianas,’’ neither of which we encountered. The Iguana we found is the only species of its genus, ranging in several subspecies over Central America, the West Indies and the northern half of South America. Plica is represented by two out of the four known species, the two remaining forms inhabit- ing Bolivia and Argentina respectively. Polychrus is present in one out of the two species of the genus. A single species of Tropidurus was found out of the twenty- odd known from elsewhere. Uranoscodon is a monospecific genus, confined to north- eastern South America. Out of six known species, one species of Urocentron was col- lected. This had hitherto been recorded only from Brazil. Occurrence : In order of relative numbers as observed by 11s, we have : abundant, Ano- lis; common, Iguana, Plica and Polychrus; occasional, Uranoscodon; rare, Tropidurus (4) and Urocentron (2). Size: From small to large; Urocentron (average 100 mm.), Tropidurus (150 mm.), Anolis (175 mm.), Plica (310 mm.), Urano- scodon (360 mm.), Polychrus (450 mm.), and Iguana (up to 1200 mm.). Food : Iguana appears to be strictly vege- tarian, but all the other lizards feed on in- sects and other invertebrates. Hints of can- nibalism are observed rarely, and a small but definite amount of vegetable matter is taken by Polychrus. Eggs : The number of eggs recorded is as follows: Anolis 2 (which may be deposited simultaneously or a week apart) ; Plica 2; Iguana 4 to 7 ; Polychrus 7 to 8 ; Uranosco- don 6 to 11. Habitat : All fifteen species, with the ex- ception of Urocentron, are essentially ar- boreal. Anolis and Iguana are found occa- sionally on the jungle floor, and the latter breeds in subterranean burrows. Among anolids fairly distinct haunts may be dif- ferentiated. Anolis aeneus and fusco-aura- tus are at home on low and rather small tree trunks, the latter preferring those with pale, lichen-covered bark. Anolis biporcatus, chrysolepis and nitens by preference choose the foliage of low jungle growths, and Ano- lis punctatus is found usually on the ground. Escape Methods: Even the smallest ano- lids, when once seized, will bite ferociously but futilely, as their teeth are too minute to puncture human skin. The bite of a large iguana is very different, as the powerful muscles and sharp teeth can make an ugly wound. All iguanid lizards, however, strive to escape observation by immobility, plus con- fusing postures and more or less swift change of pattern and coloring. Anolis ni- tens trusts so completely to its protective pattern that it can usually be picked up in the hand. Plica rests upsidedown on self-colored tree-trunks, and Polychrus assumes strained positions and holds them for hours, at the last moment rushing off headlong and again assuming a posture which allies it more to twigs and tendrils than to a living saurian. In Uranoscodon this trust is carried to an extreme, and not only can these lizards be picked from their perch on a tree, but car- ried home in what appears to be a trance- like condition. Once awakened from this, the phenomenon may reoccur from time to time in captivity. Iguanas feed usually high up in trees, and their favorite method of escape is to fling themselves headlong into water far beneath, or to flatten out and crash unharmed into thick foliage near the ground. Competitors: As far as insect food is con- cerned, the only vertebrate competitors of the iguanid lizards are toads, brevicipitids and ant-thrushes on the ground, tree-frogs and low living jungle birds in low foliage. The tree-trunks are shared with dendroco- laptine birds. Enemies : Snakes are the source of great- est danger to these lizards, with jungle hawks a close second. Birds belonging to the genus Attila of the jungle cotingas feed ex- tensively on small lizards, as do some of the terrestrial ant-thrushes. At Caripito, Mr. Swanson of my staff saw a good-sized jaguar in full pursuit of a large iguana, crossing the trail only a few yards away. Anolis aeneus Gray, 1840. Name : Marbled Anolis. Range: British Guiana, Venezuela, Trini- dad and Tobago. General Account: More than any other anolid, this species is semi-gregarious, oc- curring in limited areas of the jungle in relatively large numbers. I have found eleven in a space of ten square yards and no others anywhere nearby, although there was no apparent change of plant growth or of elevation. The marbled anolis is a tree- trunk lizard like Plica, rather than a fre- quenter of low, leafy bushes. I have seen and shot them at a height of thirty-five feet on the trunk of a mora. In general, the head and neck are grayish. Body, limbs and tail buffy brown, indistinct- 1944] Beebe: Iguanidae of British Guiana and Venezuela 197 ly mottled, banded and striped with dai'ker. Beneath, head, throat, legs and terminal two-thirds of tail whitish; body and basal third of tail pale lemon yellow. Two males displaying before females well up on the trunks of trees exhibited unusual- ly extensive dewlaps. They began half way between the eye and tip of the snout and extended down the ventral surface half-way between the insertions of the fore and hind limbs. The vertical breadth was large in proportion. There is no symmetry of dorsal markings, the pattern being of irregular blotches and reticulations, and I have noticed very little change in day and night alteration of color. The tail is exceedingly deciduous and it is difficult to secure a perfect specimen. In two individuals shot from trees, the tail was freshly lost in one and semi-detached in the other, not from a shot injury. It is probable that this occurred from the shock of striking leaves or foliage in the fall of the anolis to the ground. I have never seen any hint of the lizard being able, by its own exertion, to discard its tail. Locally it is fairly common at Kartabo but 1 did not take it at Caripito. A blue-tinged individual was captured at the pitch lake at Guanoco, a few miles down river from Caripito. In my journal I find references to de- tailed observations on this species but the notes themselves have disappeared. Food: A specimen taken March 10, 1919, had its stomach filled with more than sev- enty-five mosquitoes, and another anolis had devoured fifteen termites. Anolis biporcatus (Wiegmann, 1834). Name: Brown-backed Anolis. Range: “Western Ecuador, northward in- to Central America,” British Guiana. General Account: Only two specimens were taken at Kartabo. Color Plate 1171 is of Specimen 204, a female, total length 160 mm., April 22, 1919. This individual is mottled with several shades of gray, and from the nape back to half-way down the tail extends a wide, vertebral band of light olive-buff. The dewlap of the second speci- men, a male, when distended showed no heightened color, just a grayish-white ex- panse of skin. Both geckos were taken in one sweep of a butterfly net as they rested on the leaves of a low bush in the jungle. These were identified by Dr. Noble; one specimen given to the American Museum, the other to the museum at Georgetown, British Guiana. Anolis chrysolepis Dumeril and Bibron, 1837. (Plate I, Figs. 1 and 4). Names: Vertebral-striped Anolis, Gold- striped Anolis, Large Jungle Anolis. Range: Venezuela, Trinidad, the Guianas and Brazil. General Account: This is the second in abundance of Kartabo and Caripito anolids, the first being Anolis nitens. Their favorite haunts are the leaves of low shrubs and bushes in the jungle and along the streams. Female breeding. No. 2836, Kartabo, May 14, 1924, Color Plates 737, 738: Measurements : Total length 136 mm. Fol- lowing are percentages of length : Head 10.3 per cent., head and body 36, tail 64, head width 5, fore leg 13, hind leg 27, eye diam- eter 1.4 per cent. Weight 1.5 grams. Color in Life: General color above olive with a strong greenish-yellow tinge. A dou- ble, broken, backward pointing, V-shaped, dark brown mark between the eyes. Starting the same distance from the pineal scale as its distance behind the point of the ocular V-mark is a broad dorsal stripe of yellow ochre, overlaid with two narrow stripes, one on each side of a light golden. There is a fine line of the yellow ochre outside of these gold stripes on each side, then a broad edg- ing of very dark brown, black on the inner side and fraying out to an irregular scatter- ing of sepia freckles. Sides of head and body and upper parts of limbs are faintly spotted with red brown. Laterally and along sides of neck and chin are strong mottlings of lav- ender which die out to purest white on the chin, small dewlap and ventral surface. This latter is immaculate except for scattered specks of very light brown posteriorly and for a dull brown tinge to the under surfaces of the limbs. Tail is faintly barred with nar- row rings of brown, becoming mottled half way down and finally black on the last quar- ter inch. Pupil ring gold; iris orange with faint dark mottlings. Male not breeding: Above brownish-black, with wide median stripe of burnt orange variegated with gray. A broad band across top of head between eyes. Below grayish- white, with scarlet dewlap. Iris dark brown with inner ring of gold. Male not breeding: Back dark olive buff shading downward to wood brown and pink- ish-white on ventral surface. Dorsal stripe pale vinaceous bordered with natal brown. Head mottled with various shades of brown with broad eye band of olive. Face sandy shading to wood brown on snout. Irregular light markings near eye and one wide golden band from eye down and back to jaw. Legs and tail buffv brown with mottlings of darker. Female breeding: Black inter-orbital stripe edged with lighter. Vertebral stripe orange. Male not breeding: Mottled grayish-olive above with broad vertebral stripe of yellow buff, bordered with black. Mottled white be- low. At death darkens, the back stripe be- 198 Zoologica : New York Zoological Society [XXIX: 18 Text-fig. 1. Anolis chrysolepis. Hyoid. coming bright gold anteriorly, reddish- brown on tail. Female breeding: General color smoke gray with fine black stippling. Snout lighter. Top of orbits bright yenow gray, with hour- glass transverse marking of dusky. Dorsal stripe olive buff, lighter on lower back and tail. This stripe narrows over the shoulders and widens on lower back. Sides of face buff mottled with brown. Limbs smoke gray mot- tled with black. Ventral surface vinaceous with dark mottling under limbs and tail. Iris apricot yellow with fine freckling. These six descriptions will give an idea of the infinite variation. Often an anoiiu will change so radically that a description can- not be finished, some characters vanishing or being complicated by others. The two main differentiations into dorsal striped and no dorsal striped is regardless of age or sex. In the preserved specimens, flank markings and the eye band often remain the most dis- tinct. In a series of delicate measurements it was found that within a week, the change in total length from the living or just dead to the same specimen preserved, was from 1 to 3 mm. Hyoid: Coll. No. 30,083, Caripito, April 10, 1042, adult male, total length 210 mm. Text-fig. 1. The glossohyal is an extremely attenuated, needle-like element, equal in diameter throughout, extending forward 3.5 mm. from the basihyals. The latter are rather unlike their usual arch-like shape, due to amalgamation with the hypohyals. These latter are distinguishable as short, thick, forward-directed shoulders, complete- ly ossified with the basihyals. From the anterior tips of these shoulders arise the ceratohyals, externally and with no evidence of a socket, as slender as the glossohyals, 8.5 mm. in length and directed at an acute backward angle. Short, transparent, distal tips represent the remains of epihyals. From the thickened, posterior shoulders of the basihypohyals spring the first cerato- branchials, their enlarged, rounded proximal caps fitting nicely into deep, almost quarter- circle, concave sockets. These elements con- sist of dense, osseous-like tissue and their decided curve in another plane gives, from the dorsal aspect, a false impression of a rather sharp angle at mid-length. They are the heaviest elements in the entire hyo'idean complex, and support the usual short, distal vestiges of first epibranchials. The second epibranchials surpass in development even those in Polychrus, being twenty-five per cent, longer than the ceratohyals. They arise as posterior, ankylosed continuations of the basihyals, extending straight back- ward 12 mm., decreasing slightly in caliber. 1944] Beebe: Iguanidae of British Guiana and Venezuela 199 Food : The food of the gold-striped anolis did not show much variation. In order of abundance eaten, the five major ingredients were termites, spiders, ants, roaches and grasshopper nymphs. Breeding : As marked by courtship and the deposition of eggs the following months were noted in my journal: March (2 times), April (1), May (3), June (7) and August (2 times). An average egg measures 7 by 13.5 mm., the largest one 8 by 16 mm. The average weight was 2.8 grams. Dissection showed that 2 eggs might be laid simultaneously, but usually the second was several days to a week behind the first in developmental stage. The shell of the egg was rough, with numerous longitudinal striations or fur- rows. If not kept away from nest debris the egg became densely coated with whatever it touched within the first half hour after being laid. Anolis fusco-auratus D’Orbigny, 1837. Names : Plain Jungle Anolis, Lichen-col- ored Anolis, Tree Anolis, Clay-colored Ano- lis, Snub-nosed Anolis. Range : Northern South America. General Account : Next to Anolis chryso- lepis and nitens this is the most abundant anolid at Kartabo. It is decidedly a tree, not a bush or ground, lizard and is almost al- ways found on pale bark or that encrusted with light-hued lichens. This choice of habi- tat is very pronounced and observable day after day. The immediate reaction toward an intruder is for the lizard to scurry up the trunk, so it was often necessary to shoot the reptiles with fine shot in order to obtain them. I never saw one on the ground and only once away from a good-sized tree. Specimens No. 698 and 699 were captured among the topmost foliage of a ninety-foot tree within fifteen minutes after it had been cut down. The tree fell in an open secondary clearing, so it may be assumed that this pair of arboreal anolids was living at a great height. Color in Life : In addition to emotional and illumination color changes there are general types of patterning, one pale buffy with many irregular elongate spots and blotches, the other darker with a series of small, dark, mid-dorsal, roundish spots. Fe- ma.es usually show the latter tendency, but I have seen males in full breeding condition wi h both patterns well developed. Also the dewlap may be dominantly yellow or with much of the central area bright scarlet. Adult male, pale phase, No. 2823, total length 130 mm., Color Plates 728, 729, 730, May 8, 1924 : General body color oale smoke gray. Head and neck blotched with lightish gray (the only color in Ridgway that ap- proximates this is light mineral gray). Body has scattered, broken, lateral streakings of red brown and a broken, very vague line across the orbits of hair brown. Eyelids light vinaceous, limbs tinged with russet. Dewlap amber yellow, lighter toward median line; scales of lower neck becoming russet vinaceous as they spread out over the dew- lap. Ventral surface dull yellowish-white, pinkish under arms, immaculate except for the chin region which is marbled with light brownish. The body color shades to white on the posterior two-thirds of tail, which has ten wide bands of deep brown, separated by twice their length of ground color, and be- coming blacker toward tail tip. Pupil rim brilliant gold, iris orange, decked with gold. The dark phase of the same individual has the whole body clouded with sepia, the limbs darkening to uniform brown. Breeding male, dark phase, No. 248, total length 145 mm.. Color Plate 285, December 31, 1920: General color olive buff, with dull red brown head markings and broad tail bands. Fourteen small, irregularly rounded dark spots down the back from nape to be- yond posteriorly adpressed knee. Ventral surfaces whitish. Dewlap pale chalcedony yellow at top, white front and back with large rounded patch of bright coral pink in center, the lines of scales over its expanded surface conspicuously white. The dewlap is large, extending from the vertical of ante- rior orbit, back to the line of adpressed elbow. Male, not breeding, dark phase, No. 542, total length 90 mm., April 18, 1922: General color buffy olive, sides of face, neck and body olive, faintly mottled with cream buff. Chin and throat pallid with sparse brown stippling, moderately developed dewlap white anteriorly, in general coral pink with tinge of cream in center. Median gular scales of dewlap white. Iris brilliant gold mottled with darker. Adult male and female, Nos. 698 and 699, total lengths 134 and 135 mm., weights male 1.1, female 1.6 grams: Caught in fallen tree-top foliage, August 9, 1922. Both in dark phase, ground co'or grayish-olive, darker mottlings along mid-back, others on head and snout. Limbs and feet fuscous. Tail marked alternately with fuscous and olive. Extreme body changes range from dark gray to dull green. Throughout these changes the male is paler, especially on head and tail. This makes the dorsal marks and the caudal bands much more distinct. The most radical difference between the sexes is the dewlap, which is almost absent in the female, represented by only a self-colored, slight throat fold. In the male it is relatively enormous, 20 mm. long by 10 high, extend- ing from below the anterior level of the eye to the mid-belly. Elsewhere the male is uni- form in ventral coloring, white on chin and throat, buffy white on body and base of tail. 200 Zoologica: New York Zoological Society [XXIJX: 18 The female is white below, much marbled with dark brown along the line of the jaw and down the center of the body. Breeding male, No. 2864, total length 117 mm., May 16, 1924: This anolis never lost its strong tinge of yellow green, evenly dis- tributed from snout to tail tip. Food: No. 248: 3 termites, 2 small grass- hoppers; No. 542: 3 termites, 6 flies. Nos. 698 and 699 : 14 termites, 7 flies, 2 stingless bees, 3 ants, 1 mantid and 4 salticid spiders. Breeding : Breeding males were found on May 10 and 16, August 9 and December 31. A female taken on May 12 contained an egg about to be laid, measuring 5 by 10 mm., and another egg with well developed yolk which would have been ready in ten more days. Anolis nitens (Wagler, 1830). (Plate I, Figs. 3 and 4). Names : Oblique-striped Anolis. Ho-ko- bee, Ey-a-moo-ru (Akawai Indian). Range: Northeastern South America. General Account : The most abundant ano- lis both at Kartabo and Caripito also ex- hibits the greatest variation in coloration but not in pattern. It is found in the jungle both on tree trunks and on low bushes, but usually on the latter. Two general types of coloration are found, the lizards with dominant brown and honey yellow markings, and those with black, gray and white. The most persistent character is the pink rump which is almost invariably present both in the lightest and darkest phases and in both sexes. It vanishes very soon after death. The most characteristic pattern is a series of posteriorly pointed V-shaped markings on the back from shoulder to beyond the base of the tail, accounting for my long- used field name of oblique-striped anolid. These alternating V’s may be black, buff, honey yellow, gray or brown. The head is variously striped and banded with contrast- ing colors, and always with a large orange or white stripe from eye to ear, and alter- nating bands along the edge of the lower jaw. The limbs and tail are banded with more or less regularity and the under parts are pale yellow, light gray or white. The dewlap, when well developed, is invariably scarlet, with a whitish base and dark, widely spaced scales over the expanded surface. There is little distinction in the relative size of the sexes. In 12 breeding males, all with perfect tails, and with extreme total lengths of 166 to 209 mm., the average was 186 mm. In six breeding females, measuring from 165 to 198, the average was 180 mm. This anolis has more confidence in its pro- tective coloring than any other species and it is not difficult to pick an individual from leaf or trunk. The hand must be moved very slowly, and a more certain method is- to wriggle the fingers of one hand in front of the lizard while the other approaches from above or behind. In neither case is there any chance of not being observed. The bright eyes of the anolis watch keenly, but in the divided interest the capture is more certain. It is remarkable how the lizard will watch the fingers actually close on its body without moving. The least jerk or quick motion will send the creature away in a lightning dart. Food: The food of fifteen individual ano- lids was as follows: termites (occurring 12 times), grasshopper nymphs (10 times, all in April), beetles (5), ants (4), roaches (4), caterpillai's (2), spiders (2), the ano- lid’s own skin (2), mole cricket (1). Breeding : I find records for only a portion of the year at Kartabo, for breeding of this species. This, of course, is only a part of the whole picture. April (13 records), May (9 l, June (5 records), July (2) and August (2 records) . An average egg measures 8 by 16 mm., and is rough in texture with numerous long- itudinal striations. Usually a single one is deposited but in one female I found two eggs almost ready for laying, measuring respectively 13 and 14 mm. in length. A full-grown female kept in a vivarium ate 114 termite workers in 11 days, and probably many more when I was not watch- ing. When a choice of other food was offered, the termites were always taken first. After a two-day fast, the lizard ate a yellow and black hemipteron which would be called warningly colored. On five succeeding days eight other insects of the same species were refused on sight without any attempt at tasting. After another two days’ fast she refused a very small Dendrobates trivittatus and when the juices of one of these frogs was rubbed on a grasshopper nymph the anolis rejected the grasshopper after taking it in her mouth. Anolis puncfatus Daudin, 1802. Names: Blue-spotted Green Anolis, Large Pug-nosed Anolis. Range: Northern South America. General Account: This species seems rather rare, only three specimens being seen and taken at Kartabo and none at Caripito. One is No. 3129 (21293 in American Mu- seum), Kartabo, 1919, total length 195 mm. The second, No. 3269, male, total length 250 mm., September 1, 1920. Both of these lizards were taken on leaves on the ground in the jungle. Male adult, No. 3024, Kartabo, August 30, 1920, Color Plate 219: Measurements: Length 220 mm., head 11 per cent, of length, body 39, tail 50, fore leg 15, hind leg 24 per cent., weight 7.4 grams. Color in Life: This male is essentially 1944] Beebe: Iguanidae of British Guiana and Venezuela 201 dark yellow green including all upper parts, head, body, limbs and toes, with a sparse scattering of small, irregularly rounded spots of pale turquoise blue, largest on the 1 sides of body. The middle third of the tail is brilliant neva green, and the terminal third Vandyke brown. Dorsal head scales are touched with bright chalcedony yellow and orange. Sides of head anteriorly green, scales below the eye shaded with primuline yellow; eyelids russet vinaceous with a dou- ble row of yellow scales around the eye. The j dewlap is very large, beginning at the ver- tical in front of the eye and reaching back almost to mid-abdomen. Its depth is consid- erably more than from the posterier eye to tip of snout. It is distinctly but not bril- liantly patterned and colored cadmium yel- low, and when widely distended 10 or 12 rows of separate scales show as curved lines of white dashes, those in the central rows flanked with black at the bases. Color Change and Habits: The blue-spot- ted anolis has considerable power of color change, and both dorsal and ventral greens, under stress of emotional excitement, be- come varying shades of russet and terra- cotta. When chloroformed it becomes in general brown with the pale spots turtle green, and below dull brownish. The dewlap remains unchanged except that it darkens in general. An hour after death, in two in- stances, the normal green colors of life re- turned until placed in preservative. In pre- servative this anolis turns black or dark , wine color, all trace of the green and blue disappearing. Unlike the more common anolids this seems unadapted to captivity and refuses to i feed readily. In the first excrement pellet were the remains of two winged queens of Cryptoceris atratus. It was seen to feed on j termites. Anolis sagrei Cocteau, 1837. Range : Northern South America, north to j Honduras ; Cuba. General Account: I have records of only five specimens taken at Kartabo, two of which (21272 and 38675) are in the Ameri- can Museum. Usually dark gray with a series of thin V-shaped lines down the back. These mai’ks are sometimes more or less broken up into spots and blotches on a dull greenish back- ground. This general color can shift to a reddish-brown. A female, No. 193a, secured August 10, 1919, laid an egg a few minutes after being placed in a vivarium. I can find no further notes in my journal. Anolis schiedii (Wiegmann, 1834). Name: Small Brown Anolis. Range: Northern South America. General Account: Only one specimen re- corded from Kartabo. No. 3119, September 6, 1919. Length head and body 40 mm. (tail broken). In my field notes not distinguished from other anolids. The type of gibbiceps, a synonym (named by Cope in 1864) was taken at Caracas, and measured 49 mm. from snout to vent. The color description given by Boulenger ( Cata- logue of Lizards British Museum, Vol. II, p. 52) is as follows : “Coppery above, uniform; sometimes with darker spots, or a light vertebral line in the female; lower surfaces whitish with metal- lic gloss.” Iguana iguana iguana (Linnaeus, 1758). (Plate II, Figs. 5 and 6. Plate III, Figs. 7 and 8) Names: Iguana. Kuana (Boviander). Why-ah-mah-kah (Akawai Indian, “one who lives in the bush”). Range: Northern South America with other forms in Central America and the West Indies. General Account: The iguana is the larg- est and most generally distributed of the lizards in the two districts under considera- tion. It is a vegetarian, essentially arboreal and seldom seen on the ground. While it can be considered a green lizard, yet its frequent banding with brown and its power of both pattern and color change make it very diffi- cult to describe in general. A typical individual is an adult female No. 2877, Kartabo, June 3, 1924, Color Plates 754 and 755: Measurements: Total length 635 mm., head 6 per cent, of length, body 34, tail 60, fore leg 15.7, hind leg 24.5 per cent., weight 249 grams. Color in Life : General color above gray- ish-olive with wide wash markings, gener- ally pointing backward, of bluish-green with interspaces black variegated with small spots of light grayish-olive. A few of the lateral patches are vivid green and there are four very irregular, rectangular areas, edged posteriorly with black, of pale yellow green with dark stippling. The neck region has a black-edged cape of teagreen, fading to pale olive along its scalloped rear edge, which extends from the beginning of me vertical crest of spines on the nape, to a point a little back of the armpit. This cape is streaked and mottled with black and merges into the olive buff of the dewlap and the faintly bluish-green and pinkish scales of the head. The dewlap is marked with sev- eral almost vertical streaks of dark gray. Top of head strongly tinged with olive. A patch of black behind the orbit and one just below it, also a streak of blue green along the gape. The spines of the dewlap are buff, and there is a shade of wine color alone the lower edge of this structure. The large cheek 202 Zoologica: New York Zoological Society [XXIX: 18 scale is pale green with a patch of pink. Tympanum apricot yellow. Ventral surfaces pale gx-een along the middle, olive buff on each side. Under limbs olive buff, mottled with darker. Fingers and three inner toes bright green. Limbs very dark brown with faint water markings of green, thickly sprinkled with olive. Iris cinnamon rufous tinged with dark brown and an inner gold ring. A young iguana of 200 mm. showed a simple pattern of general brilliant grass green above with five irregular bands of dark brown around the body and six around basal third of tail. The caudal rings shaded backward from deep green to pale blue green. Text-fig. 3. Iguana iguana. Fundus Oculi. Fundus oculi : The eye of female No. 2877 was examined and described by Dr. Casey Wood and Mabel Satterlee on the day the lizard was caught, June 3, 1924, Text-fig. 3. Eyeground gray with darker stipples cover- ing the entire fundus. The optic nerve is a rough cii’cle almost entirely covered by the pecten. It is white and has white opaque nerve fibres radiating from it into the sur- rounding fundus. The pecten is dark brown, almost black. Its base is fairly circular with uneven edges, but the lower end is pro- longed in two little swallow-tail points. It comes up into a high point which overhangs the optic at the upper edge for quite a dis- tance. White opaque nerve fibres radiate from the papilla and there are dark brown- ish lines interspersed between them. Hyoid : Coll. No. 3128, Kartabo, half- grown male, July 4, 1920, total length 635 mm., Text-fig. 4. The glossohyal extends for- ward as a slender rod, of equal diameter throughout its length of 8.2 mm. At its proximal end, it enlarges into the basihyals, which form an arch of only slightly greater Text-fig. 4. Iguana iguana. Dorsal view of Hyoid. diameter, and which is given an artistic touch by two scallops, evident both on the outside and on the inside of the arch. Their practical use is apparent on the outside where these indentations act as sockets from which arise the proximal elements of the hyoid and the first branchial arches. From the anterior facet spring the hypo- hyals, 3.3 mm., in length, ossified wedges, diminishing slightly in diameter and di- rected slightly anteriorly. From their tips, the very long, slender and somewhat curved ceratohyals extend backward at a sharp angle from the hypohyals, each 25 mm. in length. Although the tips are very attenu- ated and curved I can detect no differentia- tion into distal epihyals. From the posterior, basihyal, concave sockets arise the pair of first ceratobran- chials, each 15 mm. in length. The juncture is by a rounded hyaline cap within which is an ossified core. These elements are slightly greater in diameter than the cerato- hyals, and are directed along a closely simi- lar course. Continuing from the distal ends of the first ceratobranchials, are two con- secutive elements, the first very short, 1 mm., and the end portion, 7.5 mm., long, slender and curved. If this condition was present in only one of the ceratobranchials, I should consider it an abnormal break, but it is identical in both. I believe the short intercalated rectangle is of the nature of a connecting element of use in providing in- creased mobility, and the longer, curved distal element as the first epibranchial. The second ceratobranchial arises as pos- terior, undifferentiated extensions of the basihyal arms, well separated proximally, and only slightly nearer one another at their 203 extremities. The left is 23.2 mm. long and its companion slightly shorter. Hyoid, Side Vieiv: In the usual or dorsal view of a lizard hyoid the tips of the cera- tohyals and the first ceratobranchials often cross one another, and the second cerato- branchials appear to extend straight back between the other elements on the same plane. To reorient this optical illusion of a ' flat plane point of view I have had the lateral view of the iguana’s hyoid drawn. Text-fig. 5. Here we have in dotted outline the toothed upper jaw, and a corresponding outline of the scale-denticulated dewlap or gular wattle. The glossohyal is seen to project forward in support of the tongue, the ceratohyals and first ceratobranchials curve up and around toward their former connection with the auditory meatus. The downward and back- ward directed second ceratobranchials play an important part in supporting the great throat dewlap. It is difficult, however, to understand how they can play an efficient part in the extension of the wattle, without simultaneously elevating the tongue. The apparent strong ossification suggested by the scarlet dye of the KOH is very evi- dently completely subordinate to the elas- ticity of the cartilage foundation. ■ Food: Two iguanas shot on March 24, 1919, 450 and 640 mm. in length, had the crops crammed with hundreds of small, un- chewed leaves, packed tightly together like the leaves of a book, each le&f being about a half-inch across. The stomach was equally filled with the same material crushed and comminuted into a greenish pulp. Leaves and petals of flowers seem to be their chief diet, but they will eat pawpaws, jungle cher- ries and wild guava. Enemies: On January 4, 1921, a small iguana, less than 300 mm. over all, was dropped almost at my feet by a swallow- tailed kite, Elanoides forficatus yetapa, as it swooped across the trail east of the Kartabo laboratory. We have seen a jaguar in full pursuit of a large iguana. Breeding: On September 26, 1922, four eggs were found buried under a bank, partly washed out by the rains. They corre- sponded exactly with ripe eggs taken from a female iguana. These eggs were one-third developed and showed an interesting dis- parity in shape and weight. One was round, the others elongated ovals. The shell was chalky white where not stained, rough, mi- nutely and irregularly pitted. The size varied from 42 and 46 to 47.3 and 50.8, with an average of 44.5 by 49.4 mm. The weight extremes were 44.4 and 75.8, averaging 58.3 grams. 204 Zoologica: New York Zoological Society [XXIX: 18 Seven eggs were unearthed by a pet mon- key at the foot of a dead stump on October 3, 1920. One egg was a small, round, runt, the rest were straight-sided ovals. Habits: On February 24, 1942, five days after our arrival at Carpito, Venezuela, I noticed that a tall bare tree at the western edge of our compound had a single clump of buds looking like a small parasitic plant high up among the naked branches. We later learned that this tree, a relic of the cut jungle, was eighty-six feet high, and was a Yellow Puoi, Tecoma serratifolia. On March 3 the tree showed 17 widely scat- tered panicles of golden yellow bloom, be- sides scores of opening buds. On the eighth of the month I saw a four-foot iguana high up among the scanty flowers. From 6:30 A.M. to 1:30 P.M., at least, he remained motionless. The next day the lizard was in the same crotch at 7 in the morning and as far as I know had not moved at 5:45 P.M. From March 11 on, the iguana was a permanent resident of the tree. An account of his activities on this one day will suffice for all others. He seemed to slumber soundly during the morning, although with my 40- power binoculars I could see his eyelids move now and then. At 11 o’clock he began to climb slowly up and about, and at 1 P.M. he was in the midst of the very topmost cluster of bloom, eating all he could reach and he could reach any he chose. His long tail hung down and he sometimes slipped on the slender bending twigs, and hung on with only the toes of one fore limb. The toes of all four have the power of locking tight about twigs which bend far down with his weight. At first he selected the full- blown flowers but later began on the buds. When sprawled out at the very apex of the tree, one hind leg would occasionally hang loosely, the other clinging with only the middle toe, as he edged slowly out toward the outermost flowers, although four or five other clusters were much nearer on stronger branches. He seemed to enjoy taking chances, and he never gave up. If a first attempt to reach a certain blossom failed, there followed another and another stretch- ing of his neck, then out went his tongue, although it protruded hardly at all beyond his lips. At last he crept an inch nearer, the whole branch bending slowly far over, and when almost upside down he licked in the bud and swallowed it without chewing. On turning to descend he slipped twice and both times hung reversed by his hind toes alone. When he was able to keep his position by means of three feet, the toes of the fourth were used to hook and pull the twigs or panicle to his mouth. The tree was 150 feet from the laboratory porch and 86 feet high, yet the 20 or the 40 diameter binoculars showed every wrinkle of his scaly skin, every tiny fly that alighted on the edge of his eyes. On March 19, I noted that the iguana was back after three days’ absence. In mid- morning he climbed up the tall bare trunk, and on to the very topmost branch and as all the blossoms were gone he devoured a precocious, long, slender, bean pod, the only one on the tree. It hung vertically over his head from the slenderest of twigs and his lips barely touched the end of it. The wind was blowing and as it swung back and forth he snapped at it and now and then bit off a short section. He never was able to secure more than a fraction. On April 11, the tenant- of the Puoi again returned after a week’s absence, spending the morning climb- ing about, plucking and eating the last blos- soms of the season, even taking the small, green leaf buds as well. Again he showed his ability to cling by hanging with the toes of one hind foot, both fore feet engaged in pulling a branch toward him, while the other foot dangled in mid-air. On July 10, 1 noticed the great lizard asleep in a crotch among a mass of new green leaves, and I saw him for the last time on July 24. In the night of July 26, the great tree blew down and the iguana vanished forever. Plica plica (Linnaeus, 1758). (Plate III, Fig. 9). Names: Spiny Tree Lizard. Agama (Cre- ole). Ing-wallack (Akawai Indian). Range: Northern South America. General Account: A medium-sized, ar- boreal, insect-eating lizard. Found occa- sionally on rocks along the river, usually on bark of jungle trees. General characters, broad head and body, slight spiny crest on back and several spiny tufts on neck. Dark green above with wide dark bands mottled on body and solid on limbs and tail. Chin whitish, throat black and necklace black. Sides of head and under parts orange when in full color. Male adult, No. 68, Kartabo, July 23, 1920: Measurements: Total length 373 mm., head 10 per cent, of length, body 23, tail 66, fore leg 20, hind leg 30 per cent. Color in Life: No. 68; Head yellowish- olive, deeper toward snout ; back deep grape green, with seven wide mottled bands of black, the anterior one a collar, the pos- terior on the tail just back of the hind legs. The upper arms and legs similar in color and pattern, there being three mottled cross bands on the arm and four on the leg. On the lower part of the limbs, digits and tail, the bands become solid black and the green gradually pales to olive gray. There are six bands from elbow to tip of toes, eight from the knee down, and thirteen on the tail. The black, mottled nuchal collar is mixed 1944] Beebe: Iguanidae of British Guiana and Venezuela 205 Text - fig. 7. Plica plica. Position on tree. with smoke gray and changes abruptly on the side neck to black, which broadens out on the throat. Anteriorly this fades grad- ually through light olive gray to cartridge buff on the chin. Ventral surface: black throat narrowly bordered posteriorly with citron yellow, shading into apricot orange which covers the entire under body, thighs and ventral tail, paling to yellowish-white on distal limbs, soles and tip of tail. Eyelid light grayish-vinaceous; pupil al- most round with slight irregularity at bot- tom. Iris maize yellow, densely flecked with orange and gray, and a narrow inner ring of gold. Arboreal Adaptations'. Almost every scale has a sharp protruding spine and a sharp keel. Rubbed the wrong way on its soles, thighs, belly, chin or under tail it feels like razor grass, and these projections must be of great help in sustaining its position on bark. The fingers are very long and the claws exceedingly long, sharp and bent downward. The exposed dorsal and throat patterns are highly protective, mottled dark and green. In its usual position, upside-down on trunk, the throat patch becomes a shadow and the head an irregular branch stub. The orange of head and sides are of slight conspic.uousness. A female about a foot in length was dis- covered by accident about 12 feet up on the trunk of a smooth-barked tree. She was in a most remarkable attitude, raised high so that no shadow was cast. I went on down the trail and two hours later, when I re- turned, the lizard had not moved a muscle. As seen from the sketch made on the spot 206 Zoologica: Neiv York Zoological Society [XXIX: 18 by my artist, it was head downward, with practically the entire weight supported by one hind leg, which, attached by two or three claws, was stretched out at full length. One front leg extended straight downward, and the other curved slightly upward. The dorsal crest and spiny neck tufts broke up the outline, while the mottled green color above made it the same color as the bark. It was by the merest chance that I dis- covered it. Text-fig. 7. Hyoid: Coll. No. 3549, Kartabo, June 30, 1924, adult male, total length 325 mm. Text- fig. 8. Glossohyal rather short, relatively wide, like a narrow tongue in shape, 7 mm. in length. The mid section is very slightly wider than where it merges into the basi- hyals. These elements show very little lateral thickening, and only a slight, shallow con- cavity indicative of a lateral socket. From the upper part of this area arise the hypo- hyals, short, slender, 3.5 mm. in length, di- rected slightly forward. From their tips, with no evidence of specialized juncture, spring the long, very slender ceratohyals, 16.5 mm. in length, at a backward angle of 50 degrees. Transparent, short epihyals terminate these elements. Below the origin of the hypohyals arise the stout, first ceratobranehials, with a flat- tened, proximal cartilaginous cap. In dor- sal view, these bones bend sharply outward for a distance of 17 mm., terminating in curved, hyaline, first epibranchials. The arms of the basihyals are continued pos- teriorly by the long, slender second cera- tobranchials. Proximally these bend inward toward each other, and maintain a closely parallel but quite separate course for their entire 15 mm. Typical short second epibran- chials are at the extreme ends. A young (24 hours old) Plica plica, Coll. Text-fig. 9. Plica plica. Hyoid of newly-hatched lizard. No. 870, Kartabo, April 25, 1919, total length 185 mm., shows distinct differences from the hyoid of the adult. Text-fig. 9. The relative lengths of the glosso- and hypo- hyals are the same, but there is relatively much less of antero-posterior development in the three elongated elements of the hyoid complex. The ceratohyals and first ceratobranehials in the young lizard pre- sent no change in their respective rela- tive lengths, but the second ceratobran- chials, instead of extending posteriorly an appreciable distance beyond the others, are hardly more than three-fourths of their length. This delayed ontogenetic develop- ment of an exceedingly primitive character, whose secondary function will be to assist in the operation of the very specialized wattle or dewlap, is significant and in- teresting. The only other noticeable character peculiar to the juvenile lizard is a distinct thinning, almost a perforation, in the cen- ter of the basihyal mass or the base of the glossohyal. There is also a decided con- striction of the proximal attachment of the second ceratobranehials into a neck. We can easily imagine that at a slightly earlier stage these elements would be pinched off altogether, thus isolating them and approxi- mating a still more primitive condition. Gross Anatomy: Adult male, Coll. No. 68, Kartabo, July 23, 1920, total length 373 mm. Text-fig. 10. The stomach is elongated, with the short, narrow duodenum turned right at an acute angle. The pancreas is narrow but long, extending along the duodenum and half way up the side of the stomach, partly overlying the long and narrow spleen. The small intestine is narrow and much con- voluted, the folding recalling that of a young alligator. The cloaca is large and volum- 1944] Beebe: Iguanidae of British Guiana and Venezuela 207 inous. The liver is shaped like an isosceles I triangle. It is large and long, covering most of the intestines. It is deeply incised around the gall bladder, with a deep groove extend- tending anteriorly for some distance. Meas- urements are as follows : small intestine 145 mm., large intestine 23.5, cloaca 23.5, duo- denum 15, spleen 9, liver 48 mm., liver weight 1.1 gram. Weight of lizard, 76.7 fl grams. Text-fig. 10. Plica plica. Gross anatomy. Food : The contents of six stomachs were (1) five beetle grubs; (2) large millipede, cicada, many red and black ants; (3) solid mass of more than 100 ants; (4) 85 termites, few winged ants, small scorpion ; (5) large black bee; (6) bee, 40 Atta ants, several beetles, large roach and a spider. In captivity these lizards accepted almost any kind of insect, including butterflies and moths of many species. With my large binoculars I was watching a Plica thirty feet up a tree when a cicada which attempted to alight on the lizard was seized like a flash and eaten. In Caripito I saw three of these lizards high up in tall jungle trees, all of them on the bark, and all quietly waiting upsidedown. Breeding : A female taken May 13, 1924, contained two eggs, soft, but well developed, and soon to be laid. On May 24 of the same year a male was shot in full breeding con- dition. Plica umbra (Linnaeus, 1758). (Plate III, Fig. 10. Plate V, Fig. 14). Name : Smooth-necked Tree Lizard. Range: Northern South America. General Account: At a distance, the chief difference between adults of plica and umbra is the black dewlap and, when closer, the presence of the spiny neck tufts in the former. P. umbra lacks both of these. In habitat, usual position on tree trunks, gen- eral reactions, and diet, the two species are indistinguishable. The head is short and broad, eye ridge prominent, color above in general olive with darker bands and- mottlings; greenish-white or yellowish be- low. If we concede the power of conceal- ment is equal and the chances of discovery the same, then umbra is about three times as abundant, judging by the number seen or collected. The relative number- of young individuals observed is very much greater in umbra than in the other species. Male adult, No. 14, Kartabo, April 23, 1919: Measurements: Total length 256 mm., head 7 per cent, in length, body 24, tail 69, fore leg 16, hind leg 24 per cent. Weight 12.2 grams. The weights of male lizards varied from 10 grams for a 203 mm., specimen to 22.5 grams for one of 300 mm. Color in Life: The principal pattern and color variations are covered by the following four descriptions. Specimen No. 14, male: Upper surfaces and sides generally pea-green, becoming more yellow and changing to light yellow drab on the tail. Everywhere irregularly blotched with black, the black spots coalescing more and more as they approach the hind limbs, and finally on the tail becoming large spots of sepia. Upper arms and legs irregularly spotted with blackish-brown, much more concentrated on the tibia and fore arms, forming bars. At the articulations of the fingers and toes a spot of sepia'. Upper and lower labials, head below a line from the nostrils over the canthus and eye to the tympanum, irregularly spotted and mottled with various shades of green, light green- ish-yellow predominating. Lower surface of head, body, arms and legs pale pinkish-buff, changing to apricot yellow on tail. A decided orange cast between the two longitudinal folds of the gular region. Specimen No. 14a, male breeding, July 27, 1920, total length 284 mm. : General color of back absinthe green becoming more in- tense on shoulders and dull lime green on limbs and tail. Scales of head and face a mixture of brilliant and pale grenadine pink, mottled with gray and green. Pineal scale lime green with gray central dot. Small dusky spot on back of neck. Five backward- pointing, transverse dorsal bands, the first deep black with central portion of dragon’s 208 Zoologica: New York Zoological Society [XXIX: 18 Text-fig. 11. Polychrus marmoratus. Posture in tree. blood red, connected laterally with the sec- ond which shows less intense black and more red. Last three bands paler red, with black framing before and behind, the sides of the bands breaking more and more into various sized black spots. Beginning at the thighs, 17 dull black bands extend to tip of tail, the bands being twice the width of the interspaces. Limbs and feet banded like the tail. Specimen No. 808, male breeding, Septem- ber 29, 1922, total length 300 mm.: Above, head blue green, body and limbs pale jade green. Broad, black irregular bandings, six across body, two on upper and two on lower arms and legs, and 15 on tail. The color on the tail changes gradually to citrine brown. Below head and jaws tea green, shading centrally into old gold ; body and limbs smoke gray. An unusual character is four marks of strong yellow ochre around vent, form- ing three, narrow, tall triangles pointing forward along abdomen and outward along thighs; and a central spot directly in front of the vent. The dewlap, usually hidden, is lemon yellow. Specimen No. 2661, female, March 24, 1924, total length 280 mm. : Top of head and neck gray. A short crest of cream and black pointed scales running from above tympa- num to mid-body. Face and all labials bluish- glaucous with patches of straw yellow in front of typanum each side of eye. Side of neck straw yellow, clouded with gray. Gular pouch light ochraceous buff becom- ing darker on chin and with a patch of red between humeral and gular folds. A broad band of black obliquely across shoulder. Upper body and limbs forest green with eight transverse bands of mottled brown and black, others on upper limbs. Body color changes to red brown on tail, brown at tip, which has nine mottled spots of black. Gen- eral ventral color vinaceous fawn, deeper around vent and brown under tail. Iris dark speckled golden, with clear inner pupil ring. Food : Four stomach contents were as fol- lows: (1) 8 beetles, 2 ants, 1 katydid; (2) 27 large black ants, 3 small beetles, 4 small centipedes; (3) numerous black ants and beetles, 6 centipedes; (4) 2 June beetles, 30 small black ants. Breeding : A male and female lizard (No. 191a) were taken while mating on Septem- ber 9, 1919. Other lizards in full breeding condition were collected in April (3 rec- ords), May (3) and October (2). Surpris- ingly small lizards, less than 220 mm., both male and female, were found to be breeding. A female, No. 216, October 14, 1920, kept in captivity for several weeks, laid two oblong eggs, 10 by 20 mm., white, with 1944] Beebe: Iguanidae of British Guiana and Venezuela 209 Text-fig. 12. Polyclwus marmoratus. Posture in tree. equally rounded ends, the surface rough, with many irregularities tending in a trans- verse direction. Courtship : Several times we saw the vio- lent bobbing of the head, the lizard stand- ing upsidedown on a trunk, stretched high on its fore legs, always in the presence of another individual. In the case of the pair shot while mating, both sexes bobbed when first observed. This mated pair were dis- tinguished by the following characters: Male, total length 290 mm., weight 17.7 grams. In general mottled dark green above, with dark bluish head, and much yellow buff about the vent and thighs. Female, total length 246 mm., weight 15.7 grams. Dull olive green above with greenish head and no yellow buff. All dorsal colors subdued and so merged that no specific description is pos- sible; exactly like the mottled, lichened tree trunk to which she clung. Polychrus marmoratus marmoratus (Linnaeus, 1758). (Plate IV, Figs. 11-13. Plate V, Figs. 15 and 16). Names : Many - colored Tree Lizard ; Marbled Lizard. Gamma or Colonial Chame- leon (Guiana Creole). Ee-wang-quee-buh- nah (Akawai Indian meaning, “Hungry father”) . Range : Northeastern South America. Occurence: Both at Kartabo and Caripito this species is fairly common and probably might be called abundant if our eyes were keen enough to detect more than a small per- centage of those clinging motionless and simulating their backgrounds on the branches of bushes and moderately tall trees. Color in Life, General: This lizard defies any exact description as to pattern or color. The most usual appearance is of a large. 210 Zoologica : New York Zoological Society [XXIX: 18 slender, green saurian with two jet black lines radiating back from the eye, one over the ear, and one to the gape. Usually there is a wide whitish band along the side and a brownish cast to the tail. The entire lizard can become wholly green or altogether brown, and often the back develops a series of five or six broad, oblique, straw yellow bands separated by darker bands. The under parts may be immaculate white, or mottled and marbled with green or brown. The markings of the dewlap are varied, often being very obscure, or occasionally flaring as several longitudinal scarlet slashes. The iris shares in this variability. Of three full-grown lizards one had the eye golden-brown marked with silver; a second deep wine color with inner ring of gold, and a third showed the upper quarter warm buff and the rest tawny. The males are smaller than the females, the average total lengths of nine males be- ing 385 mm. and of 13 females 478 mm. Part of this extra length is taken up by extra body length in the females, as the tails of the males average 72 per cent, of the total length as compared with 68 per cent, in the females. The largest female I have measured was 506 mm. The weights varied considerably, being greater of course in breeding females. Ex- tremes in adults of both sexes were 18.6 and 36 grams. The femoral pores varied from nine to fourteen, sometimes in the same individual. Of 12 males, two were symmetrical, in one the left leg had the larger number while in the remaining 10 there were more femoral pores on the right than on the left thigh. Color in Life : No. 637, Kartabo, May 14, 1919, total length 330 mm., although small was fully adult. It permitted a close ap- proach and was colored as follows : lumiere green on head, sides and under parts of the body, darkening to forest green on the dor- sal surface of back and limbs. Labials, chin and throat pale lumiere green, A few irregu- lar and indistinct cephalic, dorsal and femor- al spots of vinaceous brown, and the entire tail, beginning abruptly at the vent, of this color, with about sixteen nodes of lighter bands bordered with darker. A broad ir- regular band opalescent white from shoul- der back along sides, breaking into irregu- lar spots on the outer border of femur and ulna. The most conspicuous marks are the two lateral head lines of black which I have mentioned. A third almost as distinct ex- tends directly downward across the upper and lower lips, while four other broken ones radiate upward to the supraoculars, each ending in a distinct black spot. No. 185, female, August 16, 1920, 452 mm. in length, was of a general vinaceous brown over all the dorsal parts, and fawn color below. This color held after death. Another female was also brown in general with the dorsal oblique bands well developed. This pattern also held after death and in the flat mounted skin, twenty-five years after capture, the pattern and colors are distinct as ever. Change of Color : September 14, 1917, caught a male Polyehrus 432 mm. in length. It was green in general but after five minutes in a vivarium it turned almost uni- form clove brown above, with abdomen and broad tail bands light pinkish wood brown. Faint traces of green remained around the eye and ear, and a large round patch on the left side, just in front of the insertion of the fore leg, remained brilliant green. The change to this sombre color occurred within thirty seconds. After fifteen minutes the green color began to reappear. One week later when put in preservative, the green color returned in full strength, turning at once to an abnormal blue. Habits: Polyehrus, more than any other lizard I know, depends on two factors of safety, change of color to approximate the immediate surroundings, and most remark- able strained and posed attitudes which are maintained for considerable lengths of time. An example is specimen No. 637 caught on a low branch on the edge of the jungle near the Kartabo laboratory. It was hug- ging the end of a brown branch which was covered with a mottling of lichens of almost the same color and pattern as its body. Its tail hung loosely outward exactly like a side twig of the branch, and as I approached it only gripped more tightly. At last it raised itself on its fore legs, distended its dewlap as wide as possible, and then rushed through a tuft of leaves for a few inches, and again clutched the branch, allowing the right fore leg and the left hind leg to dangle free. It now changed to a dominant green and before my eyes melted into its new environment, wholly unlike the brown pre- vious color. I seized it with a sudden, swift motion, when it opened its mouth and did its best to bite me. A moment later it suc- ceeded and the sharp teeth and jaw muscles gave force enough to make several deep tears in the skin. The claws of this lizard are long, sharp and curved and exceedingly un- comfortable when tightly gripping one’s hand. Several times when by accident we spied one of these lizards near the laboratory I took the artist out and had a sketch made direct from the position. In one case the lizard kept the pose for more than three hours, although it looked most strained and uncomfortable. Two of these may be seen in Text-figures 11 and 12. In the first, the lizard is twisted around a small branch. lBeebe: Iguanidae of British Guiana and Venezuela 211 j( gripping with one hind leg and its thigh. The left hind leg has two claw holds and two more on the left fore leg. The right fore leg is wide spread in mid-air and the tail is curved in a horseshoe shape in defiance | of the pull of gravitation. In the second, the body is vertical between two separated j; branches, the whole weight resting on the right hind leg, the only other contact being through three toes of the left fore leg and one of the widely extended left hind leg. The right fore leg is extended in mid-air and | the tail is curved in an S-shape out away : from the branches. Only by sheer accident are these good-sized lizards ever detected, at least by human eyes, and it is reasonable ! to suppose that somewhat the same difficulty I attends the endeavors of enemies. Food: Polychrus is omnivorous. The con- tents of six stomachs were: (1) green katy- i did; (2) large green cetonia beetle; (3) sev- ' eral green-fleshed berries, insect remains; (4) roach and grasshopper ; (5) large grass- ;! hopper, two moths; (6) three beetles, four ! winged ants, cicada, six seeds. Breeding : Lizards in full breeding condi- tion were found in January (1), July (3), and August (2). Closely associated pairs were seen well up in high bushes in June and July. In breeding males the testes were from 8 to 9 mm. in length. No. 210, Kartabo, female, January 1, 1921, length 372 mm., laid three eggs in a vi- varium seven days after capture and then died. Dissection showed four more eggs fully shelled and quite ready to follow the first three. During this entire time the liz- ard remained an almost solid brown, in spite of the fact that the cage was well filled with green leaved plants. The seven eggs av- . eraged 11.7 by 26.2 mm. in size and in weight 1.9 grams. They were long, even- ended ovals or very slightly curved, hinting of a kidney shape. The shell was hard, deeply and longitudinally rugose or fur- rowed. A female taken August 13, 1922, con- tained eight eggs one-third developed, three in the right ovary 3 mm., in length, and four in the left ovary, all 4 mm. in length. Enemies: On August 26, 1922, I took a slightly damaged Polychrus from the stom- ach of a white-collared hawk, Leucopternis albicollis albicollis. This hawk hunts in deep jungle, often rather low down, and would be one of the few dangers to which this lizard is exposed. Hyoid: Female adult, length 330 mm., Cat. No. 637, KOH No. 2021, May 14, 1919. Text-fig. 13. The glossohyal extends for- ward as a stout rod entering a lingual core so dense that the rod is lost and does not appear again. This tongue-shaped lingual core is split anteriorly, and behind it sends back two lateral cornua. It frays out into the tongue tissue, and occupies about half the area of that organ. Posteriorly the glos- sohyal widens slightly into the head of the basihyals which afford attachment to the two anterior arches, and which bifurcate immediately behind. From the slight head of the basihyals two important elements arise, the hyoid and the first branchial arches. The hypohyals are short, rather thick and strongly curved, thus presenting a suitable surface for the support of the posterior lateral branches of the tongue. Their proximal ends approach closely to the basihyal body, just at the shoulder above the forks, and they curve out and forward. Their distal ends are rounded and lie above the broad ends of the ceratohyals, quite unattached except by tissue which permits the two elements to move rather freely upon each other. The ceratohyals are long and straight, and ex- tend obliquely backward at right angles to the hypohyals, beneath whose ends they arise. Their proximal ends are wide and flat, almost scapular in shape, while distally they terminate in a curved hyaline tip, all that is left of epihyals. Returning to the shoulder of the basi- hyals, the first ceratobranchials arise in close association with the other two ele- ments, united to them by hyaline cartilage. They have a well-developed, dense, hemi- spherical cap, which may be the last evi- 212 Zoologica: New York Zoological Society [XXIX: 18 dence of hypobranchials. The ceratobran- chials curve gradually outward and are slender, narrow rods, ending in a moderate, curved, hyaline, pointed tip of epibranchial derivation, which curves around in close proximity to the tips of the hyoid arch. The second ceratobranchials are extremely elongated, parallel and very close together, being surrounded by a hyaline envelope, and joined together by a delicate membrane. The tips end in a transparent spatulate en- largement. Proximally, no distinction is visible between the ends of the basihyals and the beginnings of these second cerato- branchials. Measurements : Lingual core length 7.7 mm., visible glossohyal rod 4.8, hypohyals 3, ceratohyals 13.7, first cerato- branchials 15.4, second ceratobranchials 18.6 mm. Text-fig. 14. Polychrus mwrmoratus. Gross anatomy. Gross Anatomy: No. 185, Kartabo, fe- male, August 16, 1920, length 452 mm., nostril to anus 124 mm., weight 37 grams. Small intestine 117 mm., rectum 23, cloaca 16 mm., liver weight .8 gram. Text-fig. 14. The body walls of the abdomen are lined with a blackish membrane. Each lung sends down three long, narrow, noded tentacles which may be inflated, and which pass be- tween the folds of the intestine, around the ovaries and down to the kidneys. The lower portion of the stomach is curved to the right. Each ovary contains four large eggs with yolk almost of full size. There are a few very small ova. The large epigastric vein enters the larger portal vein some distance below the liver; the mesenteric enters it more pos- teriorly, as does the splenic. The portal passes over the pancreas and under the duo- denum, being attached to the walls of the intestine, after the mesenteric has entered it. Arising from the cloaca, it passes for- ward to meet the mesenteric, being joined by branches from the rectum and a large vein from the lower portion of the small intestine. A system of veins drains from the stomach into the splenic and thence into the portal. Several gastric veins run directly into the left lobe of the liver from the stomach. The spleen is small and bean- shaped. Tropidurus torquafus hispidus (Spix, 1825). (Plate VI, Fig. 19). Name: Spotted Tree Lizard. Range: Northern South America. General Account: A rare lizard at Kar- tabo and not recorded from Caripito. Only four specimens taken of which two are lost. At a distance probably confused with young Plica. Measurements: Adult male. No. 306, April 25, 1919, length 175 mm., head 8 per cent, of length, body 24, tail 68, fore leg 16, hind leg 25 per cent. Juvenile male, No. 2768, April 21, 1924, Color Plate 706, length 127 mm., head 8.6 per cent, of length, body 28, tail 63.4 per cent. Color in Life: No. 306, olive brown above marked with black. Anteriorly these mark- ings are in the form of vertical bands, three passing through the eyelids, followed by eleven from the nape to the mid-back. When the fore limb is adpressed to the body the 5, 6, 7 and 8th bands are continued across the upper arm. The posterior back and the tail are unmarked. Under parts from chin to vent bright salmon. Sides of chin and throat marked with broken lines, with a very heavy converging broken band on the mid-throat and side fore shoulder. Two transverse broken bands extend across the ventral fore body, and the lower sides of the body are rather finely spotted. The dorsal crest is well marked, starting on the nape and ending abruptly half way down the tail, although this member is not regenerated. The crest is low and even, as far as the sacral region where it doubles its height and maintains this to the posterior end. In the juvenile specimen, No. 2768, the 1944] Beebe: Iguanidae of British Guiana and Venezuela 213 upper surface is bluish-gray, the occipital plate black, with a wide, conspicuous back- ward-pointing V of creamy white. Dorsal 1 crest with wide spaced bands of black. Four narrow black lines radiate from the eye- lids, forward, down and back. Five addi- tional, oblique broken black bands are visible between the nape and shoulder, the posterior one a strong, solid streak in front of the insertion of the fore arm. Adult female breeding, No. 306a, length 141 mm., is dark above, paler below, spotted and mottled coarsely on chin and throat, ■ back to mid-body. Breeding : Specimen No. 306a contained ; two eggs about to be laid, apparently full- I sized, measuring 10 by 15 mm., one on each ■ side of the body. Uranoscodon superciliosa (Linnaeus, 1758). (Plate VI, Figs. 17 and 18). Name: Gray, or Brown Tree Lizard. Ag- ama (Creole). Yamu-koo-roo (Akawai In- dian, “crossing over water, because he swims creeks and makes bubbling noises when swimming”). Range : Northeastern South America. General Account : These lizards are not uncommon, both at Kartabo and Caripito, j but are doubtless more often overlooked than discovered. They are among the dullest of monochrome saurians with habits directly correlated with their pigmentation. They live chiefly in dense jungle, but are also found on rank growths along the shores of creeks and rivers. Most of the day is spent in clinging close to a twig or branch as brown as they are, and only by accident did we ever locate them. They never try to es- cape, but allow themselves to be picked off, legs dangling in mid-air, eyes half open, making no effort to escape or bite. They may be hung on a sleeve and carried back to the laboratory with no concern about their moving. I have put one on a branch in a vivarium and had it remain without movement for 38 hours, although the eyes were open, and the lizard semed fully conscious of what went on around him. When the gray lizard dqes move it is with a headlong rush. On level ground I have seen them more than once work up to bi- pedal locomotion for a few yards. They are gentle and never make any attempt to bite. Measurements: In lizards of 226 to 400 mm. total length, the measurements show little variation. The average is, head 7 per cent, in length, body 23, and tail 70 per cent. Weights, length 150 mm. (3.4 grams), 226 (10), 320 mm. (20 grams). Color in Life: Male, No. 810, Kartabo, length 400 mm., September 30, 1922 : Gen- eral color above grayish-brown mottled faintly with spots of buff which are scat- tered along the side of the head and dorsal crest, and in six large, irregular groupings along the side of the body. Inner edge of eyelids, throat and slight dewlap, and an elongated patch of mottling along ventral surface colonial buffy yellow. Limbs strong- ly banded with darker color. Whitish-brown below. Iris orange rufous flecked with gold and darker on outer edge. Narrow inner iris rim yellow. In very young lizards there are several pale buffy white lines radiating downward and forward from the eyes across the snout and jaws. A broad whitish band begins back of the eyes and curves over the shoulder and widens into a festooned band along the sides of the body and tail, edged above with black freckling. Hyoid: Coll. No. 3006, Kartabo, August 19, 1922, total length 240 mm. Text-fig. 15. The glossohyal is 5 mm. in length, wide at the base and gradually narrowing to the anterior tip. The basal two-thirds of its length is perforated by a longitudinal aper- ture, so wide that the lateral osseous boun- daries are as narrow as the beginning of the solid distal third. The basihyals show little lateral thickening, but present two dis- tinct although shallow sockets on the ex- ternal lateral aspect. From the upper one arise the hypohyals, fairly heavy, 2 mm. long and extending at right angles to their point of origin. Equally slender ceratohyals arise from their tips, and extend obliquely back- ward for a distance of 9.5 mm., giving rise, in turn, to long, slender epihyals. The first ceratob ranch ials originate at the lower sock- et of the basihyals, and are considerably heavier than the ceratohyals, and 8.2 mm. 214 Zoologica: New York Zoological Society [XXIX: 18 Text-fig. 16. Uranoscodon superciliosa. Embryo. in length. The terminal first epibranchials are longer and thicker than the epihyals. The central foramen in the body of the basi- hyals is a good-sized oval, and the lateral arms are continuous and fused with the pos- teriorly directed second ceratobranchials, which extend straight backward for 5 mm., separated from each other by a narrow, clear channel. No signs of second epibran- chials are visible. In the perforation of the glossohyal and the relative shortness of the second cera- tobranchials the hyoid of this two-thirds grown Uranoscodon bears a striking resem- blance to that of a half-grown Plica plica. Food : In capitivity one gray lizard lived for a month and a half voluntarily without food, although surrounded and in the midst of plenty. There was little evidence of emaci- ation. When at last the prolonged fast was broken, the lizard readily devoured quanti- ties of orthoptera, caterpillars and grubs, but showed little relish for termites. A young male taken August 19, 1922, had eaten a millipede and three lepidopterous larvae. Breeding : On August 5, 1920, three miles out along the Puruni Trail, Kartabo, Dr. William Morton Wheeler called to me to come and see a large U ranoscodon on her nest. She was in a hollow in a half-broken knot-hole of a decayed log, and had deposited 11 eggs. Unlike the usual reaction of this species she was nervous and timid, and at our close approach she rushed out and up the nearest tree trunk. 1 stunned her with a bullet on the bark beneath her and put her in a snake bag. She was No. 312, August 5, 1920, length 405 mm., weight 26 grams, Color Plate 216. The hollow bore evidence of considerable scraping, as if enlarged by the lizard. Re- mains of old bits of shell of the same char- acter as those freshly deposited hinted at a former occupancy. One egg, accidently broken, was perfectly fresh. The 10 remaining eggs weighed 24.4 Text-fig. centro7i Drawing- 17. Uro- azureum. of adult. 1944] Beebe: Iguanidae of British Guiana and Venezuela 215 grams, averaging 2.4 grams. The average size was 13.4 by 24.3 mm. The shape was rather variable, one end slightly smaller than the other, three noticeably wider across the middle, the rest with sides parallel. The ; shell was leathery, rather soft, roughened, with low striae extending transversely around the shell. The shells were already brownish, stained from contact with the soil, but they washed to a dull white. Sixteen days later the embryo was well developed, limb pads entire. Text-fig. 16. At 30 days the embryo measured 33.5 mm., with other percentages as follows: head 19 j per cent., body 35, tail 46, eye diameter 12, arm bud 10, leg bud 14 per cent. Either from lack of moisture or other cause the eggs failed to hatch. On August 22, 1922, six eggs of this spe- cies were brought in by our Indian hunter, taken from a hollow 12 feet up a tree. The eggs averaged 13.6 by 25 mm., and in weight 2.7 grams. Urocentron azureum (Linnaeus, 1758). Names : Spine-tailed Lizard. Tali-tali (Akawai Indian). Range : Brazil and the Guianas. General Account : Two specimens were taken at Kartabo, both of which have dis- appeared. Specimen No. 41 was found by me among some rocks near the shore at Kartabo, on April 28, 1919. It was 100 mm. in length. Color Plate 35. Identification from the color plate seems certain. Text-fig. 17. A second specimen was brought in by my collector in 1922, and either escaped before being killed or was lost in transit. References. Cocteau, J. T. 1837. In Dumeril and Bibron, Erpetologie general ou histoire naturelle complete des reptiles. Vol. IV, p. 149. D’Orbigny, A. D. 1837. In Dumeril and Bibron, Erpetologie general ou histoire naturelle complete des reptiles. Vol. IV, p. 110. Daudin, F. M. 1802. Histoire Naturelle des Reptiles, Vol. IV, p. 84. Dumeril, A.M.C., and G. Bibron. 1837. Erpetologie general ou histoire natu- relle complete des reptiles. Vol. IV, p. 94. Gray, J. E. 1840. Catalogue of the Species of Reptiles collected in Cuba by W. S. MacLeay, with some notes of their Habits ex- tracted from his MS. Ann. and Mag. Nat. Hist., Vol. V, p. 114. Linnaeus, Carl 1758. Systema Naturae, Edition 10, p. 200, p. 202, p. 206, p. 207, p. 208 and p. 697. Spix, J. P. von 1825. Spec. Novae Lacert. Bras., p. 12. Wagler, J. 1830. Nat. Syst. Amph., p. 149. Wiegmann, A. F. A. 1834. Herpetologia Mexieana, p. 47 and p. 48. 216 Zoologica: New York Zoological Society Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. EXPLANATION OF THE PLATES. Plate IV. Fig. 11. Poly emus marmoratus, Caripito. X %. Head, body and legs bright green, dotted with gold; belly and tail gray marked with brown. Plate I. 1. Anolis chysolepis, breeding male, Spec. 2778. X 3. Upper parts brown and gray, dewlap scarlet. 2. Anolis fusco-auratus. X 2. Pale green with gray markings. 3. Anolis nitens, Spec. 530. X 2. Light brownish-yellow marked with dark brown, dewlap scarlet. 4. Anolis nitens (right) and Anolis chry- solepis (left). Photograph from life. Natural size. Plate II. 5. Iguana iguana on branch. X %. 6. Head of adult Iguana. Natural size. Plate III. 7. Iguana iguana half-grown, Spec. 2877. X 2. Head and dewlap gray, marked with olive green on body and legs ; feet green. 8. Seven eggs of Iguana. X %. 9. Plica plica, Spec. 2844. Natural size. Face and upper labials orange, dewlap pale brown and black; body pale blue and green marked with black; legs and toes black banded with lemon yellow. 10. Plica umbra. X IV3. Side of face and labials pale blue, top of head gray, dewlap golden yellow, body and legs green marked with black. Fig. 12. Polychrus marmoratus, Caripito. Head, natural size, green marked with black, dewlap green with scarlet streaks. Fig. 13. Polychrus marmoratus, Spec. 3544 Fe- male. X %. Flat skin, brown phase, light golden brown marked with dark brown. Plate V. 14. Plica umbra. X %. Female with eggs. 15. Polychrus marmoratus in jungle. X !4. 16. Polychrus marmoratus. Seven eggs. Natural size. Plate VI. Fig. 17. Uranoscodon superciliosa, Spec. 515. Half-grown. X IV2. Fig. 18. Uranoscodon superciliosa, Spec. 312. X iy2. Fig. 19. Tropidurus torquatus hispidus, Spec. 306. X 2. Head, body and upper sides of legs olive green, supraocular scales and all under parts bright pink marked with black. Fig. Fig. Fig. FIG 4. FIG. 2. FIG 3. FIG. 1. BEEBE. PLATE I. FIELD NOTES ON THE LIZARDS OF KARTABO, BRITISH GUIANA, AND CARIPITO. VENEZUELA. BEEBE. PLATE II FIG. 6. FIELD NOTES ON THE LIZARDS OF KARTABO. BRITISH GUIANA, AND CARIPITO, VENEZUELA. BEEBE. PLATE III FIG. 8 FIELD NOTES ON THE LIZARDS OF KARTABO, BRITISH GUIANA, AND CAR1PITO, VENEZUELA V-* BEEBE. PLATE IV. FIG. 13. FIELD NOTES ON THE LIZARDS OF KARTABO, BRITISH GUIANA, AND CARIPITO. VENEZUELA. BEEBE. PLATE V. FIG. 14. FIG. 16. FIELD NOTES ON THE LIZARDS OF KARTABO, BRITISH GUIANA, AND CARIPITO, VENEZUELA. BEEBE. PLATE VI. FIELD NOTES ON THE LIZARDS OF KARTABO, BRITISH GUIANA. AND CARIPITO, VENEZUELA. Breder: Materials for Study of Tarpon atlanticus 217 19. Materials for the Study of the Life History of Tarpon atlanticus. C. M. Breder, Jr. American Museum of Natural History. (Text-figures 1-9.) Introduction. Although the tarpon, Tarpon atlanticus (Cuvier and Valenciennes) is a well known sportsman’s fish and is sought for sporting purposes throughout most of its range, aside from a considerable sportsman’s lore there is surprisingly little known about its life history and habits, in spite of the fact that most of its life is spent in thoroughly accessible places. An attempt to obtain scientific data on the life history and behavior of the species was initiated in 1938 under the encourage- ment of the late Mrs. Gracia Rhinehart. For this purpose she gave financial support and provided a location in the heart of tar- pon country. To this end a field laboratory was established on Palmetto Key, in Pine Island Sound on the West Coast of Florida, under the aegis of the New York Aquarium. The cooperation of the Peabody Museum of Natural History, made possible through the good will of the Director at that time, Professor A. E. Parr, and the Zoologist, Dr. Stanley Ball, has been invaluable to the work. It took the form, to a large extent, of lending the Museum Collector, Mr. Mar- shall B. Bishop, for long periods of time. His indefatigable services have been of the greatest value. Among other things, practically the entire burden of the tagging operations devolved on him. The United States Bureau of Fisheries (now part of the Fish and Wildlife Service), through the good offices of Dr. Elmer Higgins and Mr. Robert Nesbit, enabled us to use govern- ment tags for the tagging operations, a feature we found of much value. Later Mr. Nesbit was kind enough to undertake to make celluloid imprints of certain of the tarpon scale samples, a matter which great- ly facilitated the study of the markings on them. He employed a new method suitable for these large scales which was devised by him and without which we would have been considerably handicapped. Scales of the smaller fish were mounted for study by Mr. Paul Benzer of the National Youth Administration at the laboratories of the old New York Aquarium. When the labo- ratory work was about one-half finished, operations were transferred to the Ameri- can Museum of Natural History where most of the remainder of the work was carried out in the laboratories of the Department of Animal Behavior, through the courtesy of Dr. Frank A. Beach. The manuscript was finished in the Department of Ich- thyology. The field laboratory and its location, about four miles south of Boca Grande Pass and three north of Captiva Pass, was found ideal for the purposes. Both of these places are used to a great extent for tarpon angling. Notwithstanding these advantages, the accumulation of> pertinent data was not easy and there is still much to be done, now largely forcibly suspended by the re- strictions imposed by war. For this reason, as well as others, it was thought best to make available the data and their analysis so far as we have been able to carry it. A considerable number of students not already mentioned assisted with many of the items under study. They include the follow- ing list, to whom we are grateful: Mr. L. L. Babcock, author of “The Tarpon;” Dr. Richard Cox, New York University; Mr. B. Dontzin, Cornell University; Mr. L. A. Krumholz, University of Michigan; and Mr. Stewart Springer, Shark Industries, Home- stead, Florida. To this list should be added many anglers, local fishermen and others to whom thanks are due, for one item or another. Especial mention must be made of Mrs. Mary Roberts Rinehart and Mrs. J. Coggeshal, and all the active commercial fishermen of the Spearing family. As this study progressed many other items of ichthyological interest appeared incidentally. Those pertinent to the study of the tarpon itself which have been pub- lished follow: Breder <(1939a, b and c, 1940a, 1942e) , Shlaifer (1941), Shlaifer and Breder (1940). Publications on other subjects directly derived from the work of this field station are as follows: Bishop 218 Zoologica: New York Zoological Society [XXIX: 19 (1940), Breder (1939d, 1940b and c, 1941a, b and c, 1942a, b, c and d, 1943 and 1944 Breder and Springer (1940), Breder and Krumholz (1941 and 1943), Cox and Breder (1943), Gregory and Conrad (1943), Merriman (1940), Storey (1940). Numer- ous other items are in press or still under study. Activity at the field station covered vari- ous periods, as needs demanded and the press of other matters permitted. From 1938 up to and including 1942, parts of each year were spent in field work. Most of this occupied the spring and summer months so as to cover the tarpon spawning season. The Habitat of Tarpon. Tarpon are known mostly in their larger sizes. In lengths mostly upward of three feet they are generally to be found in coast- wise waters not far from shore and most frequently in inlets, estuaries and passes between islands. Sometimes they may be found well up rivers and frequently they enter fresh waters for considerable dis- tances. They are not normally found in the open ocean at any great distance from shore. Fish under two feet in length are only rarely taken in such places as described above and those of considerably smaller sizes are known almost entirely from small bodies of water, and most frequently are land-locked when found. Such places may be a considerable distance from the sea- side. Otherwise the smaller sized fish are generally found well up rivers. In the vicinity of the place where this work was carried out, the Peace River and the Shark River are favored haunts of the species in its smaller sizes, although large fishes are often taken, especially in the lower reaches of the latter. Specimens may range down to sizes of a little over two inches in length. Below that their whereabouts is still a mystery and this item is discussed under a consideration of the larvai tarpon. The finding of small tarpon has frequently been the subject of notices by the finder. Such are discussed by Evermann and Marsh (1902), Coker (1.921a and b), Eigenmann (1921), Beebe (1927 and 1928), Beebe and Tee-Van (1928), Storey and Perry (1933) and Breder (1933, 1937, 1939b and c). All but one of these refer to the occurrences in land-locked or practically land-locked places. These they evidently enter from a previous existence in the sea, as indicated by Bab- cock (1936) and Breder (1939c), at times of temporary connections with open water. In one case Mr. Bishop actually saw them entering a newly storm-made cut which was again closed in a short time, securely land- locking these fishes. Of this case Breder (1939c) wrote: “It has long been suspected that small tarpon may get into the odd places in which they are sometimes found by being driven in on the wings of a hurricane or late summer storm. Mr. Bishop was exception- ally fortunate in being in the field when two earlier-than-usual blows piled up the water into arms and other places that are seldom invaded by high water. Wherever there was such a connection with the sea Mr. Bishop plied his nets to see if the new ocean water brought anything in with it. Since he well knew from last season’s work just what was in these places, he was in an excellent position to obtain a proper under- standing of what was going on — if any- thing. Last year there were no such early storms and the places were either com- pletely dry, or stagnant to such an extent that the only local species capable of sur- vival was the air-breathing tarpon. “Place after place was visited and noth- ing but the usual run of small shore fishes could be found in these freshly inundated pools. Mr. Bishop was about to give up and almost ready to invoke the old and feeble dodge of attributing the presence of fishes in odd places to wading birds carrying eggs adherent to their feet— only tarpon eggs are not sticky — when he came upon the last accessible place he knew about. This is on the south shore of Sanibel Island, near Point Ybel. Here, on overflowed land destined to become dry again shortly, he found his quest — tiny tarpon just as they entered these sequestered lagoons. The first time he saw this place it was still connected with the sea, but later was cut off and rapidly evaporating. These particular little tarpon were doomed to desiccation — or as Bishop puts it, before that, to be cleaned up by the abundant bird life when the water subsides a few inches, and before real desic- cation commences. This would appear to be the frequent fate of many little tarpon every time they attempt to venture inland.’’ Such land-locked places vary considerably, some being actually more saline than the ocean, due to evaporation, while others may be practically fresh water. Frequently these pools are exceedingly foul and filled with decomposing organic matter and are of such a nature that other fishes cannot endure. Tarpon by virtue of their ability to breathe atmospheric oxygen survive easily in such places. The significance of these peculiarities of habitat plays an im- portant part in an attempt to understand the nature of the growth of these fishes as will develop in the discussion of the inter- pretation of the markings on the scales. The anatomy of the lung-like swim blader is discussed by Hildebrand in Babcock (1936) and its importance in respiration 1944] Breder: Materials for Study of Tarpon atlanticus 219 by Shlaifer and Breder (1940), Shlaifer (1941) and Breder (1942e). The occurrence of large tarpon in fresh water has been frequently noted. Such places as Lake Nicaragua, Simmons (1900), Gnl (1907), and Miller (1936) ; the Chagres River, Panama, Meek and Hildebrand (1923), Breder (1925), Hildebrand (1937 and 1939), and Babcock (1936), are both famous for tarpon. Babcock gives an ex- tended list of such places covering, in ad- dition to the United States, Central Amer- ica and the West Indies. He mentions reports of tarpon being found more than one hundred miles inland and the author has seen tarpon in the Rio Tampaon, near Pujal, S.L.P., Mexico, at a point which is considerably more than 100 miles from the sea as the river winds, while Dr. Myron Gordon reports having seen them in the Rio Tonto not distant from the town of Papaloapan, in the State of Oaxaca, Mexico, some eighty miles from the sea measured along the river. The ability of tarpon to live in fresh water is considerable, if not indefinite, and there are still living in a small fresh-water aquarium at the Bingham Oceanographic Laboratory, Yale University, several small specimens taken in 1938, with no indication of deterioration due to the nature of the water, during this period of six years. Body Proportions of Tarpon. In order to be prepared to make as much use as possible of anglers’ data, it became necessary to determine correlations between the various measures that have been used to measure tarpon. No single method is com- pletely satisfactory, but the customary “standard length” is least objectionable for most ichthyological purposes. A series of specimens have been measured according to the several measurements in use by tarpon anglers. This material has been reduced to percentage of the standard length and is given in Table I. The various measurements as here understood may be defined as fol- lows. Standard length. The distance in a straight line from the tip of the snout to the caudal base. Total length. Similar to the standard length but measured to the tip of the caudal fin in a “normal” expanded position. Overall length. Similar to the standard length but measured to the tip of the caudal fin with the lobes pressed together. Length to fork. Similar to standard length but measured to the tip of the shortest central caudal ray. Length including jaiv. Similar to overall length but measured from the tip of the lower jaw when the mouth is closed. Aside from faulty tapes and poor tech- niques, there is considerable difficulty in obtaining accurate measurements of tarpon in the larger sizes. One of the sources of difficulty is the changes in length induced by stretching the fish as it is dragged about, generally by the lower jaw. There is evidently only a slight amount of heterogony in the growth of these fishes but the somewhat changing relationships of the various measures of length is broken down in Table I into a series of size groups with which material may be compared. More of such measurements must be obtained, especially in the larger sizes, before a nomo- graphic or other similar treatment would be warranted. Nevertheless this table can yield estimates from any one of the mea- surements that are of sufficient accuracy to give reasonably satisfactory estimates. These figures have been based on wild fish exclusively, for it has been found that specimens kept in aquaria for long periods showed a tendency for the tail to grow proportionally longer. This condition is pos- sibly related to lack of vigorous swimming enforced on such specimens. Due to the fish’s efforts to swim through the glass, sometimes continual, an effort carried on longer in this species than in any other known to the author, there frequently re- sults a deformation of the jaw, which, while not apparently inconveniencing the fish, further modifies the relations of its lon- gitudinal proportions. The weight-length relationships of tarpon need many more data to be worked out in a thoroughly satisfactory manner. Normally tarpon are slab-sided fish, relatively light in weight for their length, but during the spawning season the females may increase considerably in weight. Although anglers like to weigh their fish, it is probably the least satisfactory measure for any analy- tical study and little, of use in present connections, can be done with catch records that give weight only. This is further com- plicated by the fact that in the warm climates in which the tarpon is found, the weight of the fish changes considerably after lying around in boats or on docks for variable periods, sometimes wet and sometimes dry. Babcock (1936) discusses this at length. He gives a formula showing the relationship of length and weight, in which the girth of the fish is taken into account. At this writing we are unable to improve on the accuracy of this purely empirical expression and probably it is adequate to the present uncertain nature of the variations that come into any mea- surement made on large tarpon. The for- mula which Babcock (1936) gives may be expressed as follows. 220 Zoologica : New York Zoological Society [XXIX: 19 X* O OCJiONtoO c3 06 1 | >-H 05 05 GO >H GO GO GO GO GO <55 "5 a CD OhH^HOH . o3 o v CD 1 CO H r-H 05 05 O 00 SH! CO CO CO CO 05 CO CO 05 I-H t-H i-H t r— 1 t-H i-H i-H m m a O ^-o lo 00 05 Eh £ 1 C) ^f-oo >H o«! ^ H GO >H GO 05 05 GO 05 05 >-| >-s T-H T— ( »-H a 2 | Hoo?N£ig. m Ph P O « 0 H si m to o ‘55 H o O ’S cn fc C -g •— < 2 -3 m & §13 | §!> g aS S g 2 > fl a ^ S

^ 0 QI^H ■> — < "* — I '' — I '' — I — l -I — I ■» — j ■< — l '< — t >'S >~H PR O £ ^(MhCOOOO^HOCOO CM E-» c3 I in CD (MMhhhO^OOCOCON I tJ« ffiUH t— li— (,— (r-Hr— (r-HOOOOOO O r-Hr— I r-H r— I i — I t-H i-H rH t— I t-H i-H t — I l — l o £ , >0 ® ®! O A Ol Hi N Ol H^-Oj 3 ;2 C) >-h O O 05 05 oj ?h-’ ^ | 00 CO I ! H3 'a-hh-hh'-hOOOOO oo ■’ — < ■< — I H > — i > — ( ■> — ( ■> — 1 — I >H >H oq x ffi cS o d O S § O ( OCjlQl^SOCi co I I I GO Go GO Go GO Go >-h >-h t-h >» | >H >-h >h Gi> rH lO 0 CO N 00 iO ^ I doiooodood I CO Tt< CO (N 05 CO CM t— I t— ( i— ( t-H t— I t-H rH t-H O >-h ^"t-h GM Go I I Go Go ^ O cd I Go Go ^ ^ (OCC ^ CM CM 05 ^ ^ Gl O Ci Oo IQ i M « o §s ° a o fc “-s O «4H 0 OQ o £ § fe.§ S -2Q S S o p O £ H d PQ H > — I ■»— H -* — 1 cc ffi S g O <3 3s ^ . GO Oi GO GO O 2i.S Qi oo | cd <^£ co cd cd vh G^ O G^ Gi ^ >H >~H T-H >H >H >H >H »H CO CM ’ — i O O ^ 05 05 05 oi oo oo lo 05 N^010050C00005LO 0000000^^^00 O O'. O O Go O O O O ^GO Co ^ cd ^ Go o O* >d 05 Go ^d*d ^ I >- -f-H T-H ',~H >-H ’’“H >"( ’'H CO o d T^05cD»-0^rL0-t'^G0C005O'^C0' & o locdt^odcdo’ t-H ^ 1— I T m ^ W O * ^.S H ^ Q Ph <1 Q <1 Eh CZJ 0 fafi d c3 PR <00000 000 O^cDClNOO^CO 0-HC5CO^LOOOOOC5rHC5C5rtir}-H >H 76.3 18. 143 152.5 43. 091 101.5 11.339 152.5 40. 823 05 lO 101.5 13.607 152.5 43. 091 CD TfH 101.5 7.711 152.5 38. 555 T- H T-H 106.6 20.411 165.0 40. 832 106.6 20.411 165.0 34. 019 o I 107. 0 20.411 167.5 40. 823 "3* I 111.8 11.339 167.5 31.751 >~H 114. 1 13.607 167.5 40. 823 122.0 15.875 175.5 36. 287 /H 137.0 31.751 193.0 47. 626 137.0 27.215 195. 5 52. 162 137.0 27.215 197. 0 56. 698 04 r^' ^' 137.0 29. 483 197.0 56. 698 '^^CD 137.0 39. 462 198.0 249. 0 56. 698 158. 756 1 From Storey and Perry (1933). 2 From Beebe and Tee-Van (1928). 3 From Babcock (1936). 4 From Breder (1925). 1944] Breder: Materials for Study of Tarpon atlanticus 221 Text-fig. 1. Length and weight relationships of tarpon. Based on data of Table II. Solid line represents best fit of the exponential relationship of length and weight. See 0.9 L3 text foi full explanation for formula W = which show only slight heterogonic growth characteristics, than in those displaying large proportional changes with increasing size. Since the tarpon is conspicuously one of the former, as is indicated by the mea- surements of Table III, it should be ex- pected to subscribe to such a formula in a reasonably satisfactory manner. The basic formula states that the weight varies directly as the cube of the length, as follows : W — = c or W = cL3 L3 where W = weight, L = length and c = an empirical constant derived from the above formula and differing for each species dependent on its characteristic slimness or rotundity. When the measurements are ex- pressed in the metric system, it is con- venient to introduce another constant, k, where if, as in the present case, the mea- surements have been made in centimeters and kilograms: k — c 100,000 This then yields the equation: kL3 W = 100,000 100,000 Since it has been found by purely empirical means that for tarpon the value of k giving the closest fit to the data at hand is 0.9, the working expressions become : 0-9 L3 3 1100,000 W W = and L = ■*/ 100,000 y 0.9 when weight is expressed in kilograms and length in centimeters. This equation may be most conveniently transformed for use with inches and pounds in the following manner: L is multiplied by 2.54, the number of cms. in an inch, and W is multiplied by 0.45359, the number of kilos, in a pound. The expressions then be- come as follows : 0.9(2.54L)3 0.45359 W = — — or 100,000 (2.54L)3 SV50,400W W = and L = 50,400 2.54 These, of course, could be transformed in other ways but the above is perhaps as convenient as any. Some of the deviations from the mean in the present data are fairly large, as in- dicated in Text-figure 1, but probably have more to do with individual variation, as- sociated with the sex and the fatness of Table III — Change of Proportions in Relation to Size in Tarpon. Number of specimens precedes each measurement in parenthesis. Dimensions expressed as per cent of standard lengths. 222 Zoologica: New York Zoological Society [XXIX: 19 (>:go®!Co©^go^?'-©}Cooooo|»-i| I © I ,£3©oioooooooo^e^^.cococo to to § 5 o Is- t>* co co co co co ^ ^ co co co « ^ o ^ , ‘C)OiN>N'^>NOo^Co(^OoO | Go | | *0 | • GSoSoSoSoCoSo^oSoCoSoSoSo ^ * § OO^SoSoSoCoSoCo^OSo^-O Vf* p G o d QHNrfiOOOOOihOcDcDiO £ ^ So | vf* D- -f S S N icScd C^Cd C? CO 'i-H (N sosc>so^loio^r^h'^' £ OCOCOH^HTfiOlOHfNH cd M © IO*OCOIO^IO^T^IOCO^CO £ Jg Ol ~i to to to ©3 to to to to ©3 »-< to Q • to "3-© to to to to "S-to c totototototototototototo x totototototototototo ~~ito’-'to’~H’~o co ^7 SS . S . tH '—-h ca W g COCOCOO0(M(M!NCClC I I n ^S^^^SSSSSSSSSSSSS' X ©! © to ^t-to 00 Oo ©J go "i © © ^©©©©°O^N.N.^CoCo?^|goI I©jI ^ <3^ ^ >~~1 I — I 'I — ( >~«-t >"N « J C OiOCCClCD^iOOCKNHNMOi^iOOO^ tf o; OOOOOOOOI>I^I>t^COcDCOCDCO(MCOCO(NC^ d ^ IP ^^^SoSoSoSoSo^OSoSoSo I <3Q I I | w 2 ?! ^ sSSSSSsiS^SSS OiNOiCsOoOiOiOiOi *^<50 So 1 ^ 1 O^OO°0OOOO >-h 1 ^ ^ 1 ^ 1 ^ to©f^oo©©>^©jGo'^to^ 1 to 1 So o So o £ ^(MOiC^iOrf^TfOOCOC'IO^COOOOiCOO cS 0> lOONOOOOHdOO^iONOOrHCDOCOCl eH hhhhhhhiOCO^OCOJO < H H H (M '^(MOiOrlHiO^^GOCOC'IC^COGOOiOOO W H . O £ pa § fi <1 OOOOOOOnN^OO o . 1 1 ° 1 H o 1 1 5 o pa p rj 1^ O O O I^/£3 ^ Od_ ’ , 9 | 1 GO So o CO a) bfi G O ^ CO OHMCO^iOcDOOCld^OOOJO CDNOOCJhhhhhhhhiOCDhhhiM I I I I I I I I I. I I I I I I I I I LOCOI^CXiO^OT-iOlCOTtHiOt^OOt-HCDOOCOOl cn _ O) t}H CO hr, OHiMCO^iOCOOOOiiM^OOCiO p0NOOO5hhhhhhhhiOChhh(N 5 i i i i i i i i i i i i i i i i i i 3 bOcDNGOCiOHMCO^iONGOHCDOOCOiM P5 HHHHHHHlOCD^COOlO HHH(N (164) (147) (144) (144) (144) fc o (U <3 H X w Csi 55 o H O H <3 W Ph X O H 23 O Ph c « o w o X <3 w o o> 4H w £ ,-c £ «j 2 ^ - s, m 0) co T5 c3 a; O C ft o «H C O -s to ^ <3 qo c3NOOOOhhcOOiO^O»OCDcO c3 J ft,ft)-''Hft,ft) I Oo § S*2 ^ s s 0S s SSSSS . ft ft vh>H ft>-Hftftftftftft ft g NNON~HOOO0l0lt)0C!) I os k— I >H^>-H>-H>H>H>'H>'H ^ 'Hftft>'HftOO ft c u2 ^~ft ft ft ft ft ft ft ft ft ft Go I vf- I JH > ft ft ft ft ft ft ft v*- H ^ « . H C3 (NOiHHNWCCNNCOGCO^OfN co c8 ■ • • £ COOO lO CO cfST IM 1-H IN 1—1 CO C« (M v-'v-''-''-''-'v-''~^w-— ''— ' g C 2 1:0 JH ft ft A ftfte^oofto^Oift ^ft ^ I >h I I ^ I f>. >~S >H >~H >H'>*S >--< ft ft >H Q ^oi©»o^»o^^oocoOOT-(©©CO(N ‘-H rHHHHHHrHlOCD^OOO^O HHHd ftftOOOOft-^r-^OO ft ft ftft^QOftft^©i©OVhft^-| I I ft I f> j "^-s >~~H H H > — I "^-l ft ft y ftftCo^-ftftftftCo ^ d ci ^ d d os 00 | I | id c ©©©CO©TjHt^t^.t^ cocoes© 0) OOlOaoOMOOON | COTtHCDiO OQ t— H 03 rH r— I i-H 1— I 03 t-H 1— I H rH iH ^ ftftft>-sft^ftft03 fs. S d 00 Oo N. d ^°0 03 d I I I I Vf. ^ T'H >~H >~~l >~H >~~l '’“-I >~H >~1 wS'S'd'd'S'^i'c^co | dr^dToj ^OOj^-Oo®S<^i©ciC3slC3 GOSS GO m ,5 to®)«®l®j®i®)do i >^C) I Co g^®}o> ^ C’^OSC5>~OOOSC!SOSOSC3 looOs I CO ’'S >*S >H >~-H ^ >~S ’’H ^"4 ©0 ©S' S'^^SSSSSSS y ft ft ft ^3 ft ^hft ft ft ft £j >5 ft >h ^hvh^-ft oo ft I I I ft >-'l ^3 J < fl MN®'t,'^’-t050N(»rt>0(N® g ® OOhhOOOOOOhoOMOCI O r— I t-H rH r-H t-H t-H t-H t-H t-H t-H ’“H m *io>-hCoooCooo®s^C30o us jV§Qodoodddoodoodl I I 03 'do? B ^ Co GO Co 05 Oo ^ H^Ose^OoOof^CoCoOolo Igogo l>~t X ^ H J C *2 ft ft ft ft Vhft “ft ft vf> | <3^ | ft to nco d»dto??co_ | >< ft ft ft ft ft ft ft c3 ft ft ft ft ft ft ft ft ft ft 1 ft ft ft ft ft ft ft >-H ft ■^H >~H © GO CO © © © © © ft 00 © ft O © © © © © GO © GO © © GO © ft <300o^i^'^-i><33GOO©9‘JO cocoos.oooocof^ooasosoo lOr-tcOH<(NCO(M~( >0 Co N 00 C5 o N ®S 93 NT->i3 ft ft ft I £ rH © ft GO © » SI!'1 £ ft © ft GO < Eh ft © Tf* CO O © Tjl GO o ^ CO O 05 in N 05 N b* CM OM 03 CM CM 03 g ONWONOJH CO rH co co CM O CM CM CM CM CM CM CO 1-H 05 ^ I CO „ 1C F-< rt< CM O 00 O rH 00 40 O Is* rS 03 CM CO CM CO 11^1 1 I ^ CO CD ^ b O 05 CO CM CM CM CM *0 O CM CM CM 2 00 C3 CM ~ 05 Pc3 w ^ 'v rd (s. r— I t— I 05 S3 "D O ci QJ KJ A3 2 ~ S- t 05 O *c a; CQ i C «*H . m ^ §J « Jg ca 2 Lh 05 C3 w rS ^ ,2 A3 U -g - o ® a> § -O ,C $ 0) O 35 ? 05 O 0> o > ® ^ (C o c c3 s tH 05 > w C3 o3 "O ffl "O ' s°° *-i CM o 05 r & co H 2 A3 » r o O D O -O' (V) O f-D ,_0 ^ •| COO " ” “ w- d) eg S o> . -+f oj p .52 » _ w *5 PJ C C 3 cj ^ " at u ® o ^ o « o o w ^ a> T3 C , pg -*-> a> co t-i ji *-> o CJ M ^ CU 4> 05 f- J CU M Ph CL. Ph 226 Zoologica: New York Zoological Society [XXIX: 19 mens. Males with running milt were ob- tained without difficulty but it was found impossible to obtain females with eggs in a sufficiently advanced stage to enable suc- cessful fertilization to be accomplished. It is true that anglers often take females from which eggs are running and which give the superficial appearance of being mature. Microscopic examination, however, shows that they are not entirely ripe and exper- iments have demonstrated that they are not yet ready for fertilization. Evidently the vigorous contortions that these fishes make on being boated cause a premature expul- sion of the roe in well distended females approaching ripeness. It is possible, and not unlikely, that females thoroughly ripe do not feed until after spawning. Sexual maturity is reached in this region at a total length of about four feet (122 cm.). The smallest ripe female seen was 139.8 cm. on July 3. A fish with fully undeveloped gonads of 142.2 cm. t. 1. was taken on June 22. This was the largest sexually immature specimen examined. See indication of this in Text- figure 1. At this size they weigh close to 20 kilograms. Since, during May, June and July at least, these fishes do spawn in the region as has been repeatedly observed in shallow water between the numerous islands, the eggs should be recoverable in tow-nets. Often pairs or three fish, evidently two males and one female, may be seen milling around, usually in water shallow enough to cause them to rile up the sand to a considerable extent. During this procedure white streaks or clouds appear, which are almost surely composed of quantities of milt emitted by the males. This was noted by Storey and Perry (1933) from near Sanibel and is in essential agreement with these personal ob- servations. The sand pits described by Bab- cock (1936) may well be produced by spawning tarpon. Several were seen close to areas of unusual activity. Tow-net operations at the surface have yielded no eggs that by any stretch of the imagination could be the eggs of tarpon as compared with the nearly ripe eggs taken from specimens. In this connection it was noted that these nearly ripe eggs sank in sea water even after they had water hardened, a fact first reported by Babcock (1936). Taking this as a hint, towing was undertaken close to the bottom in places near where tarpon were spawning. This method yielded very few eggs of any kind, but among them was one type that we have been unable to eliminate as clearly not that of the tarpon. Many such, in the aggregate, were taken during the years from 1938 to 1942 during the known spawning time of this species. Such an egg is shown in Text- figure 2, a. If this is not the egg of the tarpon it is surprisingly like what the near- ly ripe eggs must approach on full develop- ment. These eggs averaged a little less than 2 mm. (1.8 mm.) in diameter. This com- pares with nearly ripe ovarian eggs which, when water-hardened, averaged nearly this size. Babcock (1936) gives 0.6 to 0.75 mm. for preserved ripe ovarian eggs, which agrees well with our data on the above. The eggs taken by tow-net sank in the sea water in which found, which had a specific gravity of 1.0140 at 25°. They just about floated in the denser off-shore water, where, in fact, eggs could sometimes be taken at the surface tows. They were, however, too disperse to collect in any numbers by the means available. These eggs were hatched and carried on into larval development as far as was pos- sible in the laboratory. The longest any were successfully held in a viable state even under elaborate care was a matter of three days. The various stages that these pass through up to that time are shown in Text- figure 2. The egg shown measured 1.7 mm. in diameter. The details of the other stages passed through are given in the legend to Text-figure 2. It will be noted that the most advanced stage shows characters not unlike those seen in the pre-leptocephalid larvae of apodal fishes and also that there has been considerable shrinkage in the length of the last two stages. Whether this is associated with its evident consolidation or is associated with unsatisfactory condi- tions in the laboratory bowls is uncertain. There was more than the usual difficulty in rearing these larvae for reasons ad- ditional to their own delicacy. It so happens that the shore waters of this region are extremely rich in plankton, as has already been noted by Breder and Krumholz (1943). In fact, it is so rich that a glassful dipped from sea at the end of the laboratory dock looks slightly turbid and if held up to the light can be seen to be full of “jumping” minute crustacea. Some of these are ex- tremely voracious and prey on the newly hatched fish from the numerous kinds of planktonic eggs. It was soon found impos- sible to rear anything hatched from such eggs without placing them in water that had been first filtered through cotton, for only a single organism was necessary to ruin a bowl of hatchlings. It was noted that in these tows there was never any fish life except unhatched eggs and fairly large postlarvae. Knowing the time of hatching of numerous species, tows made at that time would sometimes yield shells, but never recognizable larvae. This was carefully checked on the easily recognizable and abundant eggs of Anchoa mitchilli (Cuvier and Valenciennes). These 1944] Breder : Materials for Study of Tarpon atlanticus 227 Text-pig. 2. Development of larval fish from an egg which may be that of Tarpon atlanticus. A. An egg as taken from tow-net catches made near the bottom, 1.7 mm. dia. B. Ventral view of a larva at the time of hatching, 3.0 mm. long. At this time they float passively in an inverted position. C. Dorsal view of a larva 24 hours from the time of hatching, 3.9 mm. long. By this time they have righted themselves and are active. D. Lateral view of a larva 43 hours from the time of hatching, 2.9 mm. in length. The eye has become pigmented, the yolk is nearly gone and the mouth is open and func- tioning and shrinkage has commenced. E. Lateral view of a larva 72 hours from the time of hatching, 1.9 mm. in length. Two series of chromatophores have appeared ventrally, the otic sapsule is evident and the eyes have become inclined somewhat forward. The gape has reached considerable size and the behavior is voracious. The general appearance approaches that of the pre-leptocephalus of apodal fishes. 228 Zoologica: Nciv York Zoological Society [XXIX: 19 are elliptical and even the empty shells may be recognized at once. This condition was found to be uniformly true for any of the species with which we were sufficiently familiar to make such determinations with- out difficulty. In the same tows would be found large numbers of the postlarvae in the sizes studied by Breder and Krumholz (1943). No variations from this condition were found in tows made in the “green” shore waters. However, when tows were made in the outside “blue” Gulf waters, the newly hatched larvae were taken in abundance. The chief obvious difference be- tween these two colors of water is the relative abundance of plankton, the blue waters representing relatively “desert” con- ditions. Evidently most of the local species of pelagic egg producers spawn both within the blue and green waters, the delimitations of which make a surprisingly sharp line. Consequently it is deduced from this fact and the experiments with filtered water in the laboratory that few, if any, eggs hatch- ing within the limits of the “passes” escape destruction on hatching. Those spawned in the blue waters evidently have a reasonable survival rate, because of the many fewer predators waiting for them, and it is evi- dently the postlarvae developing from these that work back into the inside waters at a size no longer vulnerable to the crustacean elements. Their presence there at this time, make the crustaceans their chief food, in turn. This leaves one with the conclusion that the spawning of such fishes in these shore waters is a complete peripheral wast- age and that the only fishes successfully spawning, in Pine Island Sound, for ex- ample, are those which in some manner protect their young through this vulnerable stage from various planktonic elements. Checking through the species about which there is sufficient data, which fall within the latter categories, the following forms are listed together with their protective mechanism. Fish. Protection. All Elasmobranchs Much too large at birth or hatching Bagre marinus (Mitchill) Galeichthys felis (Linnaeus) Oral incubators and with eggs and young too large at hatching and release. All Cyprinodontidae Too large and advanced at hatching. All Poeciliidae Too large and advanced at birth. Strongylura notata (Poey) Hyporhamphus unifasciatus (Ranzani) Too large and advanced at hatching. All Syngnathus All Hippocampus Too large and advanced on release from brood pouch. All Atherinidae Too large and advanced on hatching. Bathygobius soporator (Cuvier and Valenciennes) Protected by parent in a shell cavity re- mote from the pelagic crustaceans. Gobiosoma robustum Ginsburg Same as preceding. Opsanus beta Goode and Bean Same as preceding and in addition too large at hatching. Paraclinus marmoratus ( Steindachner) Protected by parent in lumen or cavi- ties of a sponge and remote from pel- agic Crustacea. This list covers all species of which we have positive knowledge of spawning in the Sound, many, by their natures, being practic- ally confined to such places. All the others, pelagic egg producers, are also found in out- side waters. It would thus appear that fishes here have two possible courses in regard to reproduction. If producing pelagic eggs, hatching at an early time into small fragile larvae, they must have recourse to at least some reproductive activity in outside waters. Such is evidently the case. The other course, breeding only within the inside “green” waters, can only be specifically successful if some morphologic, developmental or be- havioristic habit insures the protection of the young until they are too large to be preyed upon by planktonic elements. The tarpon has clearly taken the first- mentioned course and only in the “blue” outside waters have larvae of the eggs, that may be those of this species, been taken. None have been taken beyond the size of those reared in the laboratory, the finding of which should lead to their positive or negative identification as tarpon. Clearly this approach to the postlarval tarpon has been no more successful than the attempt of an approach from the larger sizes down, but it does indicate that the larvae develop- ing from this egg, whether tarpon or not, does not move into the estuarine waters as soon as of sufficient size, as do many of the others. Almost certainly this missing and per- haps off-shore stage is a leptocephalus, as would be expected from general considera- tions and the fact that the eastern near relative, Megalops cyprinoides fBrous- sonet), has had such a stage identified. Since the leptocephali of this species have been taken in places not unlike the waters well within Pine Island Sound, van Kampen (1909), an intensive search by all methods available to us has been undertaken. This included towing extensively in all manner of likely and unlikely places, dredging the bottom with small fish trawls, using night lights both from docks and afloat and even digging in sandy and muddy places where amphioxus was found. This last examina- tion was instigated by Mr. Stewart Sprin- 1944] Breder: Materials for Study of Tarpon atlanticus 229 ger, who said that he once saw something along with the amphioxi that may have been a leptocephalus. No such specimens were found. After five years of this exploratory effort in both summer and winter, it would seem fair to assume that the leptocephali are not to be found regularly in these inside places. In attempting to carry on the larvae from the eggs supposed to be those of Tarpon, a variety of treatments were given, including reducing the salinity of the water and transfer to fresh water from pools in which young tarpon lived. In none of these ex- periments was there either an improvement or impairment of viability that could be noted. The latter is the more remarkable since other pelagic eggs similarly treated promptly died. Included in these were An- choa, Paralichthys, Lactophrys as well as many others not fully identified. Extensive towing in the lower reaches of the Peace River and in numerous ponds produced no such eggs. A single pool, on Captiva Island, into which high surf washed over a low sand ridge, contained both adult Anchoa and their eggs, but nothing was seen of the supposed tarpon eggs. Hildebrand (1934) recorded what he be- lieved to be a young tarpon in the transition from leptocephalus to adult form. Unfor- tunately this specimen, which was taken at the mouth of Core Creek, Beaufort, North Carolina, was inadvertently destroyed be- fore a figure was made of it, with the result that all that is now available is the brief description given by Hildebrand. There is no reason to suppose that it is not the young of the tarpon, except for a peculiar feature of the fin ray counts. Hildebrand gives dor- sal 12 and anal 20. It so happens that these are the counts generally given by taxono- mists on adult material, as is shown in Table V. As pointed out in the section on the body proportions of tarpon, in the small- er sizes, of three inches and under, the extra rays are clearly evident. It would seem that surely the counts to be found in a transforming leptocephalus of this species would be, dorsal 15 to 17 and anal 23 to 27. No doubt at this stage and size, 2 cm., some of these rays destined to become consolidated are relatively small, although in the smallest specimens available they actually grade up to the longest ray in each fin. Hildebrand in a personal communication stated that “. . . the figures given, I am sure, were based on the total number of rays (or fulcra) visible under magnification.” This specimen, then, could hardly be the young of the tarpon, un- less there is some unexpectedly late develop- ment of these elements. The other two pos- sibilities, Elops saurus and Albula vulpes, are clearly eliminated on the bases of their much different fin counts, as has been indicated by Hildebrand. The Beaufort re- gion is well out of tarpon country, except as a place for strays to turn up, as also is indicated by Hildebrand. It thus is con- ceivable that this might equally well be a stray from most any place south of North Carolina and might be referable to some other isospondyle, although what could pos- sibly be a suitable form is not evident. It thus appears that this specimen must rest as a probable, but uncertain, record until more material serves to clarify the fin count matter. Juvenile Tarpon. Tarpon are next known from fishes from about 5 cm. upwards. The smallest that we have obtained measures 4.2 cm. in standard length. Various features of these sizes are given in Tables I, II, III, VI and X. Table VI indicates the sizes of the small ones that Mr. Bishop actually saw entering a small pool from the open sea. A second small sample taken twelve days later from this same place, then land-locked, suggests that as a group they showed a mean growth of a little over 2 cm. which is not out of line with what one might expect of a species reaching such a large size. The actual growth rates of some individual fish under a variety of conditions is given in Table VII. These are all fully positive records of marked individuals and serve to indicate the response in growth variation that these fishes show to various kinds of environ- ments. It has been long suspected that the tarpon found in small land-locked pools may be stunted. The first part of the table indicates that certain individuals under such conditions may show no growth whatever in periods up to and over five months. In respect to growth these fish are comparable to those kept in aquaria of various sizes, which is in reasonable accord with what one would expect from the evident mean growth of those taken as they entered such places. In this same table are given calculations of the rates of increment in length, reduced to a standard annual basis, for purposes of comparison with the apparent increase of the fishes measured in Table VI. From this it is clear that the smallest tarpon, evidently just out of the leptocephalus stage, are growing at a relatively terrific rate, but as soon as they reach the size of the smallest otherwise to be found in this pool they drop down to a relatively slow growth. Those kept in the laboratory pool and in small aquaria showed a mean growth of the same sort, while those in the Sanibel pools showed marked stunting, as already mentioned. These relationships are indicated in Text- figure 5. Larger specimens kept at the old New York Aquarium showed a higher growth 230 Zoologica: New York Zoological Society [XXIX: 19 Table VI. Measurements of Collections of the Smallest Tarpon. These include the smallest specimens of definite record. They were collected, on the first date, just as they entered a pond on Sanibel Island, from the Sea. Cm. standard number of specimens lengths Aug. 8, 1939 Aug. 20, 1939 4.2 1 - 4.9 1 - 5.0 5 - 5.1 1 - 5.2 4 - 5.3 8 - 5.4 8 - 5.5 5 - 5.6 2 - 5.7 5 - 5.8 5 - 5.9 3 - 6.0 2 - 6. 1 - - 6.2 1 - 6.3 - - 6.4 - - 6.5 - - 6.6 - - 6.7 1 - 6.8 - - 6.9 - - 7.0 - 1 7.1 - - 7.2 - - 7.3 — - 7.4 - 3 7.5 - 2 7.6 - 3 7.7 - 2 7.8 - 3 7.9 - 1 8.0 2 2 8.1 - - 8.2 - 1 8.3 - 2 8.4 - 2 8.5 - 1 8.6 - - 8.7 - 1 8.8 - - 8.9 - - 9.0 1 - 9.1 - 1 9.2 1 - 9.3 - 1 9.4 - - 9.5 1 9.6 - - 9.7 - - 9.8 - 1 10. 1 1 - 10.6 1 - 11.2 1 - Date Maximum Mean Minimum 8/8/39 11.2 6.0 8/20/39 9.8 8.0 7.0 Measurements in cm. total lengths. 8/8/39 U.7 7.7 5.3 8/20/39 12.7 10.3 8.9 Table VII. Fully Positive Records of Tarpon Growth. All measurements in cm. total length. Initial Initial Final Elapsed Incre- Growth Tag date length length days ment rate* number 7/11/38 34.5 39.0 258 4.5 18.5 15038 7/20/38 37.0 38.0 167 1.0 5.9 15136 8/ 4/38 36.0 36.0 152 0.0 0.0 15137 8/23/38 35.5 35.5 133 0.0 0.0 15129 8/23/38 36.5 36.5 133 0.0 0.0 15124 The first listed in a small pond two miles north i of St. James City, Pine Island, “A’: ’ in text-figure 2. The rest in a pond three and one-quarter miles north-west of Point Ybel, Sanibel Island, “B’ ' in text-figure 2. 3/26/39 24.0 33.0 472 9.0 29.0 13420 3/26/39 28.0 29.3 472 4.2 11.6 13421 3/25/39 23.0 33.6 427 10.6 39.4 13122 3/30/39 35.0 52.9 468 17.9 39.9 13424 12/31/41 43.6 46.5 183 2.9 13.3 — The above released in the laboratory pool on Palmetto Key on initial date. laboratory specimens At the New York Aquarium Aquarium no. 9/28/39 9.8 11.7 314 1.9 22.5 1 9/28/39 11.0 12.7 314 1.7 17.9 1 9/28/39 12.0 15.7 314 3.7 35.8 1 9/28/39 14.5 17.5 314 3.0 24.1 l 9/28/39 9.4 11.7 222 2.3 40.2 2 9/28/39 10.6 13.0 314 2.4 26.3 2 9/28/39 11.4 13.5 113 2.1 59.6 2 9/28/39 13.2 16.7 314 3.5 30.8 2 9/28/39 9.5 11.0 314 1.5 18.4 3 9/28/39 10. 1 11.5 222 1.4 22.8 3 9/28/39 10.2 12.7 314 2.5 28.5 3 9/28/39 10.8 12.5 314 1.7 18.3 3 The above kept in aquaria of standing sea water, 24" x 12" x 12". Specific gravity 1.024. Temperature 22 to 24° C. Water 10" deep. comparative growth rates of tarpon Based on the above fully pos'tive records. Location Number of fish Mean initial length Mean growth rate Pine Island “A" 1 34.5 18.5 Sanibel Island “B” 4 36.2 1.5 Laboratory pool 5 30.7 26.6 Aquarium 1 4 11.8 25. 1 Aquarium 2 4 11.2 39.2 Aquarium 3 4 10.2 22.0 All Aquaria (1, 2 & 3) 12 11. 1 28.8 Sanibel (newly arrived) (See Table VI) 7.7 616.2 1 Rate calculated as % of increase in total length for one year, according to the following formula : I *36500 T L = Initial length I = Increment T = Elapsed days X = Calculated rate of increase 1944] Breder: Materials for Study of Tarpon atlanticus 231 CHECK NUMBER Text-fig. 3. Length of tarpon at the time of development of checks on the scales (annulae?). Based on imprints in celluloid of large scales. The numerical data are given in Table XII. The limits of variation are indicated by the dotted lines marked “Maximum” and “Minimum.” Twice the standard error of the means, in both plus and minus directions, is indicated by vertical lines in classes of four or more fish. The number of specimens in each class are indicated below the curve. The values of the mode are indicated by a heavier line than that used for the means down to a lower limit of thirty-six fish below which clear modes could not be established. rate but still comparable to that of the smaller fish kept in small aquaria and in the laboratory pool. These specimens kept on exhibition at the old New York Aqua- rium grew from less than 50 cm. up to 122 cm. in a period of five years, see Table XIV. These fish were taken near Key West, Florida. More exact measurements were out of the question because of the demands of exhibition. The accumulated data of others on the occurrence of small tarpon are given in Table XI. Included with the data of Storey and Perry (1933) is the statement that young tarpon were taken April, 1933, of 8.46 and 12.70 cm. These are not much larger than those we took from another place on Sanibel Island in August at the time of their entry. It may be that these fish had been there a year, as they suggested, but with the demonstrated lack of growth that some- times occurs at that place it is possible that they are much older. Scale Analysis. The useful method of counting the checks on the scales of fishes as an index of years passed has never been critically investigated in subtropical marine species, although they show markings that could be so interpreted if a sound basis for such a view could be established. In an effort to cast some light on the nature of the checks on tarpon scales, both experimental study and statistical analysis were applied in order to determine if there is any regularity and any statistical significance assignable to these marks. Certain specimens were taken in land- locked pools, tagged and recaptured at a later date. The data so obtained are given in 232 Zoologica: New York Zoological Societg [XXIX: 19 Table VIII. Checks on the Scales of Tarpon on Which There are Fully Positive Records of Growth. All measurements in cm. total length. Scale edges in italics between columns of check calculations. Tag Date of scale CALCULATED LENGTH AT CHECK number: number removal 1 2 3 4 5 15038 7/11/38 15.0 22.4 27.5 34.5 3/26/39 11.3 21.8 31.8 36.0 39.0 Mean 13.1 22. 1 29.6 36.0 15136 7/20/38 13. 1 19.4 26.2 33.6 37. u 8/23/38 9.8 16.9 28.4 32. 5 37.0 1/30/39 8.9 17.5 22.8 31.2 35.0 Mean 10.6 17.9 25.8 32.4 35. 0 15137 8/ 4/38 20. 1 30.7 33.4 36. O 1/30/39 16.8 26.9 33.5 36. 0 Mean 18.5 28. S 33.5 15129 8/23/38 15. 1 23.2 25. 9 29.8 32.0 1/30/39 12.3 21.9 25.4 29.8 32.5 Mean 13.7 22.5 25.6 29.8 32.2 15124 8/23/38 11.5 20.4 30. 1 33.7 36. 5 1/30/39 11.6 20. 1 29.3 32. 9 36.5 Mean 11.5 20.2 29. 7 33. 3 L. pool 12/31/41 18.8 31.0 43.6 7/ 1/42 18.4 29. 5 41.8 46. 5 Mean IS. 6 30. 2 41.8 Table VII. Compared with these data in the same table is similar information on speci- mens held in aquaria. The markings on the scales of these fishes are given in Tables VIII and IX. Measurements of the mark- ings on the scales of large fish are given in Table XII as proportional parts of the total length of the fish. The total length has been used for this study of the scale mark- ings, as it was found impractical to attempt to induce anglers and others that were kind enough to help, to take measurements in standard lengths. For purposes of conver- sion Table I, showing the relationship be- tween the total and the standard length, should be consulted. The values given in Table XII, plotted out according to the number of recognizable checks, are shown in Text-figure 3 on the assumption that they represent some natural regularity in time. The appearance of the mean, maximum and minimum values pre- sent a not unreasonable growth curve. Fol- lowing the methods of Westman and Fahy (1940) and Westman and Gilbert (1941), twice the standard error of the mean is indicated by verticals in both directions. This gives a measure of the significance attached to these values. Large irregulari- ties are no doubt accounted for by the fact that it was impossible to determine the sex in most cases and undoubtedly some of the checks measured are adventitious rings which would make the fish appear older than it was in fact. The excess of the maximum values over the minimum as referred to the mean may be a measure of such sources of error. As a further expression of the extent of the regularity of these check marks, Text-figure 4 shows histograms of each of the checks of the fishes given in Table XII. The values of these modes is shown in Text-figure 3 comparatively with the mean values and clearly shows substantial agree- ment. There is thus clearly some distin- guishable regularity in time of the forma- tion of these checks. Furthermore they show a regular decrease in spacing with increase in number, which in other and northern fishes is taken to represent the effects of the slowing down of length in- crement with age in an environment which is responsible for the formation of such checks at uniform intervals of time. The question that then remains is whether this effect on the tarpon scale is in the nature of an annual event or not. This calls for an analysis of changes in the tarpon en- vironment of an annual or other nature. Most students have hesitated to assign the term annulae to marks on the scales of the more southern species on the basis that the annual change in temperature of the sea or other influences was perhaps insuf- ficient to cause the formation of such checks. Insofar as the fishes of the Florida west coast are concerned, this would seem to be an unwarranted assumption for here in these coastal waters there is a considerable seasonal temperature change. In fact, peri- odically it drops to lows sufficient to kill large numbers of fishes. See for example Willcox (1887), Brown (1905), Finch (1917), Storey and Gudger (1936), Bab- cock (1936), Storey (1937 ). See also Taylor (1917) who discusses other causes of sud- den mortality in this region. Gunter (1941) discusses low temperatures on the Texas 1944] Breder: Materials for Study of Tarpon atlanticus 233 Z ■ I I I I I I I 1 I I I I I ll I I □ I I I I II II I II I I I I I I I I I I I I I I I I I M I I I 1 1 I 30 60 90 120 60 90 120 150 180 CALCULATED TOTAL LENGTH AT EACH CHECK Text-fig. 4. Histograms of checks on scales given in Table XII and Text-figure 3, indicating modes, distribution and spread with increasing number. Each class covers 5 cm., the upper limit of which is indicated. The left index indicates units of one fish. coast bearing on similar matters. Tem- perature records made at the end of the laboratory dock show the following dif- ferences between summer and winter. Temperatures at Laboratory Dock in °C. Max. Mean Min. June 4 to July 11, 1940, 41, 42 (148 readings) . . 33.7 30.8 26.5 March 5 to 31, 1941 (26 readings) 22.2 19.5 15.0 It should be evident from the above tab- ulation alone that there is a considerable temperature change with the seasons at this place. This is even more marked in back- waters and streams th^it are much frje- quented by tarpon. A glance at Text-figure 8 gives some idea of the access of these estuarine waters to the open Gulf at the laboratory site. In addition to simple tem- perature change there is a large change in the fauna with the seasons. Many fishes axe absent or nearly so in winter or summer and the invertebrate faunal change is most striking. It would seem likely that there is a consequently large dietary change with the season, which might influence the whole physiology of the fish. Other factors that must be reckoned with are those of movements of the fish them- selves. There is a considerable movement in and out of streams by these fishes. If there is any regularity to it we have not been able to determine such. At this writing it appears to be entirely sporadic. Local lore is to the effect that tarpon that have been recently in fresh water have a slightly “golden” tinge and that those which have not are “pure silver.” Two such color phases are readily distinguished at sight and check 234 Zoologica: Neiv York Zoological Society [XXIX: 19 Text-fig. 5. Growth rates of fully positive records of individual tarpon under various conditions compared with the indicated growth rate of large fish according to I — • 36500 the formula T based on the data of Tables VI, VII and X. Solid = X L lines, rates of fully positive records. Dotted lines, rates of large fish based on scale readings. was kept to determine if one showed a predominance at any season. Nothing of such a nature was noted and the two kinds appeared in roughly equal numbers. As many small specimens were kept in aquaria, both fresh and salt, credence is given to this fisherman’s belief because those kept in fresh water took on a “brassy” color while those in salt water remained “silvery white.” Furthermore it is a common occurrence for many fishes, in passing from salt to fresh water, to lose even more of their bright silvery color and to regain it on returning to the sea, e.g. Petromyzon and Salwo. This is evidently some physiological matter, per- haps associated with calcium metabolism. If this movement in and out of fresh or nearly fresh water is reflected in the scales, we have no indication of it. Small tarpon can easily withstand a transfer from straight fresh water to sea water of more than normal density or vice versa. Ex- perimentally they did not throw down a check after such treatment. Migrations proper, which the tarpon are supposed to make but of which we have as yet no good experimental evidence, would hardly be expected to produce other than “winter rings” in season. The general hypothesis is that they move northward along the Florida west coast and down the Texas coast and back across open waters. Even if this is true they would be exposed annually to a marked temperature differen- tial. There is thus established the fact that tarpon are exposed to annual fluctuations of temperature of considerable sort, and they cannot be thought of as living continually in an environment of uniform temperature. The above statements refer to tarpon in open waters. Obviously those confined to small land-locked pools suffer even greater extremes of temperature. 1944] Breder: Materials for Study of Tarpon atlanticus 235 In order to attempt to determine if it were justifiable to consider these markings as winter rings, scales of various marked fish were examined at various periods. The results in calculated length at the time the various checks were formed are shown in Table VIII. It will be noted that in each case where a fish passed through a winter, an additional check was formed except in the case of the three fish from Sanibel Is- land which showed no growth at all and therefore could not form a new check. This table also gives a measure of the extent of accuracy to be expected from measurements and calculations of this kind from tarpon scales. The measurements made on different scales from the same fish, taken at different times, show for most part rather close agreement. The largest deviation from the mean value is in no case greater than 24% and in most cases under 5%. Growth rates of these measured fish are shown comparatively with similar data from Table XII. Further data of a similar nature on large numbers of small fish are given in Table XIV and the evident growth of these various groups of land-locked speci- mens is expressed graphically in Text-figure 6. It will be noted that the growth charac- teristics vary widely from pool to pool and agree with those from which two scale samples had been taken. Compare especially the Pine Island with the Sanibel fishes. Finally there have been the fishes kept for several years at the old New York Aqua- rium. These were held under practically uniform conditions as to temperature and food, in salt water. The data on their growth and scale markings are given in Table XIV and shown graphically in Text-figure 7, compared with the mean values of scales of large fish, taken from Text-figure 3. From this it is obvious that they were growing at a rate comparable with those in the sea. These fish were taken in a body of water open to the sea near Key West and presum- ably had not been stunted. Two of these, the upper two in Text-figure 7, passed 5 and 4 winters at the Aquarium respectively and showed one check mark for each winter passed. The third, lowest in Text-figure 7, and consequently the smallest, showed 6 checks as against 4 winters passed in the Aquarium. If it is assumed that all had no checks before capture, then it must follow that the checks agree with the number of winters passed through except in the case of one fish that in some way developed two “adventitious” rings. In this connection, when these tarpon arrived at the Aquarium, one died during handling. Its scales were examined for rings and none were found, Breder (1937). Since these markings and the sizes of the fish agree so well with the marks on the scales of large fish, it is dif- ficult not to believe that these marks are in truth winter rings in the ordinary sense of the word. Since, however, the formation of Table IX. Detailed Record of Growth of Certain Individual Specimens of Tarpon. All measurements in cm. total length aquarium specimens (See Table VII) Initial Second Third date date date First Second Growth Growth 9/28/39 113 days 201 days increment increment rate (1) rate (2) 9.8 10.5 11.7 0.7 1.2 23.1 20.8 11.0 11.5 12.7 0.5 1.2 14.7 18.9 12.0 14.0 15.7 2.0 1.7 53.8 22. 1 14.5 16.0 17.5 1.5 1.5 33.4 17.0 9.4 11.0 11.7 1 1.6 0.7 55.0 21.3 10.6 11.3 13.0 0.7 1.7 21.3 27.3 11.4 13.5 — 2. 1 — 59.6 — 13.2 15.5 16.7 2.3 1.2 56.3 14.1 9.5 10.5 11.0 1.0 0.5 34.0 8.6 10.1 11.2 11.5 1 1. 1 0.3 35.2 8.9 10.2 11.5 12.7 1.3 1.2 31.4 18.9 10.8 11.5 12.5 0.7 1.0 20.9 15.8 Maximum 59.6 27. S Mean 36.6 17.6 Minimum 14-7 8.6 OCCURRENCE OF CHECKS ON SCALES OF ABOVE LISTED FISHES Date Number of Number of Total lengths at check number: fish checks 1 2 0 1 2 ' Max. Mean Min. Max. Mean Min. First 12 7 5 0 12.4 8.9 5.1 — — — Second 11 0 11 0 13.4 9.8 9.1 — _ — Third 10 0 8 1 11.8 9.5 8.4 13.5 — 1 Figures marked with footnote 1 in third column represent 109 days instead of 201. 236 Zoologica: New York Zoological Society [XXIX: 19 Table X. Size Groups op Tagge,d Tarpon by Locality and Date. Measurements in cm. total length. Modes in 5 cm. intervals — the upper limit indicated. LOCALITY' Pine Island, one to two miles north of St. James City (ditch) Sanibel Island, 1 mile s. of Wulfert (pond) Sanibel Island, Y2 mile s.w. of Point Ybel (ditch) Sanibel Island, 334 miles s.w. of Point Ybel (pond) Boca Grande Island, Railroad ditch 2 miles north of city Captiva Island near Redfish Pass (pond) Boca Grande Pass DATE NO. OF FISH Max. LENGTHS Mean IN CM. Mode Min. 7/ 6/38 32 51.5 34.4 35 27.5 7/11/38 7 77.4 61.5 65 34.5 2/26/39 4 61.0 34.0 — 23.0 3/ 7/39 3 23.5 25.7 — 27.0 8/ 9/38 2 66.0 65.5 — 65.0 8/ 2/38 1 35.5 _ 8/ 4/38 47 48.0 37. 1 35 31.0 8/ 9/38 37 66.5 39.8 35 32.0 8/25/38 45 49.0 36.5 35 31.0 3/30/39 5 37.5 35.9 — 35.0 7/20/38 36 42.5 35. 1 40 32.0 8/ 2/38 54 42.0 35.9 40 32.0 8/ 4/38 24 38.5 35.8 40 33.0 8/23/38 46 41.0 36.0 40 31.5 8/24/38 22 38.5 35.9 40 32.0 1/30/39 10 38.0 35.6 — 33.5 6/28/38 1 — 37.5 — 6/30/38 2 66.5 56.7 — 47.0 7/ 4/38 1 — 58.0 — — 8/18/38 1 — 40.0 — — 7/ 6/38 3 167. 6 144.8 ■ | 127.0 7/ 7/38 2 187.9 153. 6 — 119.4 7/ 9/38 1 — 152.5 — — 3/23/39 2 130. 9 122.4 — 113.9 3/24/39 6 122.0 114.3 — 106.9 3/28/39 1 — 114.3 — — 4/ 4/39 1 — 101.7 — — 4/ 9/39 2 167.6 140.9 — 114.3 4/16/39 1 — 107.8 — — 4/20/39 1 — 137.2 — — 4/21/39 2 132.2 104.2 — 76.3 4/22/39 1 — 165.2 — — 4/25/39 1 — 183.0 ’ — — 4/26/39 1 — 107.8 — — 4/28/39 2 183.0 152.4 — 121.9 4/30/39 3 198.1 183.4 — 156. 1 5/ 1/39 2 152.6 137.3 — 122.0 5/ 2/39 1 — 122.0 — — 5/ 3/39 1 — 137.2 — — 5/ 4/39 1 — 152.6 — — 5/ 5/39 2 183.0 180.4 — 177.8 5/ 6/39 2 198. 1 152.5 — 106.9 5/13/39 1 — 137.2 — — 5/14/39 1 — 152.5 — — 5/18/39 1 — 152. 6 — — 5/20/39 1 — 127.0 — — 5/22/39 2 167.8 155.0 — 142.2 5/23/39 2 198. 1 — — 183.0 5/25/39 1 — 183.0 — — 5/28/39 1 — 175.2 — — 6/ 2/39 1 — 193.2 — — 6/ 3/39 1 — 122.0 — — 6/ 5/39 1 — 122.0 — — 7/28/39 1 — 152.6 — — 8/ 7/39 2 177.8 169.9 — 122.0 8/11/39 1 — 167.8 — — 1944] Breder: Materials for Study of Tarpon atlanticus 237 Table X. (Continued) — Size Groups of Tagged Tarpon by Locality and Date. LOCALITY DATE NO. OF FISH Max. LENGTHS Mean IN CM. Mode Min. Captiva Pass 7/12/38 1 — 127.0 — — 7/13/38 2 187.1 129.5 — 121.9 7/14/38 1 — 119.5 — — 3/26/39 3 187.2 126.3 — 109.3 4/ 9/39 1 — 139.6 — — 5/ 1/39 2 195.7 174. 1 — 152.6 8/ 6/39 2 137.2 129.6 — 122.0 8/ 9/39 1 — 137.2 — — 8/10/39 1 — 101.8 — — Redfisli Pass 5/15/39 1 — 101.7 — — Off Palmetto Key 7/10/39 1 — 137.5 — — these rings took place under the abnormally uniform conditions of an aquarium, some source other than some obscure physio- logical regularity may be sought. The only variable entering into this situation that was not controlled was that of hours of day- light, which is evidently the only item that could account for this situation. As this feature of environment is well established as having similar effects, it may be assumed, subject to further experimentation, that this accounts for the check formation. The fact that electric lights were lighted over the tanks on dark days would be presumably altogether too insignificant to be expected to have any measurable effect. Text-figure 7 also shows the growth of the smallest wild fish of which there is any record; the data of Table VI. This fits nicely with the rest, especially when it is borne in mind that this growth is probably minimal, due to the fact that these fish were newly trapped on Sanibel Island where it is known that growth is slow or even absent. These data, compared with those already discussed on long term land-locked fish, show clearly that fish from open waters grow at relatively greatly accelerated rates and that there is little in common with the check formations on the scales from the two types of environment. From this it follows that the land-locked forms have little to do with the stock of maturing fishes. No scales from larger fishes have been found that show checks at the small sizes as those from land-locked places. The latter must there- fore be looked upon as a peripheral wastage not taking part in the main stock, evidently few escaping from such places to grow up to be large fish, which is in agreement with the opinions of Babcock (1936). Another approach to an attempt to under- Table XI. Accumulated Records of Small Tarpon. Authority Sizes in CM.1 Locality Date Evermann" & Marsh (1902) 19-29.1 s.l.? Huacares, Porto Rico February Eigenmann (1902 & 1921 ) 1 2-19.2 s.l.? Pinar del Rio, Cuba March Coker (1921a) 25 s.l.? Dauphin Island, Alabama January Coker (1921b) 5. 7-6.0 s.l. Fajardo, Porto Rico Beebe (1927 & 1928) & 5.1-20.3 s. 1. Source Matelas, Haiti January 13 Beebe & Tee-Van (1928) 7.6-33.1 s.l. Source Matelas, Hait January 23 11.4-17.8 s.l. Source Matelas, Haiti March 21 Breder (1933) 15.5-28.0 s.l. Andros Island, Bahamas February Storey & Perry (1933) 12-38 s.l.? Sanibel Island, Florida “Any time” 8.5-17.7 s.l.? Sanibel Island, Florida April Babcock (1936) 3 27.5-47.9 t.l.? Boca Grande, Florida 20.3-35 t.l.? Shark River, Florida February 1-4 lbs. St. James City, Florida small Grenada, Jamaica 7.6 t.l.? Aransas Pass, Texas 25.4 t.l.? Cristobal, Canal Zone Breder (1937) circa 15.2-22.8 s.l. Key West, Florida June Breder (1939c)4 Upwards from less than 6.0 s.l. Sanibel Island, Florida August Original5 31.1-32.4 s.l. Chagres River, Canal Zone April ^Various measures converted to cm. standard length where practicable. 2The smallest, 2 cm., has never been located as a preserved specimen and there is some question as to whether this is a misprint. See Babcock (1936) for a discussion of this specimen. "This author quotes and discusses many of these records and those of larger fish. ^Actually 4.2 to 11.2 s.l. See Table VI. 5From data and photographs obtained from Mr. William Markham. Evidently the same location as those of Bab- cock’s Canal Zone records. 238 Zoologica: New York Zoological Society [XXIX: 19 £ c Q ifi < cc a £ O £ H GO ^ * 2 1 <1 | H Q a W f3 ■oc co Oi g CO GO II I I II M I ! 03 | iO 03 1 1 ° 1 1 1 Ill'll r i 03* CO* cd 03* GO t>- CO o id >H r-H r-H 03 >~H cDnOGO | OiOO | | | lO | | O GO 00* 0 3 03* cd 03* CD* GO* cd O* rdo3* 03 Go CD 03 COQ»G ^ Go 03 | ©3 | C5 Go o | | lo Oi 00 CD 05 CO IQ | T-H o GO | os c> co m id d GO >H CD >h CO r-H >H 03 r-H f-H ^ >*H id I'r- 03 *d Co >h r-H 03 o GO 03 -h 00 03 03 *d> r-H r-H rH >H >~s co id o 03 03 CO H-S >-H t-H t-H iO id 03 03 03 Co 03 G© lO I icrj^o N Tfi Oi CO CO CO N H H O CO o " O | CO tO O »0 Oj CO N O 03 O O 03 03 03* CD oi rd o’ O »h OIONO "d id rH >d CO* CD 03* 03* cd 00* 03* id 03* o © O* "d-CO cd cd 03 03 030 00030 ^222^22203^22 £ o <1 p o r. J cj c o O £ O o A 5 ° 00 00 O lO lO ^ Go (NNOiOCi^ ’ ' — i ' — 1 03 ' — i ’’-h O GiQ0N00:C0NHC0HO030lNH^iO03iOc0O03C0OOiCOi0O050003Oi000O(N^>Ci o’ cd o4 oi id cd tjh* o’ o* oo’ go o’ id o cd ^* t-h od o cd cd ^h* cd i> id r^* cd cd cd o tjh id id ^* cd cd fOOhOOOhC3 00000^^0C00 03NiOO^i0 03 01^000^NhOOO(N03 03 T-H "t-H r-H t-H r— I t-m r— ( t-H >“H t— I H t-H i— ( rH r— I t-H rH rH t-H t— I rH rH i-H t— t t— I r-H r— I t-H t— I cOiON '>O^IOCDNOOOOM‘OO^NOCON03- •O 000303^^0^ 03 © © t-h 03 O ^NcOOOHCOcOrfCOOHOOiOOO^COOOiOOOOOOON ^hN05OO03O00 00NO00hC0 00NC5hOhO03 03O 030i0 003003000303000300rHOOOI> tJH id o co o oo o io co nioiooono>oohioocoojooonoiolocdhoooncdcoo3ocooiohcd^ooqcoio H •J PC < lO 03 lO O CO ^ T C003iOOOWOOOO^hcDOOONOhiOCOOO^COhOCOCONC0003>OiOCOOOOQh^ O3COO3O3NiO^iOiONNC5COONiONa2OOO3rHCO0O3OOOOO5OOCOrHCOO3OO3iOCOCOCOa> NOiONNNNiOiOiGiONiONOiO^iOOiONiOOOOCiNOiCOCiOOOONOCOOOOO^D CO iO 0 CO lO CO iO Ot}UO^O5ONNNNiONO3OO0hOCOOhOO3OOCOO3OO^NNCOOlOHt1CoOh0^oO o oi 6 d h d o dNdddcdcdidididddrHcdoidddoood^ id d o id n co d h o cd n cd oi ^ d d cd ^!O^00iO*t ^TtHr^^HuOTjHCDTlH^rtH'TtHi>.TtiiOI>^CO^iOCOLOCOiOOOO^'-^^iOOOOO^>-O^^HOiO^ 0001003030 '^iO(M^^G0O03r-(i000G0iOO00’^00^030003t^G0T^0300^03^t>t^C0^00O^Oi0^ o6^j>o6o3idoocdocdocdcii-Hcdi>Or-Hcd'do6idcdo3*o6o3*o6ido3*o^i>’oa50C5cdid 03 03 03 03r^(03--HC0C0C003O03--H^03 03 03i00303r-^03rt<^C^03^(Mr}H^O0303^Hl>C00303 "3 00 00 GO 00 00 OO GO 030303C303C303C3C3C30303C30303C3C30303C3C303030303C303C3C303C30303C303C3C303C3 vg CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CD CD rH CO CO CO CO ^ CD CD CD 00 CO rJH 03 Ci o CO lO 00 00 rH r-H (M 03 CO ^ lO IQ CO TtH 00 00 03 CO ID — 3 rH r-H r-H rH 03 03 03 03 03 03 03 030303 03 03 03 03 03 03 03 03 ^ ^ hhhh 03 03 03 COCOCOCOCOCOCOCOCOCOCO^^^^^^rtHTf^rfTt-TfiOiOiOiOiOiOiOiOLOiOiOiOiOiOiOiO 5/28/39 31.8 47.7 66.5 124.8 154.0 175.2 239 1944] b- b oo ^ CM iO CD CD Breder: Materials for Study of Tarpon atlanticus co b o ^ 1 l | lO d r-H »-Q> rH 1 b ! O ^ | tOO | b» o CO b- CM co b- rtf i— i b r-H r— 1 r-4 00 | lOH | 1 1 1 1 1 CM d- b- CM CM CD rtf 1—1 b i — 1 r-H 1 1 1 1 1 II 1 1 00 05 00 *0 d io 2 oi io m CO Co n GO CO b 1* r-H r— 1 1—1 1 1 00 1 1 I I ! 1 M oo 0> ^jGO N b ’"H 1 1 ° 1 ! 1^11° ClNiOO oq o Oo © o k c4 © «o r liS +IXN iO GO (N r-H b i-H b r-H r-H 1 | M | [ t- | GO 00 O ‘O 05 00 b CM b CD* d r-H r-i rH 05 bp "tl 05 r-H r-H lo CM ^T-rH r-H t-i r-H >H i-H >H r-H rH | | OOt^ | | CO Go C3 6)9OC0 r— * 00 b- !> CD GO -h ID O* 2 rH CM CO £ 'll 05 O CM C CO rtf hNcOO i-H i-H r-H r—| r-H r-H b i-H r-H & 3 0 1 1 r**00 1 | CO O CO io ^ (NO IO 05 & 0 CO rtf CD r-H rtf Co Yi tD rJH CD 05 ^ -Tib- 00 CO a CM GO O lb CM GO 0 rH r-H r-H b r-H s rS hGOOO | MOrttoin OjiqOiO ih CD d CM LQ io (M H CO d-^td co* cd ® > — 1 Tl 05 b— rtf rtf ^ r-H toOKMON 0> i — 1 b I— 1 b r-H r-H b r-H 0 CM CM to b TH CM tb CO 05 <30 O 40 05 b r-H CM CD Oo H CO rtf CM CO -d-7j CO 2 CS © ID 0 COCODCDn ICCO OD CD d i-h r-H d . . d HOO^rHN rtf lO 05 IO IO 05 cc GO b- b 44 0 o cd co d th N r-H r-H N N GIloolooiK Oo n1 lO to to HOOOIOO oo ib ib b- io « Pi lOiOCM^OOi CD to O 05 O 05 0 C5 t-4 _ S3 CD CO ' — i 00 CO O b- GO rtf O 05 jj CO* O 05 g Co Tl CO rtf to S't! b- oo b- -f to rtf rtf ^D rtf n O r-H d t-) •<- HNOGON CM tb r — 1 rtf 05 Co®Hb *0 d a) 40 CD CM rHNdoco "i jj-- 1 1- C3 g ii i-h CM to ■-* QJ d^i to CD rtf CO rtf CM CO to CM ttfHH 05 00 CD rtf LO CD 00 rtf rtf CO N = t|IO | »« r~H b» b* b» rtf O lb 00 rH bl ■b n 05 CD 05 <20 *1b* 1— 1 to 73^5 <15 03 CM CM CO CM CM CM i-H CM CO r-H o>Eh ^.d 'go 53 43 « bB 0 1 1 g || r- 5 H “ S '3 § O O O O O r-H i-H i-H i-H i-H rtf rtf tfH rtf rtf CD lb b- 00 rtf ^ ^ rtf -tf rtf 05 05 05 CM CD m h Q diH r CH C3 H 0’S X'S HO CM CM CM CM CM CM cu p a .s 00 CD CD CD CD P— CD CD CD CD CD Table XIII. Size of Large Tarpon Compared With Number of Checks on Scales. Number Number of Total lengths in cm. of fish checks 1 Max. Mean Min. 2 4 118. 1 115.9 113.8 11 5 175.2 132.5 ±7.3 92.8 10 6 204.0 135. 0 + 9.8 101.8 13 7 185.5 139.7 +6.4 111.8 12 8 193.0 156. 0 + 6.5 121.9 5 9 204.0 158.0 + 10.9 137.2 1 10 — 158. 8 ■ — 1 11 — 182. 9 — 2 12 198. 1 190. 1 182.1 1 14 — 185.4 — 1 15 — 161.3 — 1 Exclusive of the scale edge. stand the checks on tarpon scales is to com- pare the total number of checks on a scale with the length of the fish from which taken. These data are given in Table XIII for the fish listed in Table XII. They are compared with data of other smaller fishes from various places in Text-figure 8. Con- sidering the large fish as a random angler’s sample, it is clear that the bulk of the fishes caught have 7 checks and evidently have passed through that many winters. Since at this age they show a mean length of 139.7 cm. and we have obtained sexually immature fish as large as 142.2 cm. and the smallest ripe fish at 139.8 cm., it is evident that the time of greatest hazard, as far as capture by angling is concerned, is just at about the time of first spawning. Further, just about as many are caught below this age as above it, i.e. 36 of this size or less and 33 above it, in Table XIII. Comparison with the small land-locked specimens shows the same kind of separation already noted in the earlier treatment, while the aquarium specimens fair within or close to the range of the large fish. Movements of Tarpon. The actual data on which this analysis is based are expressed in various forms in Tables XV and XVI and are summarized in Table XVII. Much of the tagging was of an experimental nature, undertaken to determine the suitability of tags, the nature of growth in small land-locked ponds and similar matters. Such fish were not available to the general angling and fishing public and consequently no returns could be ex- pected from these, which numbered 226 tagged fish. Those which were available to the public numbered 254 tagged fish. Of these 177 were small land-locked specimens released in the open sea or bays and 77 were large “angler’s” fish, released where caught ex- cept for 3 which were taken from Boca Grande Pass and subsequently released at Palmetto Key. Thirteen tags taken from fish Table XIV. Calculations from the Study of the Scales of Small Tarpon . No. of specimens in parenthesis. Standard errors after means. A. calculations of size at time checks were formed, based on SCALES OF SMALL LAND-LOCKED TARPON Calculated total lengths in cm. at check number: 240 Zoologica : New York Zoological Society [XXIX: 19 § TjH +1 o lO co O 00 Co Co <>0 o >X, o o Mir 1 >4 <20 GO Vf .. <20 CM lO ft 2 + io *3 P K* co co O • GO • -r^ CO 00 d- 05 O Tfl g >, A rt ® S - >-i to to °o o o^to^ o s o3 cd on ^ •• o go w <2o a> "i , P o g w a® » c3 - O ° ^+1 H o 5 Z 2 S3 % = E-t S j c S Sj a* a 2 co z “ >< O 9 03 W -4-^ CQ bl) ^ M P S'g S g o co +1 3 S faO p H 43’ N O 02 £h PP oo^-n os co co co cs ^ cs T— I r-H T3^. T3 P ^ P jS 02 §0 02 -H HH 0^0 ! .9 ^ .S ! 2 CM CM co CM o 00 >o r-H co CM O >*H fx< o 50 O i 1X5 °o Co Co o’ 1 d >X( (M 1 1 GO 1 d <50 >H >H >H ®4 co GO’ CO io io 1 UO o 1 ^ 'S 1 •• 1 CM 1 1 1 CM* CM 1 | 1 CO o CM 00 tn 1 d 1 CM CM CO CM CM CM CM CM CM cm a to CM co >H 00 Co vt- e» c O o n! 1 o oo oo d 1 xs t o3 g ! 1 Nt- GO GO GO GO o 1944] Breder: Materials for Study of Tarpon atlanticus 241 CO *n 1 ° 2 Mill II 1 1 II 1 1 1 1 1 U GO 1 Go N Go »n o 4-1 00 m CM in in CD 05 co co m in m » 1 GO GO Co ^ ^ t-H T—t CM V~-^ ^ ^ ' ^ O ‘n ‘n Co O lo GO* GO GO GO GO GO •S{- m I I 1 1 1 1 1 co CO m C5 So 03 +1 C2 +| r— H O n CO s°. "3 is ‘sg is .go go ^ ^ s O GO ° CD • lO • w £ GO r~v 0 . ■"3 ^ S CO ©» '3^5 -- c Go CO X3 d d 0 ^ o0 o +-3 d fi o d hP m CO ' CO 00 00 d Ph tT d d d a CO d o ^ O S « « -S ^ Jr? H 0 o ^ 1 GO § g d co H O O *rd JjtJ £ £ 9 £ o cn ^ 3 1 o _ ® 3 t» o rd o ++> * d s o d 0 o 6 IS o bfi d co s *> d f-i O D d hP m d > 02 d Q GO 00 00 GO \ COCOCOCO N •*tooco "2 2 CO CO CM d d ww c5 d CD CD CD *7n d o d T5 O S 8 « O C t-h £0 . O CM • o C5 m o o 1 co 1 1 1 1 *— i Co Go 1 >H *n CD CO *n Co V _d b£) d PH O CD CD O O O CM ^OHGi cm o ©£ co 2 ^ w ^4- I> CO Co Co CO ShcDN o" < < CO CD CM CD >nCONOOO0 Co CD lO in lO (M O Co ^ Go Co ^ go o °o co *n r}H ^-©0 GO Go ^ GO GO OMO^WQ 00ONICN00 rti >n ^t1 in rj^ r— : <4 5© IO CO GO W OT tfi a< a» a> 33D .£ £.5 D D "D c c c o o o QUO co ^ n go ci m ^ CD o CD m m o o cd ^ ^ o cd co ■rt* in go n* o co co co co m Tf< rH cwchmo in 05 CD (M 00 00 CM f-i CO CM 242 Zoologica: New York Zoological Society [XXIX: 19 Text-fig. 6. Length of land-locked tarpon at the time of the development of checks on the scales (annulae?). Based on the direct examination of scales. The numerical data are given in Table XIV. Notation as in Text-figure 3. A. Pine Island, 1938, and single fish, 1939. B. Sanibel Island, % mile s.w. Point Ybel, 1938. Bi. Same, 1939. C. Sanibel Island, 3 miles s.w. Point Ybel, 1938. C,. Same, 1939. D. Gasparilla Island, 1938. E. Laboratory pool, Palmetto Key, 1940. F. Sanibel Island, Wulfert, 1938. were returned by various fishermen to the Government offices by their finders, as shown in the schedule comprising Table XVI. Recoveries from open waters amount to 5-)-% while recoveries from land-locked places amounted to 4-%. Of the former l + % represents larger “angler’s” fish and 7-% represents small fish which were taken in land-locked places and released in open waters. The details of these data are given in Table XVII. Since the latest land-locked fish to be recovered had all cast their tags in some- thing over a year, it is to be presumed that the wild ones showed a complete loss in at least the time that has elapsed since then and at present there are no tarpon bearing our tags. A consideration of Table XVI shows that actually the last fish in open waters known to be still bearing its tag was taken 118 days after tagging. Two longer records refer to cast tags picked up on Sanibel beach, while a third is uncertain as to date. The greatest known length of time for a tag to be retained in a land- locked pool is 258 davs for tag no. 15038 in Table VII. While these data are many too few to base any final conclusions upon, as far as they go they would indicate that retention of either button or strap tags cannot be expected to exceed eight months. Also the loss of tags in pools and the open sea would evidently seem to be about equal in speed. There is no evidence of any unknown inter- communication between the open waters and the supposedly land-locked pools. Such con- nections are evidently irregularly intermit- tent due to storms or other changes, as generally supposed. Although most local people and anglers claim that small tarpon are extremely rare, a condition which we did not find borne out in the light of the material collected, we found that all but one of the returns pertained to relatively small tarpon of a size that most people disclaimed any knowledge of ever seeing. Obviously, however, the catchers of these fish must have been angling for other fishes, such as squeteague, for example, or using small seines. That a return of 7-% was obtained by such means would suggest on face value slight loss of tarpons in these size groups incidental to fishing operations 1944] Breder: Materials for Study of Tarpon atlanticus 243 Text-fig. 7. Comparison of known growth of three tarpon in aquaria and the small- est wild fish with the mean values given in Text-figure 3. Numerical data given in Tables VI, XII and XIV. Notation as in Text-figure 3. Dark circles represent measured lengths of fishes and are placed according to date on the assumption that the checks (open circles) represent winters. for other species. However, since there is such unanimity of opinion concerning the absence of these small sizes, our inability to find such in open waters after prolonged combing of the region would clearly in- dicate their substantial absence from places frequented by fishermen. It must be re- called that these tagged fish were removed from pools to open water and were thus made available to fishermen by our activities. The low return on the larger fish, 1 out of 77 tagged, would seem to indicate a very small take of the available stock, for here there is a vigorous and active angling fraternity that does catch larger numbers of tarpon in the course of a year, as evidence the records of the hotels at Useppa Island and Boca Grande alone. Consequently there is no indication from this of evidence that the present rate of catch of tarpon as a sportsman’s fish is in any way depleting the stock. Furthermore it must be borne in mind that a goodly proportion of the fishes are released without ever having been fully removed from the water. While the fate of the majority of these is uncertain, un- doubtedly a considerable number survive. Suggestive in this connection are the ex- perimental data obtained on laboratory fishes discussed by Shlaifer (1941). Two chief hazards would seem to appear. One, a physiological matter, includes the shock effect so marked on handling tarpon and the amount of success they have in re- gaining sufficient composure to take the necessary first breath after release. Since most of the tarpon are released in fairly open or deep water and we know the chief hazard to their continued survival at such times to be the proximity of soft flocculent mud, there should not be much loss from this source unless the continued “playing” of the fish or some unknown condition in the large adults or their environment comes into action, working toward their destruc- tion. The second factor involves the presence 244 Zoologica: New York Zoological Society [XXIX: 13 Text-fig. 8. Length of tarpon compared with number of checks (annulae?) on scales. From data of Table XIII and XIV. Notation for large fish as in Text-figure 2. The means only are given for the small fish. Fish from Aquarium indicated in solid circles. of predators in the form of sharks. In the later winter and early spring, sharks are not numerous in Boca Grande and Captiva Passes and probably cut no figure in the matter of any consequence. In the late spring and summer they become exceedingly numerous and voracious. In fact, in the summer it may become almost impossible to land a tarpon, once hooked, because as soon as one gives evidence of being in difficulty, shown by its peculiar swimming movements, the sharks close in, leaving the angler half a fish or only the head. This fact gives some indication of the size and capabilities of these sharks, some of which are easily able to bite through a six- or seven-foot tarpon at one stroke. This condition forced Mr. B’shop to abandon angling in August of both 1938 and 1939 for purposes of tagging. The fish generally become more numerous during that month and many could be hooked, but it was almost impossible to bring any to the boat intact. Consequently it is infer- red that early in the year the shark factor is practically negligible but later nearly completely neutralizes any useful effect that the release of hooked tarpon might have, for during the quarter or half hour period that the tarpon take to recover they would evidently be easy prey to any marauding shark, including those that might not have the courage to strike a hooked tarpon not very distant from a boat. Although no distinct migratory movement is apparent from these returns, an examina- tion of Text-figure 9 indicates that more were taken to the north of the place of release than to the south, actually 10 to the north and 2 to the south. Expressed in terms of northern distance in miles from the place of release, the captures stand as indicated in Table XVI. It is also evident from these calculations that the group as a whole showed little drift to the west — actually 9 to the west and 3 to the east. The averages of the distances north and west traveled, considering south and east as negative, show that there was a mean drift 1944] Breder : Material s for Study of Tarpon atlanticus 245 of 2.39 miles to the north and a mean drift of 0.29 miles to the west. These fish were all tagged in Juiy or August when local belief has it that the fish are moving north along the coast. The State of Louisiana, Mathes (1940) reported, started efforts to tag tarpon in that area. Monel strap tags were experimen- tally affixed to the dorsal fin. This method was earlier tried by us but not put into practice for two reasons. One was the rather- discouraging experiences in the past with other species and the other that since much of the tarpon fishing in the Charollette Harbor region is done at night and only the head of the fish brought out of water before releasing, it was felt that a more anterior position for the tags was desirable. There have evidently been no results re- ported so far from the Louisiana venture. In connection with the fate of the tags used by us, it was found that the strap tags used either on jaw, operculum, tail, dorsal fin or fleshy part of the back were generally unsatisfactory, tending to set up local ir- ritations, become loose and be shortly cast. The button tags secured to the operculum, two celluloid discs fastened by a pin of pure block tin, did not have such objections if placed well in from the edge. If close to the edge they would frequently tear out. Those that were retained by small land- locked fish would remain for the periods noted, and the external disc frequently became the focus of long, streaming, algae growths, which, however, did not seem to bother the fish in any way. Specimens held in the thoroughly controllable laboratory pool on Palmetto Key, after losing their tags, very quickly healed the perforation so that after a month or so no mark show- ing the former location of the tag could be found. It is not clear just in what manner the tags were lost. The tin pins seemed perfectly intact on all those recovered. If some corrosive action of environment and physiology is not involved, it is conceivable that the action involved might be entirely mechanical, as the periods of time would give ample opportunity for the exterior disc to become slightly loosened and catch on aquatic growths. The tags were placed firmly against the operculum but there was in no case any evidence of an overgrowth of tissues, the tags remaining completely free at all times. Discussion. The accumulated data on the absence of larval and postlarval tarpon from shore waters indicate that they presumably spend this part of their life history off-shore either in deep or surface waters. In order to determine this, elaborate tow-netting operations from an ocean-going craft would seem in order. Such a project was planned but the restrictions due to the war time conditions made prosecution impossible. It is hoped to undertake this program as soon as conditions make such activity possible. The tentative scale analysis indicates that the markings on tarpon probably do in- dicate numbers of winters passed and that those small specimens taken in land-locked pools are stunted and not part of the per- petuating population. Other estimates of age based on similar specimens have been given by Breder (1933), who found from 2 to 4 checks on land-locked specimens on Andros Island, Bahamas. Coker (1921a) reported no checks on a fish 25 cm. long from open waters in Alabama, and Gill (1907) thought fish of 5.7 to 8.5 cm. in Puerto Rico to be “probably young of the first year.” These cases are in keeping with our present find- ings. Babcock (1936) after much study gives what he calls “little more than a guess” as to the size of tarpon at various ages. Translated to metric measures his values stand as follows: Length in cm. Years 30.5— 35.6 1 50.8— 63.5 2 127.0—152.5 3 The growth indicated for the first and second year is in close agreement with our present data as indicated in Text-figures 3 and 7. Since Babcock’s “total length” is probably equivalent to our “length includ- ing jaw,” the agreement is even closer than is at first apparent. The third year is, however, much higher than our indications would call for. The lower estimate is within the present range of extremes but the upper falls without. It would seem that the growth between the second and third year is much too great in Babcock’s calculations, for it is more than three times as large as that which he indicates between the first and second. This is obviously a most unlikely condition in practically any growing animal. In another place he indicates that he believes that tarpon of 142.0 cm. may be in their ninth or tenth year, which again agrees closely with our data as indicated in the same Text-figures. This suggests that his figures for the third year are in the nature of some inadvertency of typography. In all these calculations of the possible age of tarpon, based on scale analysis, it seems to be taken for granted that they form a check the first winter in life. On this we obviously have no data. It has been shown that very small fish do form a check under land-locked conditions but we have no data as to the length of time that is spent as a leptocephalus. It may be a very brief period or a very extended one, covering- even a year or more. The discussion here 246 NO. OF Zoologicn: New York Zoological Society [XXIX: 19 Table XV. Tarpon Tagged, by Localities and Dates. Fish were released at site of capture unless otherwise noted. type of DATE FISH tag SERIAL NUMBERS OF TAGS FISH RELEASED AT FISH TAKEN IN PONDS AND DITCHES PINE ISLAND One to two miles north of St. James City. (Ditch). 7/ 6/38 32 Disc 15004 to 15035. San Carlos Bay. 7/11/38 7 Disc 15036 to 15039, 15070, 15071, 15073. — 2/26/39 4 Disc 13420 to 13423. — 3/ 7/39 3 Disc 13411 to 13413. SANIBEL ISLAND One-half mile south of Wulfert. (Pond). 8/ 9/38 2 Strap 10338, 10339. One-half mile south-west of Point Ybel. (Ditch). — 8/ 2/38 1 Strap 10202. — 8/ 4/38 47 Strap 10258 to 10267, 10269 to 10284, 10286 to 10288, 10290 to 10291, 10293 to 10308. — 8/ 9/38 37 Strap 10340 to 10342, 10344 to 10347, 10349 to 10351, 10353 to 10354, 10356 to 10357, 10359 to 10376, 10378 to 10380, 10382 to 10383. 8/25/38 20 Disc 15190 to 15195, 15212 to 15221, 15231 to 15234. — 25 Disc 15196 to 15211, 15222 to 15230. South shore of Sanibel. 3/30/39 5 Disc 13424 to 13428. Three and one-quarter miles south-west of Point Ybel. Laboratory pond, Palmetto. (Pond). 7/20/38 10 Disc 15078, 15080 to 15084, 15086, 15090, 15093, 15094. — 26 Disc 15095 to 15120. South shore of Sanibel. 8/ 2/38 54 Strap 10203, 10205 to 10257. South shore of Sanibel. 1 Strap 10250. — 8/ 4/38 7 Strap 10309, 10311, 10312, 10314, 10315, 10335 to 10336. 17 Strap 10316 to 10332. South shore of Sanibel. 8/23/38 46 Disc 15121 to 15166. — 8/24/38 22 Disc- 15167 to 15184, 15186 to 15189. South shore of Sanibel. 1/30/39 10 Disc 13400 to 13405, 13407 to 13410. GASPARILLA ISLAND Two miles north of Boca Grande. (Railroad ditch). 6/28/38 1 Disc 15000. — 6/30/38 2 Disc 15001, 15002. — 7/ 4/38 1 Disc 15003. CAPTIVA ISLAND Near Redfish Pass. (Pond). Charlotte harbor. 8/18/38 1 Strap 10384. — FISH TAKEN IN PASSES AND OUTSIDE WATERS BOCA GRANDE PASS 7/ 6/38 3 Disc 15050 to 15052. 7/ 7/38 2 Disc 15053, 15054. — 7/ 9/38 1 Disc 15055. — 3/23/39 2 Strap 10448, 10450. — 3/24/39 6 Strap 10401, 10431, 10436, 10440, 10411, 10406. — 3/28/39 1 Strap 10415. — 4/ 4/39 1 Strap 10460. — 4/ 9/39 2 Strap 10408, 10414. . - — 4/16/39 1 Strap 10509. — 4/20/39 1 Strap 10510. — 4/21/39 2 Strap 10513, 10413. — 4/22/39 1 Strap 10438. — 4/25/39 1 Strap 10512. - — 4/26/39 1 Strap 10439. — 4/28/39 2 Strap 10404, 10405. — 4/30/39 3 Strap 10519 to 10521. — 5/ 1/39 2 Strap 10534, 10535. — 5/ 2/39 1 Strap 10533. — 5/ 3/39 1 Strap 10488. — 5/ 4/39 1 Strap 10532. — 1944] Breder: Materials for Study of Tarpon atlanticus 247 Table XV (Continued) — -Tarpon Tagged, by Localities and Dates. DATE NO. OF FISH TYPE OF TAG SERIAL NUMBERS OF TAGS FISH RELEASE^) AT 5/ 5/39 2 Strap 10487, 10489. 5/ 6/39 2 Strap 10522, 10523. 5/13/39 1 Strap 10402. 5/14/39 1 Strap 10403. 5/18/39 1 Strap 10486. 5/20/39 1 Strap 10491. 5/22/39 2 Disc 13415, 13416. 5/23/39 2 Strap 10466, 10469. 5/25/39 1 Strap 10490. 5/28/39 1 Strap 10449. 6/ 2/39 1 Strap 10476. 6/ 3/39 2 Strap 10477, 10531. 6/ 5/39 1 Strap 10480. 7/28/39 1 Strap 10723. 8/ 7/39 2 Strap 10712, 10713. 8/11/39 1 Strap 10716. 6/18/41 1 Disc 46926. 7/ 3/41 3 Disc 46927 to 46929. Palmetto Key.' 7/12/38 1 Disc CAPTIVA PASS 15074. 7/13/38 2 Disc 15075, 15076. 7/14/38 1 Disc 15077. 3/26/39 3 Strap 10409, 10441, 10443. 4/ 9/39 1 Strap 10506. 5/ 1/39 2 Strap 10478, 10479. 8/ 6/39 2 Strap 10706, 10707. 8/ 9/39 1 Strap 10714. 8/10/39 1 Strap 10715 . 5/15/39 1 Strap RUDFISH pass 10451. 7/10/39 1 Strap PALMETTO KEY Off west shore. 10701. 1 These fish were collected by Mr. M. B. Bishop late in this date of release. They constituted some of the material May, and with others were held at the laboratory until discussed by Breder (1942e). as to age is made with this in mind, pending the discovery of the leptocephalus stage and an understanding of its duration. Babcock (1936) gives extended data on the seasons when “angler’s” tarpon appear at various places on the Gulf coast. This could be interpreted as a northward migra- tion with the coming of spring, or in various other fashions. The populations may be fairly static without any marked migratory behavior, simply keeping out of sight and not taking anglers’ offerings if the water be sufficiently cold. Throughout their range it seems that there are some present at all times and it has been shown that the res- piratory rises are less in cold water, as would be expected. This could easily give the impression in the field of comparative absence. On the other hand, the data on tagging give a suggestion that there may be a slight northward drift of the fishes, but these data are too few to be taken very seriously. This is another matter that cannot be re- sumed until world conditions become more nearly normal, for tagging at this time could not be expected to produce any reasonable number of returns with so few people angling. The assembled data indicate that tarpon may not reach sexual maturity until after passing their sixth or seventh winter. This may be a little too old, for there is still considerable uncertainty about the develop- ment of adventitious rings or perhaps spawning checks. In this connection it should be mentioned that in the larger fishes of about this size upward, there were in most cases vague ill-defined bands between the fairly sharp “winter” checks. These were evidently different in nature and placed about half way between the latter. They were not counted in these studies, but it is suspected that they may be spawning marks, which tends to support the general views here expressed. Since tarpon spawn at the warmest time of year and feed actively nearly or quite up to the time of spawning and are ravenous immediately thereafter, it may well account for the diffuse nature of these bands that have been tentatively considered as possible “spawning” checks. Although tarpon are found on the west coast of Africa, Fowler (1936) and Inter- national Game Fish Association (1943), 248 Zoological New York Zoological Society [XXIX: 19 1944] Breder: Materials for Study of Tarpon atlanticus 249 Text-fig. 9. Chart of region showing deployment of tagged specimens. Based on data of Table XVI. Serial tag numbers, capital letters, at place of release, small letters at place of recovery refer to that table. The connecting lines indicate the short water route. Palmetto Key is indicated in black. s Zoologica: New York Zoological Society [XXIX: 19 250 Zoologica: New York Zoological Society [XXIX: 19 Table XVI. Recovery of Tarpon, with Detailed Data on Each Fish. Measurements in cm. total length. Statute miles Miles from point of covered release3 Tag Tag Site of1 Recovery Place of1 Captor’s Elapsed3 Dis- Short no. date release Length date recovery data days tance route North West 15006 7/ 6/38 “A” 33. 5 10/10/38 “a”' 4 Yi lbs. 96 7.5 9.0 1.48 5.62 15008 7/ 6/38 “A” 27.5 9/ ?/38 “b” About 15” 56 2.0 2.0 -1.35 0.66 15015 7/ 6/38 “A” 34.0 11/ 1/38 “c” 5 lbs. 118 12.0 17.0 8.35 6.88 15016 7/ 6/38 “A” 38.0 9/19/38 “i” — 75 7.5 9.0 2.35 — 5. 62 15054 7/ 6/38 “E” 1193.8 9/17/38 “d” About 3H' 72 19.0 25.0 14.50 -2. 66 15097 7/20/38 “D” 39.0 3/ 5/41 “e” No fish 228 •? ? — — 15110 7/20/38 “D” 36.0 10/ ?/38 “f” — 88 2.5 12.0 1.09 1.09 15116 7/20/38 “D” 38.5 4/ 5/39 “e” No fish 254 ? ? — — 10207 8/ 2/38 “D” 38.0 3/ ?/42 “i” — 1326-7* 2.5 10.5 — — 10229 8/ 2/38 “D” 35.5 10/ ?/38 “f” — 75 2.5 12.0 1.09 1 . 09 10257 8/ 2/38 “D” 33.5 10/ ?/38 “f” — 75 2.5 12.0 1.09 1.09 10321 8/ 4/38 “D” 37.5 10/ ?/38 “f” — 73 2.5 12.0 1.09 1.09 10330 8/ 4/38 “D” 33.0 10/ ?/38 — 73 2.5 12.0 1.09 1.09 15167 8/24/38 “D” 36.0 9/24/38 “h” i w 31 11.0 18.0 3.50 3. 18 15202 8/25/38 “E” 49.0 10/ ?/38 “g” ay2 ibs. 63 16. 5 19.0 Mea -5.57 n 2.39 -10.01 0.29 1 Letters under “Site of release” and "Place of recovery” refer to the localities indicated in Text-figure 9. 1 1talicised dates of recovery and days elapsed refer to date of letter sent by captor where accurate dates were not obtainable. 8 South and east expressed as the negative of north and west respectively. 4 Presumably this fish was caught much earlier than date of letter. Further information unobtainable. Table XVII. Summary of All Tarpon Tagged. Location of Original Capture Number of Fish Locality 46 Land-locked Pine Island (1 ditch) 352 Sanibel Island (2 ditches, 2 ponds) 4 Gasparilla Island (1 ditch) 1 Captiva Island (1 pond) 403 All ponds and ditches 61 Open waters Boca Grande Pass 14 Captiva Pass 1 Redfish Pass 1 Off Palmetto Key 77 All open waters 480 All fish tagged Small fish Places of Release released where tagged, in pre- sumably land-locked pools and ditches. . 226 Small fish released in open waters 177 Large fish in open waters 77 Total 480 Total available for retaking in open waters 254 Total available only in ponds and ditches 226 Classification of Recoveries Numbers retaken in open waters 131 Percentage recovered 5 + Numbers retaken in ponds, etc 8 Percentage recovered 4 — Numbers of large fish retaken 1 Percentage recovered 1 + Numbers of small fish retaken 12 Percentage recovered 7 — 3 Omitting 2 tags cast and subsequently found on beach. and are general throughout the West Indian islands, the coasts of northern South Amer- ica, Central America and the Gulf of Mexico and do occur in less numbers on the east coast of Florida, they do not occur north of there except as stragglers. At Bermuda they are uncommon, Beebe and Tee-Van (1933). This type of distribu- tion is not uncommon. Whether the Ber- muda fish are to be considered as larvae caught in the eastward edge of the Gulf Stream, which have grown up in Bermuda waters, or whether they are accidental strays of larger fish like those that occur irregularly all along the Atlantic coast as far north as Massachusetts, is uncertain, but the indications would seem to be that the larvae are kept, or are able to stay, largely within the Gulf circulation and that it is mainly the larger specimens that have moved outward to form these accidental strays. Summary. 1. The development of an egg and larva believed to be that of the tarpon ie described, down to the pre-leptocephalus stage. 2. Dimensional changes with age and size are discussed and found to be remarkably small, from young to adult specimens. 3. The fate of small land-locked specimens is discussed and the conclusion reached that they are in the nature of peripheral wastage and do not partake in the sup- port of the adult population. 4. The area of successful spawning is con- sidered to be in the blue non-estuarine waters, because of the great abundance of predaceous plankton in the inside green waters. 1944] Breder: Materials for Study of Tarpon atlanticus 251 5. The checks on the scales of tarpon are compared with some of known growth under both wild and captive conditions, which suggests that they are true an- nulae. 6. Tagging operations suggest that there might be a slight drift to the northward of fishes during the summer months at least. 7. Reproduction is attained at a length ap- proximating four feet in total length, which is also the size of greatest hazard by anglers, but there is nothing in the present data to suggest that the current level or methods of angling is dangerous to the population. 8. Indications are that tarpon become sexually mature after passing their sixth or seventh winter. Bibliography. Babcock, L. L. 1936. The tarpon. 4th Edition, pp 175. Pri- vately printed, Buffalo, N. Y. Beebe, W. 1927. Tarpon nursery in Haiti. Bull. N. Y. Zool. Soc., 30 (5) : 141-145. 1938. Beneath Tropic Seas. G. P. Putnam, N. Y. pp. 234. Beebe, W. and Tee-Van, J. 1928. The fishes of Port-au-Prince, Haiti. Zoologica, 10 (1) : 1-279. 1933. Field book of the shore fishes of Ber- muda. G. P. Putnam, N. Y. pp. 337. Bishop, M. B. 1940. Notes concerning two broods of young of Natrix compressicauda. Copeia, (2) : 128. Breder, C. M., Jr. 1925. Notes on fishes from three Panama localities. Zoologica, 4 (4) : 137-158. 1933. Young tarpon on Andros Island. Bull. N. Y. Zool. Soc., 36 (3) : 65-67. 1937. Young tarpon at the Aquarium. Bull. N. Y. Zool. Soc., 40 (1): 21-22. 1939a. Report of the Director of the Aquari- um. (In the 43rd Ann. Rept. N. Y. Zool, Soc.) : 36-50. 1939b. On the trail of the tarpon. Bull. N. Y. Zool. Soc., 42 (4) : 98-110. 1939c. The tiniest of tarpon now at the Aquarium. Bull. N. Y. Zool. Soc., 42 (4) : 154-155. 1939d. On the life history and development of the sponge blenny, Paraclinus mar- moratus (Steindachner) . Zoologica, 24 (31) : 487-496. 1940a. Report of the Director of the Aquari- um. (In the UUth Ann. Rept. N. Y. Zool, Soc.) : 49-76. 1940b. The expulsion of young by the male of Hippocampus zosterae. Copeia, (2) : 128. 1940c. The spawning of Mugil cephalus on the Florida west coast. Copeia, (2) : 138-139. 1941a. On the reproduction of Opsanus beta Goode and Bean. Zoologica, 26 (21) : 229-232. 1941b. On the reproductive behavior of the sponge blenny, Paraclinus marmora- tus (Steindachner). Zoologica, 26 (22) : 233-236. 1941c. Respiratory behavior in fishes not especially modified for breathing air under conditions of depleted oxygen. Zoologica, 26 (25) : 243-244. 1942a. On the reproduction of Gobiosoma ro- bustum Ginsburg. Zoologica, 27 (11) : 61-64. 1942b. An octopus on guard. Animal King- dom, 45 (4) : 91-94. 1942c. On the phenomenon of locomotor dis- organization induced byt strong light in small plectognath fishes. Copeia, (4) : 211-213. 1942d. On the behavior of young Oligoplites saurus (Bloch and Schneider). Copeia, (4) : 267. 1942e. Social and respiratory behavior of large tarpon. Zoologica, 27 (1): 1-4. 1943. The eggs of Bathygobius soporator (Cuvier and Valenciennes) with a dis- cussion of other non-spherical teleost eggs. Bull. Bingham Oceanographic Coll., 8 (3) : 1-49. 1944. The metamorphosis of Synodus foe- tens (Linnaeus). Zoologica, 29 (3): 13-16. Breder, C. M., Jr. and Krumholz, L. A. 1941. On the uterine young of Dasyatis sabinus (Le Sueur) and Dasyatis has- tatus (De Kay). Zoologica, 46 (10): 49-53. 1943. On the locomotor and feeding be- havior of certain postlarval Clupe- oidea. Zoologica, 28 (10): 61-67. Breder, C. M. Jr. and Springer, S. 1940. On the electric powers and sex ratios of foetal Narcine barsiliensis (01- fers). Zoologica , 45 (26): 431-432. Brown, E. J. 1905. Florida fish killed by cold. Forest and Stream, 64 (6) : 118. Coker, R. E. 1921a. A record of young tarpon. Copeia, (93) : 25-26. 1921b. Smallest tai’pon. Copeia, (96) : 37. Cox, R. T. and Breder, C. M., Jr. 1943. Observations on the electric dis- charge of Narcine brasiliensis (01- fers). Zoologica, 28 (8): 45-51. Eigenmann, C. H. 1921. Small tarpon. Copeia, (95) : 33. Evermann, B. W. and Marsh, M. C. 1902. The fishes of Porto Rico. Bull. U . S. Fish. Comm,, (1900). 20 (1): 51-350. 252 Zoologica : New York Zoological Societg Finch, R. H. 1917. Fish killed by the cold wave of Feb- ruary 2-4, 1917, in Florida. Monthly Weather Rev., 45: 171-172. Fowler, H. W. 1936. The marine fishes of west Africa. Bull. American Mas. Nat. Hist., 70 (1): 1-605, (2): 606-1493. Gill, T. N. 1907. The tarpon and lady-fishes and their relatives. Smithsonian Misc. Coll., 48 (3) : 31-46. Gregory, W. K. and Conrad M. 1943. The osteology of Luvaris imperalis, a scombroid fish : a study in adaptive evolution. Bull. American Mus. Nat. Hist., 81 (2) : 25-283. Gunter, G.; 1941. Death of fishes due to cold on the Texas coast, January 1940. Ecology, 22 (2) : 203-208. Hecht, S. 1916. Form and growth in fishes. Journ. Morph., 27 (2) : 379-400. Hildebrand, S. F. 1934. The capture of a young tarpon. Tarpon atlanticus at Beaufort, N. C. Copeia, (1) : 45-46. 1937. The tarpon in the Panama canal. Sci. Monthly, 44: 239-248. 1939. The Panama canal as a passageway for fishes, with lists and remarks on the fishes and invertebrates observed. Zoologcia, 24 (3) : 15-45. Hile, R. 1931. The rate of growth of fishes of In- diana. In “Investigations of Indiana Lakes” Pub. No. 107 of the Division of Fish and Game, Dept. Cons. State of Indiana, (2) : 8-55. International Game Fish Association 1943. Yearbook (for 1943) : 1-112. Kampen, P. N. van 1909. Kurze Notezin jiber Fische des Java- Meeres. III. Die Larvae von Megalops cyprinoides Brouss. Bull. Dept. Agric. Indes-N eerl., 1908 (1909), (20): 10-12. Mathes, J. 1940. Louisiana pioneers in tarpon banding. Louisiana Conservation Rev., ( Sum- mer 1940) : 7-8, 19. Meek, S. E. and PIildebrand, S. F. 1923. The marine fishes of Panama. Field Mus. Nat. Hist. Pub., 214. Zool. Series. 13 (1) : 217-374. Merriman, D. 1940. Morphological and embryological stud- ies on two species of marine catfish, Bagre marinus and Galeichthys felis. Zoologica, 45 (13): 493-512. Miller, I. W. 1936. Fresh water tarpon. Field and Stream, 41 (1): 32-33. Shapiro, S. 1943. The relationship between weight and body form in various species of scom- broid fishes. Zoologica, 28 (12): 87- .104. Shlaifer, A. 1941. Additional social and physiological aspects of respiratory behavior in small tarpon. Zoologica, 46 (11): 55-60. Shlaifer, A. and Breder, C. M., Jr. 1940. Social and respiratory behavior in small tarpon. Zoologica, 45 (30) : 493-512. Simmons, W. E. 1900. The Nicaragua canal. Harper and Bros., N. Y. pp. 334. Storey, M. 1937. The relation between normal range and mortality of fishes due to cold at Sanibel Island, Florida. Ecology, 18 (1) : 10-26. 1940. Suppression of two generic names ( Auchenopterus and Cremnobates) of tropical American blennoid fishes, with notes on systematic characters. Copeia, (2) : 81-87. Storey, M. and Gudger, E. W. 1936. Mortality of fishes due to cold at Sanibel Island, Florida, 1886-1936. Ecology, 17 (4) : 640-648. Storey, M. and Perry, L. M. 1933. A record of young tarpon at Sanibel Island, Lee County, Fla. Science, 78 (2022): 284-285. Taylor, H. F. 1917. Mortality of fishes on the west coast of Florida. Appendix III to The Re- port of the U. S. Commissioner of Fisheries for 1917. Bureau of Fish. Doc. No. 848: 1-24. Westman, J. R. and Fahy, W. E. 1940. The carp problem of the area. In “A biological survey of the Lake Ontario watershed.” Suppl. 29 Ann. Rept. N. Y. Cons. Dept., (for 1939) : 226- 231. Westman, J. R. and Gilbert, P. W. 1941. Notes on age determination and growth of the Atlantic bluefin tuna, Thunnus thyyinus (Linnaeus). Copeia, (2): 70-72.' WlLLCOX, J. 1887. Fish killed by cold along the Gulf of Mexico and coast of Florida. Bull. V. S. Fish. Comm., 6: 123. 1944] Zoologica: Index 253 INDEX Names in bold face indicate new genera, species or varieties,- numbers in bold face indicate illustrations; numbers in parentheses are the serial numbers of papers containing the Plates listed immediately following. A Acanthoderes circumflexa, 10 nigricans, 1 1 venezuelae, 1 1 Acanthonyx petiverii, 120, 122 Achryson surinamum, 4 Acrocinus longimanus, 10 Aethopyga, 33, 36 boltoni, 34 christinae, 34 duyvenbodei, 34 eximia, 34 flagrans, 34 gouldiae, 34 ignicauda, 35 mysticalis, 34 nipalensis, 34 pulcherrima, 34 saturata, 34 shelleyi, 34 siparaja, 34 flavostriata, 18 Ageniella delila, 108 micans, 108 re versa, 108 Alasagenia, 106 erichsoni, 107 Anaiinomma brevicornis, 1 Anolis aeneus, 196 biporcatus, 197 chrysolepis, 197, 198, (18) Plale I fusco-auralus, 199, (18) Plale I niiens, 200, (18) Plale I punctaius, 200 sagreii, 201 schiedii, 201 Anoplius varius, 112 Anthreptes, 21, 22 anchieta, 22 aurantium, 22 collaris, 22 fraseri, 22 gabonicus, 22 griseigularis, 22 longuemarei, 22 malacensis, 22 neglecta, 22 pallidigaster, 22 platura, 22 rectirostris, 22 reichenowi, 22 rhodolaema, 22 simplex, 22 singalensis, 22 Aplochares, 1 1 1 imitator, 1 1 1 Arachnophroclonus crassidentatus, 111 Arachnothera, 35, 36 affinis, 36 chrysogenys, 36 crassirostris, 35 flavigasier, 36 juliae, 37 longirosiris, 35 magna, 36 philippinensis, 36 robusta, 36 Aslyanax mexicanus, 131 Astyochus mucoreus, 12 B Balboana auripennis, 104 fulvipes, 103 Balazonus decedens, 111 fervidus, 1 1 1 polistoides, 111 Balus barbicornis, 8 Belostoma asiaticum, 129 dentatum, 129 Brasilianus, 4 baius, 4 plicatus, 4 C Calappa flammea, 123, 127 Callipogon (Enoplocerus) armillatus, 4 Calopompilus vitreus, 102 Campylenchia hastata, 193 Ceresa vitulus minor, 194 Chalcomitra, 25, 26 Chrysoprasis viridis, 8 Colobothea pulchella, 12 Compsa vana, 7 Corymbophanes andersoni, 43 venezuelae, 41, (5) Plate I Cyanomitra, 24, 25 Cycloes bairdii, 124, 128 Cyllene cayennensis, 8 Cyphonia fuscata, 194 D Darnis labor, 194 Diaphus dumerili, 88 gemellari, 89 pacificus, 90 rafinesquii, 90 E Ebmia albolineaia, 4 Eburodacrys pilicornis, 5 254 Zoological Index Enchenopa albidorsa, 193 ignidorsum, 193 lanceolala, 193 serralipes, 193 Enchophyllum quinquemaculatum, 193 Epiallus bituberculatus, 120 G Gonalodes albogularis, 146, 147, (14) Plale I annularis 148, 149, 151, (14) Plates I, II beebei, 152, (14) Plate III boonei, 152, (14) Plate III caudiscutatus, 152, (14) Plate IV humeralis, 153 ocellalus, 153 vittatus, 153 Gnomidolon confusum, 6 Goniopsis cruentatus, 117 H Hemidactylus mabouia, 155 Heteronotus strigosa, 194 vulnerans, 193 Hippopsis lemniscala, 10 Hylettus vindex, 12 Hypopomus, 39 arledi, 40 beebei, 40, (5) Plate I* brevirostris, 40 occidentals, 40 I Ibidion, 7 beebei, 7 binoculatum, 7 Iguana iguana iguana, 201, 202, 203, (18) Plates II, III L Lagochirus araneiformis, 11 Lagothrix, 169-192, (16) Plates I-V cana, 169-192, (16) Plates I-V lagotricha, 169-192, (16) Plates I-V Lampanyctus elongatus, 79 idostigma, 81 longipes, 81, 82 macdonaldi, 82 macroplerus, 82, 83 mexicanus, 84 omostigma, 85 rilteri, 86 tenuiforme, 87 townsendi, 87. Lebistes reticulatus, 49, (7) Plate I Leplocoma, 26 Leptostylus gibbulosus, 11 Leptodius parvulus, 119 Lethocerus annulipes, 129 grandis 129 melo-leilaoi, 129 Limnogonus celeris magnus, 129 * Dr. Leonard P. Schultz, Curator of Fishes of the United States National Museum, asks that attention be called to “a black ‘eye* spot” appearing on the repro- ductions of Plate I in his paper, ‘‘Two New Species of Fishes (Gymnotidae, Loricariidae) from Caripito, Ven- ezuela,” published as paper No. 5, pages 39-44, in this volume of Zoologica. Dr. Schultz states: “This little black spot in the middle of the head must have been added without our knowledge, for it does not appear in the original photograph nor on the fish. The eye is the light area far forward near the tip of the snout.” Lissonotus equeslris, 8 Lissonotypus fasciatus, 9 M Macrodontia cervicornis, 4 Megaderus stigma, 9 Megasoma elephas, 53, (8) Plates I-V Membracis c-album, 193 arcuata, 193 foliate, 193 lefebvrai, 193 tectigera, 193 Menidia beryllina peninsulae, 45, 46, (6) Plate I Micropanope spinifer, 119 Microphrys bicornutus, 122, 124 Mithrax forceps, 121, 123 Myctophum affine, 61, 62 aurolaternalum, 64, 65 californiense, 66 coccoi, 67 evermanni, 73 laternalum, 74 pterotum, 75, 76 rarum, 76 reinhardti, 77 valdiviae, 78 Myzomela sclateri, 18 N Nectarinia, 23, 27, 28 adelberti, 26 afra, 30 alinae, 24 amethystina, 25 asiatica, 29 balfouri, 24 batesi, 24 bifasciata, 31 bocagei, 31 bouvieri, 31 chalcostetha, 27 chalybea, 30 chloropygia, 30 coccinigaster, 32 comorensis, 29 coquereli, 29 cuprea, 31 cyanolaema, 24 dussumieri, 28 erythroceria, 32 famosa, 32 fuliginosa, 25 fusca, 29 habessinica, 31 hartlaubi, 24 humbloti, 29 hypogrammica, 24 johannae, 33 johnstoni, 32 nyikensis, 32 jugularis, 28 frenata, 18 kilimensis, 33 lotenia, 28 loveridgei, 30 mediocris, 30 minima, 27 minulla, 30 neergaardi, 30 newtoni, 24 notate, 33 olivacea, 24 1944] Zoologica: Index 255 oriiis, 24 osea, 31 preussi 30 pulchella, 32 purpureivenlris, 31 regia, 30 reichenbachi, 24 reichenowi, 33 rockefelleri, 30 rubescens, 25 seimundi, 24 senegalensis, 25 sericea, 27 wigglesworthi, 27 shelleyi 31 Solaris, 29 souimanga, 29 sperala, 27 superba, 33 lacazze, 31 lhomensis, 24 ursulae, 24 venusla, 29 verreauxi, 24 verlicalis, 24 violacea, 30 zeylonica, 27 Neclariniidae, 21 Neoclylus guianensis, 8 Neodrepanis, 33 corruscans, 33 hypoxanlha, 33 Noliochares amethystina, 111 Notocyphus ’tyrannicus, 112 vindex, 112 O Onychocerus crassus, 10 Oreodera glauca, 10 jacquieri, 10 P Pachygrapsus Iransversus, 115, 117 Panopeus bermudensis, 119, 121 occidenfalis, 119 Panlonyssus erichsonii flavipes, 6 Parysatis nigrilarsis, 10 Pepsis completa, 101 diabolus, 101 dimidiala, 101 elevata, 100 equeslris, 101 excelsa, 100 {estiva, 102 flavicornis, 101 frivaldszkyi, 100 fulgidipennis, 100 gracilis, 102 grossa, 100 hecale, 102 ianthina, 101 iucunda, 101 Iuleicornis, 10l' margarete, 100 niphe, 100 nireus, 102 plulus, 100 preliosa, 101 ruficornis, 101 sagana, 101 sp. near ruficornis, 101 sp. near smaragdina, 102 strenua, 100 vicina, 101 xanthocerus, 101 Percnon gibbesii, 117, 118 Phanochilus, 108 gloriosa, 108 pilifrons, 108 nobililala, 108 Planes minulus, 114, 116 Planiceps periyi, 112 Plica plica, 204, 205-207, (18) Plate III umbra, 207, (18) Plates III, V Polychrus marmoraius marmoratus, 208, 209, 211, 212, (18) Plates IV, V Pompilinus orihodes, 112 Porlunus anceps, 114, 115 depressifrons, 113, 114 sayi, 113 Potamobates unidenlala, 129 Priochilus, 104 diversus, 105 formosus, 106 imperius, 105 nobilis, 105 opacifrons, 105 plulonius, 105 regius, 104 rhomboideus, 105 scrupulus, 105 sericeifrons, 105 superbus, 105 Priocnemella difformis, 110 eurytheme, 109 Priocnemiodes, 102 aurifrons, 103 biluberculalus, 103 gigas, 103 perpunclalus, 103 purpureipes, 103 Priocnemis (Calicurgus) nubilus, 102 Priophanes fabricii, 107 Probalius humeralis, 11 Promerops cafer, 18 Psammochares echinatus, 111 inculcatrix, 1 1 1 ornamenlus, 111 Pseudagenia basalis, 109 comparala, 109 femorata, 109 incroia, 109 modesla, 109 smithi, 109 larsata, 109 Pyrodes (Esmeraldo) auralus, 4 R Ranalra macrophlhalma, 129 Rhagovelia insularis, 129 tenuipes, 129 S Sericopompilus exilis, 111 Sphaerodaclylus molei, 153, (14) Plate IV Spongophorus guerini, 193 Stegaspis laevipennis, 194 Sleirasloma breve, 10 Slendonles (Mallodon) dasystomus, 4 Stizocera rubricollis, 6 Slralegus aloeus, 53 Stromatum kariaboensis, 1 Strongylaspis corticaria, 3 Synodus foetens, 13, 14-16, (3) Plates I, II 256 Zoologica: Index T Tapeina, 9 iransversifrons, 10 Tarpon atlanticus, 217-252, 227 Tenagogonus duolineatus, 129 Thecadactylus rapicaudus, 155, 157, (14) Plates V, VI Trachyderes succincfus, 8 Tropidurus torquaius hispidus, 212, (18) Plale VI U Uca, 161-168 Uranoscodon superciliosa, 213, 214, (18) Plale VI Urocenlron azureum, 214, 215 X Xanthodius denliculatus, 119, 120 NEW YORK ZOOLOGICAL SOCIETY General Office: 630 Fifth Avenue, New York 20, N. Y. Publication Office: The Zoological Park, New York 60, N. Y. Fairfield Osborn, President Alfred Ely, First Vice-president Laurance S. Rockefeller, Second Vice-president Harold J. O’Connell, Secretary Cornelius E. Agnew, Treasurer William Bridges, Editor and Curator of Publications Zoological Park Lee S. Crandall, General Curator & Curator of Birds Leonard J. Goss, Veterinarian Claude W. Leister, Associate, Mammals John Tee-Van, Associate, Reptiles Grace Davall, Assistant to General Curator Aquarium Christopher W. Coates, Curator and Aquarist Ross F. Nigrelli, Pathologist Myron Gordon, Assistant Curator C. M. Breder, Jr., Research Associate in Ichthyology G. M. Smith, Research Associate in Pathology Homer W. Smith, Research Associate in Physiology Department of Tropical Research William Beebe, Director Jocelyn Crane, Research Zoologist Henry Fleming, Entomologist George Swanson, Staff Artist OFFICERS SCIENTIFIC STAFF General John Tee-Van, Executive Secretary Jean Delacour, Technical Adviser Claude W. Leister, Education William K. Gregory, Associate Gloria Hollister, Associate John Tee-Van, Associate Mary VanderPyl, Associate Editorial Committee Fairfield Osborn, Chairman William Beebe William Bridges Christopher W. Coates Lee S. Crandall Jean Delacour Claude W. Leister John Tee-Van *