# ZOOLOGICA SCIENTIFIC CONTRIBUTIONS OF THE NEW YORK ZOOLOGICAL SOCIETY VOLUME 42 • 1957 • NUMBERS 1 TO 14 PUBLISHED BY THE SOCIETY The ZOOLOGICAL PARK, York NEW YORK ZOOLOGICAL SOCIETY GENERAL OFFICE 30 East Fortieth Street, New York 16, N. Y. PUBLICATION OFFICE The Zoological Park, New York 60, N. Y. OFFICERS PRESIDENT VICE-PRESIDENTS SECRETARY TREASURER Fairfield Osborn Alfred Ely Harold J. O’Conaell David H. McAlpin Laurance S. Rockefeller SCIENTIFIC STAFF: Zoological Park and Aquarium John Tee- Van General Director ZOOLOGICAL PARK Leonard J. Goss Assistant Director and Veterinarian John L. George Associate Curator of Mammals William G. Conway. .Associate Curator of Birds Grace Davall Assistant Curator, Mammals and Birds James A. Oliver Curator of Reptiles Charles P. Gandal. . .Associate Veterinarian Lee S. Crandall General Curator Emeritus William Beebe Honorary Curator, Birds AQUARIUM Christopher W. Coates . Director James W. Atz Associate Curator Carleton Ray Assistant to the Director Ross F. Nigrelli Pathologist Myron Gordon Geneticist C. M. Breder, Jr Research Associate in Ichthyology Harry A. Charipper. . .Research Associate in Histology Homer W. Smith Research Associate in Physiology GENERAL William Bridges Editor & Curator, Publications Sam Dunton Photographer Henry M. Lester . . . Photographic Consultant DEPARTMENT OF TROPICAL RESEARCH William Beebe Director Emeritus Jocelyn Crane Assistant Director David W. Snow Resident Naturalist Henry Fleming Entomologist John Tee-Van Associate William K. Gregory Associate AFFILIATES C. R. Carpenter Co-ordinator, Animal Behavior Research Programs L. Floyd Clarke Director, Jackson Hole Research Station SCIENTIFIC ADVISORY COUNCIL A. Raymond Dochez Carjd P. Haskins Alfred E. Emerson K. S. Lashley W. A, Hagan John S. Nicholas EDITORIAL COMMITTEE Fairfield Osborn, Chairman James W. Atz Lee S. Crandall William Beebe Leonard J. Goss William Bridges James A. Oliver Christopher W. Coates John Tee-Van William G. Conway Contents 3993r SI Part 1. May 20, 1957 PAGE 1. Miniature Circulating Systems for Small Laboratory Aquaria. By C. M. Breder, Jr. Plates I & II; Text-figures 1-5 1 2. The Behavior of the Bottle-nosed Dolphin (Tur stops truncatus): Mating, pregnancy, parturition and mother-infant behavior. By Margaret C. Tavolga & Frank S. Essapian. Plates I-III; Text-figure 1 11 3. A Study of the Relationship Between Certain Internal and External Mor- phological Changes Occurring During Induced and Natural Metamorphosis in Rana pipiens and Rana catesbeiana. By Beulah Howatt McGovern & Harry A. Charipper. Text-figures 1-12 33 Part 2. August 23, 1957 4. On the Bramid Fishes of the Gulf of Mexico. By Giles W. Mead. Plates I-III 51 5. The Relation of Oxygen Consumption to Temperature in Some Tropical, Temperate and Boreal Anuran Amphibians. By Richard E. Tashian & Carleton Ray. Text-figures 1 & 2 63 6. Basic Patterns of Display in Fiddler Crabs (Ocypodidae, Genus Uca). By Jocelyn Crane. Plate I; Text-figures 1-4 69 JAN 1 5 1958 Part 3. November 25, 1957 PAGE 7. Studies on the Lizard Family Xantusiidae. III. A New Genus for Xantusia riversiana Cope, 1883. By Jay M. Savage. Text-figures 1-3 83 8. Nesting Behavior of the Crested Oropendola {Psarocolius decumanus) in Northern Trinidad, B.W.I. By Richard E. Tashian. Plates I & II; Text- figures 1-3 87 9. Changes in the Cytological Structure of the Adenohypophysis and Gonads in Juvenile Bathygobius soporator after Pituitary Implantation. By Louise M. Stoll. Plates I & II 99 10. The Ctenuchidae (Moths) of Trinidad, B.W.I. Part I. Euchromiinae. By Henry Fleming. Plates I-III 105 Part 4. December 31, 1957 11. The Abifity of the Saprolegniaceae to Parasitize Platyfish. By Helen S. Vishniac & R. F. Nigrelli. Plate 1 131 12. Imaginal Behavior in Butterflies of the Family Heliconiidae: Changing Social Patterns and Irrelevant Actions. By Jocelyn Crane. Plate I.. . . 135 1 3 . Habits, Palatability and Mimicry in Thirteen Ctenuchid Moth Species from Trinidad, B.W.I. By William Beebe & Rosemary Kenedy. Plates I & II. 147 14. Serological Relationships among Members of the Order Carnivora. By Ludwig K. Pauly & Harold R. Wolfe 159 Index to Volume 42 167 ZOOLOGICA SCIENTIFIC CONTRIBUTIONS OF THE NEW YORK ZOOLOGICAL SOCIETY VOLUME 42 • PART 1 • MAY 20, 1957 • NUMBERS 1 TO 3 PUBLISHED BY THE SOCIETY The ZOOLOGICAL PARK, New York Contents PAGE 1. Miniature Circulating Systems for Small Laboratory Aquaria. By C. M. Breder, Jr. Plates I & II; Text-figures 1-5 1 2. The Behavior of the Bottle-nosed Dolphin (Tursiops truncatus): Mating, pregnancy, parturition and mother-infant behavior. By Margaret C. Tavolga & Frank S. Essapian. Plates I-III; Text-figure 1 11 3. A Study of the Relationship Between Certain Internal and External Mor- phological Changes Occurring During Induced and Natural Metamorphosis in Rana pipiens and Rana catesbeiana. By Beulah Howatt McGovern & Harry A. Charipper. Text-figures 1-12 33 1 Miniature Circulating Systems for Small Laboratory Aquaria C. M. Breder, Jr. The American Museum of Natural History, New York 24, N. Y. (Plates I & II; Text-figures 1-5) Introduction Because of the requirements of certain experiments it became necessary to es- tablish various small, but fully control- lable circulating systems in small aquaria. These have included both open and closed freshwater systems and closed saltwater systems. As the designs eventually worked out have proved to be entirely satisfactory, and as many colleagues have inquired about these systems, with a view to building similar ones for their own purposes, the details of construction and operation are explained here. Primarily these systems are the outgrowth of work of earlier years at the old New York Aquarium where much larger, but similar, equipment formed the basis of operations. This equipment itself had been developed from schemes used by older institutions of similar kind. Naturally, many persons had a hand in developing the arrangements and devices em- ployed at the New York Aquarium. For these reasons the origins of the devices were not al- ways clear, but those chiefly interested and re- sponsible for them at the Aquarium were C. W. Coates and the late C. H. Townsend, and H. Knowles. Townsend (1928) and Breder & Howley (1931) reported on some of these fea- tures. It has been found that by suitable modifi- cation of the principles of the larger devices it is possible to develop very useful miniature equipment. Such need, of course, applies only to laboratories not connected with large public aquariums and which consequently lack the utilities usually only to be found in such places. These devices have been worked out in connec- tion with experimental work carried on in the laboratories of the Department of Fishes and Aquatic Biology of the American Museum of Natural History, which has been supported in part by the National Science Foundation. Open Systems An “open system” as here used refers to one in which the water is used but once and not re- circulated. That is, there is only a supply line and a drain line. This calls for little comment in present connections except where a very small, well-regulated flow is required. Such apparatus may be arranged to provide as little as a speci- fied number of drops a minute, and wUl main- tain a surprising accuracy if properly designed. The overflow provided for this system is a constant-level syphon. If such a syphon is made by a glass blower it is expensive, subject to breakage and is not readily cleaned. Syphons can be quickly and cheaply made of some straight glass tubing, a tee, some flexible rubber or plastic tubing, and two small pieces of wood or plastic strip. No dimensions are given, as these will vary with the individual needs, al- though Text-fig. 1 is drawn to scale. The two pieces of plastic are identical and should have two holes drilled in them that will snugly fit the glass tubing. These parts are then assembled to make a constant-level syphon attached to the lip of an aquarium as shown in Text-fig. 1 and PI. I, Fig. 1. The lower piece of plastic may be fastened to the aquarium by small clamps or be cemented to it. The open, upper end of the tee vents the syphon. If a cap or plug is placed on this it immediately becomes a simple syphon and will drain the aquarium to the level of its inlet tube. This is sometimes found to be an added handy feature. The level of the water in the aquarium will be that at which the overflow water spills out through the horizontal leg of the tee. Adjustments of this to a fine point may be made by raising or lowering the tee through the hole in the plastic support, or this whole external assembly may be moved up or down by altering the position of the straight length of tubing which passes through the hole in the other piece of plastic connecting it with the tube in- 1 2 Zoologica: New York Zoological Society [42:1 Text-fig. 1. Side and end view of constant-level syphon made up of standard parts. side the aquarium. By arranging the outside part of the syphon to lie along the aquarium wall, as shown, the danger of its being in the way of operations is reduced. It is obvious that cleaning presents no problem with this type of syphon. In most cases a plastic or glass strainer is placed over the intake end of the syphon. If something does nevertheless block the syphon tube from the aquarium, it almost always can be cleared by blowing into the open end of the tee and restarting the syphon by drawing on this same open end of the tee while the outlet tube is held shut. It is possible and sometimes more convenient to attach the outer portion of the constant-level syphon to a small board which is affixed to a pivot at its upper end so that it is free to rotate on the stationary part attached to the aquarium frame. A small handle pointing upward from there makes its adjustment simple and marks on the latter in reference to some stationary part make return to a former rate of flow exactly possible. The action is simply that by rotating the part of the syphon so that the horizontal part of the tee raises or lowers, the level in the aquarium follows accordingly. This in turn affects the float valve, which is described below. The rate of flow will increase if the syphon outlet is lowered and decrease if it is raised. This is useful where the exact level of water is not of any importance but where it is desired to vary the amount of water flowing through the aquarium by specific amounts and where it is necessary to repeat such changes in flow at will. If the water supply has considerable pressure, such as is ordinarily encountered in city water systems, or approaches it, a pressure-reducing valve which may be regulated should be em- ployed. This can bring the pressure down to a value which will not burst or otherwise destroy the light equipment to be employed. This valve placed someplace in the supply line should be set so as to deliver little more than the maximum amount of water which will be required of it. Another way to accomplish the same purpose is to permit the supply water to run into a small reservoir of no more than sufficient height to provide enough head of pressure. Into this reser- voir the supply water is allowed to run continu- ally, of a little more volume than the aquaria will ever need. This is necessary to maintain a con- stant head in the reservoir. A small excess will overflow and go to waste by this method. It is economic only where water saving has no sig- nificance. The water flowing into this aquarium is con- trolled by a float-valve constructed of a glass stop-cock, a suitable-sized chemical flask and some small parts of either wood or plastic. These are assembled as shown in Text-fig. 2 and held together by iron screws and two pieces of strip steel. This metal is mentioned because of the danger of toxic salts forming if brass were used, since such corrosion might fall into the aqua- rium. A one-hole cork is bolted to a piece of Incite and then inserted into the mouth of the flask as shown. A dowel or plastic rod is inserted in the other hole in the plastic piece and secured. 1957] Breder: Miniature Circulating Systems for Small Laboratory Aquaria 3 Text-fig. 2. Top and side view of float-valve for control of inflow of water and additives. A similar piece of plastic is movably secured on this rod and on a similar one at right angles to it, extending from the valve. This is so arranged that the center of the flask comes to rest directly below the horizontal rod extending from the valve. This is best seen in the plan view of the device. By loosening the two set screws in the upper plastic piece the flask may be moved vertically on the one and horizontally on the other. PI. I, Fig. 2, shows one arrangement of this device. The extending glass tubes which are an in- tegral part of the stop-cock are inserted through two snug holes in the wood or plastic end-pieces of the valve and these are held in position by the two steel strips which are held in place by four wood screws. In the center of one of these steel pieces a hole is tapped into which is screwed a set screw with a pointed end (about 60°) and a lock nut as shown. The stem handle of the stop-cock is imbedded in a piece of wood cut about as shown. For this purpose a suitable space is hollowed out in the block into which the stem is inserted, the space around being filled with plastic wood or similar product. The face of the block and the stop-cock stem must be at right angles. On the outer face of the block a small steel strip is affixed with a small drill-tip impression at its center on the axis of the stem. Into this the pointed set screw fits as shown. This is adjusted so that the valve works freely with- out being too tight or leaking. This prevents the glass stop-cock from working loose and leak- ing after a time of long-continued operation. The diagonal dotted line on the block indicates the position of the hole through the plug as well as that of the handle on the stem. It is shown in a position just fully closed. It is obvious that with a fall in the water level the valve will open proportionally to the change in water level and shut itself off as the water level rises. The interaction between the constant-level syphon and this valve is indicated in Text-fig. 3. It is clear that danger from flooding could come only from some damage to the equipment. If, for instance, something clogs the overflow in any way, the float valve shuts itself off when it has reached whatever predetermined point for which it has been set. As an extreme point of precaution a safety alarm or shut-off could be built as an entirely separate system. Such a device, which has never failed so far as the writer’s experience goes, con- sisted of an old pair of contacts such as are to be found on relays to one member of which was fastened a shell vial. This hung over the water in such a manner that when the water rose over a specified place it lifted the vial and pushed the two contacts together. It operated on two dry cells to ring a doorbell but could be used with a relay to switch on house current to operate any suitable device. This could be a normally open solenoid valve placed in the supply line. Such extreme caution would only be warranted where a little flooding would be disastrous. It is obvious that this float-valve could be used under certain experimental procedures to add chemicals to an aquarium at a prescribed rate by inactivating the float and fixing the rate of flow by hand. It also could be used to bring the con- centration of some chemical to a fixed limit and then hold it at that point in flowing water 4 Zoologica: New York Zoological Society [42:1 Text-fig. 3. Diagram of interaction between float-valve and constant-level syphon. Dashed line represents level of water in aquarium and syphon arm. aquaria. The water supply would operate as above described and a second float-valve regu- lated to add much less chemical than the water flow would move with it and act as a follower to the other if there was any fluctuation in the flow of water, thus holding the additive in pro- portion to the change of water. Also a single float could be arranged to operate the two valves in proportion to the setting of each. While the designs of these float-valves have varied from time to time, all have been built embodying the same principles as herein de- scribed. The first and somewhat primitive one has, at this writing, been in service continually for more than four years and is still entirely satisfactory and dependable. Closed Systems The term “closed systems” refers to circu- lating systems in which the water is returned to the aquaria after filtration or other treatment and none is allowed to run to waste during normal operations. Some such system is manda- tory for the maintenance of marine forms re- mote from a ready supply of sea water, and often convenient or necessary for various experimental procedures involving freshwater aquaria. This is especially true of cases where it is necessary to maintain close control of some feature such as temperature, chemical quantities and the like. By use of such means it is possible to maintain a series of aquaria with absolutely identical water conditions, as the water in all is part of a common body. Consequently no matter what transpires in one aquarium there is no oppor- tunity for the water of that one to depart from the characteristics of the rest since it is moving freely from one aquarium to the other and is being continually and effectively mixed. An especially useful arrangement for some purposes is one in which the flow between aquaria may be continuously varied from maxi- mum in one direction through zero flow to maximum flow in the opposite direction. This may be readily accomplished by the adjustment of four valves while the pump runs continuously in one direction at constant speed. The details of the arrangement of these valves are shown diagrammatically in Text-fig. 4. PI. II, Fig. 3, is a photograph of such a device. In operation the action is as follows. With valves A2 and B1 closed and the others open, the flow is out through pipe A and returns through pipe B, as indicated by the arrows, at maximum flow. If these valves are reversed so that A1 and B2 are closed and the others open, the flow through pipes A and B is reversed, although the flow through the pump remains as indicated by the arrow on it. To pass uniformly from the first position, as shown in Text-fig. 4, through a state of no flow to the reverse, either valve A2 or B1 can be gradually opened. This reduces the speed of water movement because of “back leakage.” After one of them has been opened fully the opening of the other can further retard flow. When it, too, has been fully opened, that is with 1957] Breder: Miniature Circulating Systems for Small Laboratory Aquaria 5 Text-fig. 4. Diagram of piping and valves for continuously variable flow from maximum in one direction through zero flow to maximum in the opposite direction. all valves fully open, there should be no flow through pipes A and B, as there is as much pump pressure in one branch of both A and B pipes as in the other. Then by beginning to close either valve A1 or B2 the flow begins to move in the opposite direction. When these two are fully closed, the maximum flow in the opposite direc- tion has been reached. If three valves are arranged on either line A or B or such a set on both, various water treat- ment devices may be placed in the series, such as a heating or cooling device, in which case the water may be best passed through a glass coil for heat exchange purposes. This arrangement could equally well be used for any kind of de- contamination which might be required or for the introduction of specified materials being mixed with the passing water. Unlike most aquaria plumbing, in this case the water must pass from one aquarium to another, so that as many as may be needed can be placed in series. A diagram of a more usual arrangement for a closed circulating system is shown in Text-fig. 5. This is the form which is perhaps most useful for general laboratory purposes. Here each aquarium is respectively supplied and drained directly from a common supply and return. It is advisable to use a pump of somewhat greater capacity than needed for the purpose. With this means the excess water can be returned to the reservoir without passing through the aquaria. which greatly assists in the eflBcient application of whatever water treatment is being given and insures adequate pressure for the system. Simple filters may be made by wedging a piece of glass in a small aquarium and filling the intake side with suitable filter material, while the reservoir may be used for whatever chemical or other treat- ment is to be applied. Two may be provided, as shown in the figure, and used alternately or together. A constant-level syphon takes the water to the first reservoir aquarium. Only one syphon is shown, which may be switched to the other aquarium for cleaning purposes. The pipes and fittings used in this system are standard hard rubber. Connections between aquaria at the same level may be made by “jumpers” which are preferably of hard rubber. Their use is indicated in the lower level aquaria shown in Text-fig. 5. These have been found to be fully satisfactory and in several years’ opera- tion have not clogged nor have they become airbound. They are, however, not suitable for the overflow lines of the upper series of aquaria. Here constant-level syphons may be used as shown in PI. I, Fig. 1, or, preferably, a hole may be drilled in the slate bottom of each aquarium and a one-hole rubber stopper holding a glass tube inserted. Still better is the installation of a hard rubber standpipe locked in place with fit- tings. The drilling of slate is not easily accom- plished and there is considerable danger of 6 Zoologica: New York Zoological Society [42:1 cracking or otherwise damaging the aquarium. The manufacturers will supply aquaria with such holes drilled on order. Since the supply to the aquaria of this system is preferably from the top, as indicated, the drain line as above described may be made to draw water from the bottom of an aquarium by the following simple means. A tube of glass, or other material, of larger diameter than the drain tube and as long as the depth of water in the aqua- rium, is placed over it, reaching nearly to the bottom of the aquarium. Since the larger tube extends above the surface, water leaving the aquarium must enter the annular space between the two tubes and pass up between them to spill into the open upper end of the inner tube. In addition to giving the aquarium a better circu- lation, much detritus is drawn up through this arrangement and delivered automatically to the filters. The outer tube, if of glass, may be posi- tioned by slipping a short piece of plastic tubing on its lower end and cutting various openings or notches in the plastic. The area of these passages should of course be at least equal to the cross- section area of the inner tube. The annular space between the inner and outer tubes should also have this much area, at least. On the other hand, it is best not to make the annular space much larger than needed because this will cause the water flowing through it to move with less speed. The value of this arrangement as a detritus re- mover is thus lessened, for the slower-flowing water will not lift as heavy particles as will the faster. As such a system is usually intended to be operated continuously for long periods without attention, a safety feature may be built in which would shut down the pump if the water in the reservoh' rose too high or fell too low. The one in current use, shown in PI. II, Fig. 4, was im- provised from the tube of a mercury switch. This was mounted on a rotatable glass shaft running through a support of plastic. It was actuated by a chemical flask float by means of a thread over a small drum so that the motion of the float was transmitted to the pump switch. Any unusual change in the water level, either positive or nega- tive, would indicate some radical failure at some point in the system. Since the aquaria which held the fish were drained by an overflow they would continue to hold their water level so that stop- ping the pump would insure the retention of water there. Even if one of the aquaria leaked and lost its contents the others would not suffer by draining through the system to it because of this protective device. To prevent normal evaporation from stop- ping the pump, a supply was provided which operated in conjunction with the protective switch. This supply was administered by a float valve identical with that shown in Text-fig. 2. A very nice adjustment was found possible with these two float-actuated mechanisms, so that the dripping from the float-valve supply became 1957] Breder: Miniature Circulating Systems for Small Laboratory Aquaria 1 directly proportional to the evaporation, without at any time tripping the protective cut-off ffoat- valve. On very humid days it could be seen that the number of drops per minute was notably less than on a clear dry day when evaporation was high. In the case of saltwater aquaria this device had an added important application which is discussed under the treatment of salt water. Because of the nature of the controlling de- vices above described, it is necessary to observe certain details in starting the system. The levels of water in the lower series of aquaria will be different when the pump is not running than when it is in operation. This is mostly because the drain lines of the upper series of aquaria empty themselves into the lower aquaria when the pump is stopped. Therefore the water is carried at a lower level in these aquaria so that there will be no overffowing when the circula- tion has been stopped. For this reason a switch should be shunted around the cut-off float to be used in starting the system before the operating level is reached. It will not suffice to wedge the float into a position where its switeh will be closed, because its free action is necessary to establish its proper level of operation. After a dynamic equilibrium has been achieved by ad- justing both the cut-off device and the density control device, the shunt switch should be opened, after which the system should control itself. If it does not at first, very obvious adjust- ments of either or both will bring them into the proper relationships. The diagram of the closed marine circulating system shown in Text-fig. 5 employed seven aquaria for holding experimental fishes, only three of which are shown in the illustration. Three “reservoir” aquaria were used, of which only two are shown. These were standard com- mercial aquaria measuring 2' X 1' X 1'. The two smaller, used as filters measured 10" X 8" X 6". The pump was driven by a 1/10 hp motor and was rated at 10.8 gals, per min. at zero head. The pipe sizes are not indicated, as they would naturally vary with the needs of each system. In this one the flow was slow but suf- ficient at about three gals, per hour through each of the seven top row aquaria. At the right of Text-fig. 5 the supply pipe is extended up- wards for some distance and with the upper end open. This permits building up whatever head of water is desired without subjecting the pipe to pressure greater than that produced by gravity. Although the upper series of aquaria were intended for holding fishes and the lower series were regarded as treating reservoirs, the latter too may be, and have been, used to hold fishes, that is, all but the one from which the pump draws water, since the suction and turbulence here would be destructive to most small fishes. Aerating stones and a standard aquarium glass heater and thermostat comprised the rest of the water-treating equipment. The heater which turned off when the water reached 74°F. was sufficient to keep the water throughout the sys- tem close to that temperature as it was only slightly higher than the normal room tempera- ture. It was found that the aerating stones made it possible to permit the flow of water in the lower aquaria to run through submerged pipe outlets and thereby reduce the amount of splash- ing and consequent salt deposits. This was not found necessary in the upper series, for each supply pipe carried only one-seventh of the flow in the lower pipes. Materuls It is strongly recommended for all the pur- poses for which these devices were developed that only hard rubber or some biologically inert plastic be used. In fresh water, iron plumbing is adequate for many purposes but for sea water no metals whatever should be used if any de- gree of satisfaction is to be obtained. Hard rub- ber and acrylic resin or vinyl chloride-acetate copolymer plastics have been used throughout for those parts which come in contact with the water, including the pumps.i Also it is important to see that no brass or other such metals are used in positions over the aquaria in order to prevent possible corrosion falling into the water. In all cases involving the use of pumps for aquarium purposes it is best to have a spare stand-by duplicate pump and motor unit as a precaution against the failure of either motor or pump. It is then possible to change such a unit in a few minutes, in the case of accident, with no serious interruption to the operation of the system. It is most convenient to use flexible connectors between the pump and the rigid plumbing leading to the aquaria. It is then only necessary to unfasten two screw clamps and in- sert the new unit in place. This type of arrange- ment is shown in PI. II, Fig. 3. An additional advantage of this kind of con- nection is that it dampens any vibrations, from the pump or motor, which tend to travel throughout the system along rigid connectors. The pump in PI. II, Fig. 3, was suspended by iln the systems described, the plastics used in the construction of the apparatus go under the trade names of Lucite and Plexiglas and the tubing under the name Tygon. The pumps have been supplied, complete with motors, by Broadbent-Johnston, Inc., Compton, Cali- fornia. 8 Zoologica: New York Zoological Society [42:1 four light cords, a means which is also very effective in quieting such small machines. Treatment of Water The treatment of fresh water for aquarium purposes is too well known to warrant comment m present connections and is usually necessary only under special situations. The maintenance of sea water in a satisfactory condition is quite another matter, however. It is not the purpose here to discuss the theoretical aspects of the chemical and physical conditions of sea water. Such matters may be found extensively treated by Sverdrup, Johnson & Fleming (1942) and Harvey (1955). The following is intended purely as a guide for the practical application of principles which have been found adequate to maintain a variety of marine fishes. Under this treatment regular reproductive behavior was quickly estabhshed in both Histrio and Bathygobius, which had been reared from juve- niles. It also permitted a variety of volunteer algae and microorganisms to establish them- selves. Incidentally these aquaria were kept under conditions of no daylight, the illumina- tion being supplied by fluorescent tubes of the “warm white” type necessary for satisfactory plant growth. The periods of light and darkness were controlled by a time switch. The equipment found necessary for the con- trol of sea water consisted of a small hydrometer, a colorimetric pH device and some simple titrat- ing equipment. The filters were provided with bone charcoal and the bottoms of the aquaria and reservoirs were floored with so-called coral sand, and aerating stones were placed in various conven- ient places, but not in the aquaria containing fishes. The specific gravity and pH were taken every day until the rate of change was established and from then on were taken at less frequent inter- vals. This rate of change will vary with the quantity of water, the bulk of the organisms con- tained and the temperature of the water. At less frequent intervals titrations were made to de- termine the variously-called excess base, titra- tion alkalinity, or alkaline reserve. This method, which measures the bound CO2, is not especially accurate but is sufficient for the present pur- poses. It consists of titrating a sample with N/lOO hydrochloric acid to which brom-cresol purple has been added as an indicator. After the purple color has vanished the sample is repeat- edly boiled and further titrated until the purple color no longer reappears on heating. If the sample consists of 100 cc. to which five drops of indicator have been added, the final burette reading in cc. multiplied by 0.1 gives the bound CO2 or bicarbonate in mfilimols/liter. This method is not to be generally recommended for accurate work but is sufficient as a comparative measure of how far and how fast the aging water is departing from its original value. With this information, corrective measures may be taken. The specific gravity is nearly taken care of by automatic means involving the use of the float-valve already discussed. Under normal operations distilled water is used to make up for the evaporation of sea water, which of course tends to increase its density thereby. This has been satisfactorily supplied from a five- gallon carboy on a shelf higher than the float- valve. The operation of the float-valve holds the amount of water in the system at a constant volume, which means also that the dissolved salts will remain at a constant amount. If it is desired to increase the density of the water, in- stead of using distilled water as an additive, sea water may be used until the specific gravity has reached the desired level. If it is desirable to re- duce the salinity, water may be withdrawn from the system while distilled water is used in the float-valve supply. This may be conveniently ac- complished by means of a syphon with a small hose clamp so that the flow is restricted to a drip slow enough to permit the float-valve to follow. Although distilled water was customarily used, in its absence tap water was used with no detectable effect on the fishes or the system. If the pH falls to lower values it may mean that there is an increase in the amount of free CO2 present. This could indicate too many or- ganisms for the volume and temperature of the water or too much decomposition for the anti- acid components of the system to dispose of rapidly. The calcium carbonate in the sand should react with the acids formed and unless there is overcrowding this type of decreasing alkalinity usually does not present a problem. If the placing of fresh activated bone charcoal in the filter results in an abrupt increase in the pH, it is almost certain that there is too much free CO2 present. The use of charcoal renewed at short intervals will bring the CO2 content down but the charcoal rapidly becomes satu- rated and cannot be thought of as a regular part of the regulatory process. An increase in the number of aeration stones or amount of air they pass, while much slower in its effects, is a much more satisfactory way to insure against the ac- cumulation of CO2. If on a falling pH, none of the procedures above mentioned increase the pH significantly, the titration reading should be carefully checked and it too should show a decrease. This would indicate a lowering of the bound CO2 which does 1957] Breder: Miniature Circulating Systems for Small Laboratory Aquaria 9 not normally occur in an imoverloaded system in the presence of calcareous sand. If it does, however, more sand may be added, or sodium bicarbonate may be dissolved and administered with the distilled water through the float-valve. Since the sand alone tends in a long-term sense to disproportionately increase the Ca in solution as compared with the Na, the occasional use of sodium bicarbonate, which tends to do the re- verse, aids in keeping these two quantities in more nearly normal proportions. See Breder & Smith (1932). The described procedures may seem to be somewhat complicated, but they are, in fact, not much more complex than those involved in maintaining a similar number of standing fresh- water aquaria. After the equipment is built and regulated, so that valves and controls are in balance, there is nothing to be done with them at any time and in fact there should be no tam- pering with them at all. It is probably wise to post warnings to this effect. There is little aquar- ium cleaning to be done, as most of the accumu- lating detritus is automatically deposited in the filters. Aside from feeding the fishes and some- times cleaning algae off the glass sides, the latter being controlled by adjusting the lighting ar- rangements, there are the following routine matters to be done. These will vary with each installation but may be approximated by the regimen under which the described installation of seven aquaria were controlled, as follows. Read pH and Sp. G Twice a week 5 min. Titrate sample Once a week or less 15 min. Fill distilled water Twice a week 5 min. carboy Clean filters Once in 2 weeks 15 min. Make adjustments based Once a month or on above information longer Various The need for changing the pump and motor is such a rare occurrence as not to figure in the above schedule and should not take more than 5 minutes. Every attempt has been made to re- duce the maintenance of the system to its mini- mum. It is not uncommon for the system to be left alone for as long as three days, as over a long week end. The only thing to normally expect at the end of this period is some extra-hungry fishes. It should be borne in mind that the smaller the system— that is, its total amount of water— the more rapidly decomposition or con- tamination can spread through it. It is prudent not to reduce the size of the equipment more than necessary. References Breder, C. M., Jr., & T. H. Howley 1931. The chemical control of closed circulating systems of sea water in aquaria for tropical marine fishes. Zoologica, 9 (11): 403-442. Breder, C. M., Jr., & H. W. Smith 1932. On the use of sodium bicarbonate and calcium in the rectification of sea-water in aquaria. Jour. Marine Biol. Assoc. N. S., 18 (1): 199-200. Harvey, H. W. 1955. The chemistry and fertility of sea waters. Cambridge University Press, Cambridge, i-viii: 1-224 pp. Sverdrup, H. U., M. W. Johnson & R. H. Fleming 1942. The oceans. Prentice-Hall, Inc., New York, i-x: 1-1087 pp. Townsend, C. H. 1928. The public aquarium, its construction, equipment and management. Report U. S. Commissioner of Fisheries. Appendix 7: 249-337. 10 Zoologica; New York Zoological Society [42:1:1957] EXPLANATION OF THE PLATES Plate I Fig. 1. A constant-level syphon in operation. Fig. 2. Float- valve for the control of inflow. Note: This is the arrangement as used in the closed saltwater system. The placement is for convenience, but usually valves are placed close to one end. Plate II Fig. 3. Pump provided with reversible flow de- vice. The plastic pump here shown is powered by a 1/80 hp motor and is rated to deliver 3.7 gals, per min. at zero head. Fig. 4. Safety control for closed circulation. Its relation to the system is indicated in Text- fig. 5. The pipes at the extreme right are the pump intake and excess return. BREDER PLATE I FIG, 1 FIG. 2 MINIATURE CIRCULATING SYSTEMS FOR SMALL LABORATORY AQUARIA BREDER PLATE II FIG. 3 FIG. 4 MINIATURE CIRCULATING SYSTEMS FOR SMALL LABORATORY AQUARIA 2 The Behavior of the Bottle-nosed Dolphin {Tur slops truncatus): Mating, pregnancy, parturition and mother-infant behavior^ Margaret C. Tavolga Department of Animal Behavior, American Museum of Natural History, New York Frank S. Essapian^ Marineland Research Laboratory, Marineland, Florida (Plates I-III; Text-figure 1) Introduction Because Cetacea bear their young in water, data on the normal progress of pregnancy and parturition, together with information on mating behavior and post-natal development, have been either lacking or only incompletely available to biologists interested in these animals. The greatest part of the informa- tion as yet recorded has been received from whalers and other sea-faring men who have observed these animals in the open sea, where conditions are far from ideal for such observa- tions (Liitken, 1888; Pedersen, 1931; Slijper, 1936). James (1914) first observed the birth of a Common Porpoise (Phocaena phocaena (Lin- naeus) in captivity in the Brighton Aquarium, but this was a stillbirth. Khvatov (1938) and Sleptzov (1940) both observed what they be- lieved to be normal births in the Common Dol- phin {Delphinus delphis Linnaeus) in shallow portions of the Black Sea, or in large seines in which a number of animals were confined prior to being drawn toward shore. McBride & Hebb (1948) and McBride & Kritzler (1951) were the first, however, to record their observations under conditions almost ideal for this purpose. In their 1951 paper, McBride & Kritzler pub- ^The authors wish to extend their thanks to the staff of the Marineland Research Laboratory, Marineland, Florida, for their cooperation and for the use of their facilities and equipment. They also wish to express their appreciation to Dr. Lester Aronson of the American Museum of Natural History, Mr. F. G. Wood, Jr., of the Marineland Research Laboratory, and Dr. Frank A. Beach of the Department of Psychology, Yale Uni- versity, for their special help in reading the manuscript. ^Present address: Seaquarium, Miami, Florida. lished on this subject the latest data taken at Marine Studios, Marineland, Florida, where the activities of the Bottle-nosed Dolphin {Tursiops truncatus Montagu) in a 75-foot circular tank may be viewed conveniently through portholes set in the walls. At that time five pregnancies resulting in normal births had been under ob- servation. Three animals had conceived in the tank. Since then five additional pregnancies, all resulting in normal births, have been observed, providing many additional data. The present re- port is based upon a series of observations in- stituted by one of us which was conducted over a period of slightly more than three years, and which for the last year and a half was on a daily basis. During parts of the latter portions of these observations, both authors were able to be at Marineland at the same time, and obser- vations could be made almost constantly, so that a steady daily record was kept which provided the observers with continuous data, not only on the group as a whole but also on the activities of any individual animal. Up to the time of the 1951 paper by McBride & Kritzler, new animals were introduced at ir- regular intervals and some deletions were made in order to keep the population at a desirable level and at the same time to minimize the pos- sibilities of combat which resulted on a previous occasion in injuries to a newborn dolphin (Mc- Bride & Kritzler, 1951). Since then, the colony has become a stabilized and self-perpetuating group, in which the only additions are those by birth. This fact has enabled the observers to ob- tain data on the growth and maturation of sev- eral individuals. One of the females born in the tank has reached the beginning of her eighth 11 12 Zoologica: New York Zoological Society [42:2 year and terminated her first pregnancy in March, 1954.3 Table 1 shows the names, sexes, dates of birth and relationships of the members of the stable population of the tank. In addition. Happy, a fully-grown adult male, and presumably the father of all the infants conceived in the tank, was present throughout the three years covered by the present data. Frank and Floyd, born in the tank, were removed in December, 1953. actively sought the company of the selected female by such stratagems as chasing her when she left him, hovering over her when she hunted for fish or was otherwise occupied in one spot, and by returning to her immediately after both had been engaged in other activities, such as feeding. Such exclusive companionship lasted from three or four days to several weeks. This period of prolonged companionship of- ten, but not always, led to further activities on Table 1. Stable Population of Tursiops truncatus in Tank at Marine Studios. Date of Sex of Name of Name of Duration Birth Infant Infant Mother of Birth* Remarks Feb. 26, 1947 Female Sprayt Mona 28 April 23, 1948 Male Mrs. Jones 21 Died of injuries May 8, 1949 Male Algiet Susie 117 May 12, 1950 Male Frank Pudgy May 15, 1950 Male Floyd Mrs. Jones Feb. 7, 1953 Female Mamie Mona Died June 13, 1953 Feb. 23, 1953 Female Maggie Mrs. Jones Feb. 27, 1953 Female Nellie Susie May 7, 1953 Male Mitch Pudgy 68 Died Feb. 22, 1954 March 4, 1954 Female Peggy Spray 48 Died March 19, 1954 *Duration of birth of infant in minutes, from first appearance to complete emergence. Where no figure appears, the birth took place at night and was unobserved. ■[These animals were conceived in the wild before the capture of the mother. AU other infants in the table were conceived in the tank. Precopulatory Behavior In the spring of 1952, 1953 and 1954 (Feb- ruary to May inclusive), there occurred a num- ber of special activities which appeared to be the preliminaries to copulation, since the copu- latory pattern was frequently seen to follow them. The aggregate of these activities over a period of time between two animals of opposite sex we term courtship. It should be understood that not all of these activities need necessarily take place between any two particular animals, and further, that they need not follow any par- ticular order. During the major portion of the year the adult male changed his companions often, swim- ming with one animal for a few minutes or a few hours, and rarely remaining with it for more than a few days. His companions during this time were other males, females or young ani- mals of either sex. He also remained alone for long periods. During the early spring, however, he showed preference for the company of one particular female and remained with her for prolonged periods. He either accepted the pres- ence of other animals or avoided them, but he ®This paper was submitted for publication on Feb- ruary 13, 1956. the part of the male, depending on the at- tendant circumstances. At times, without any further activity, the male suddenly broke off the partnership and shifted his attention to other interests, and there was no further association between the male and female for weeks or months. On the other hand, he often suddenly increased his attention to her and engaged in activities which were more aggressive and more directly associated with precopulatory behavior. These activities were quite varied and included posturing, stroking, rubbing, nuzzling, mouth- ing, jaw clapping and yelping. Any or all of these were included in any sequence and varied in intensity, becoming most intense just before exhibition of the copulatory pattern. Posturing. — Although this type of act ap- peared at any time during precopulatory be- havior, it was usually one of the first indications, aside from prolonged companionship, that pre- copulatory activities were about to become more intense. The male swam toward the female and took up a position in front of her, or slightly to one side, but well within her line of vision. He then bent his body in such a way that from the side it took on the shape of an S-curve, head pointing up, flukes down, as in Text-fig. 1. He maintained this position for about two to five seconds, and then resumed normal swimming. 1957] Tavolga & Essapian: Behavior of the Bottle-nosed Dolphin 13 The action was repeated immediately, after several minutes, or not at all, depending on the female’s general responsiveness and the other activities taking place in the tank. All males postured sometimes, but males were seen to begin courtship without posturing. Therefore it appears to be a subordinate portion of the pre- copulatory pattern of behavior. Although pos- turing was usually shown by the male, it was observed in the female under certain circum- stances which will be described later. Text-fig. 1. Male dolphin in posturing position. Stroking. — During normal swimming the male took up a position just behind and below the female. When this happened, the female often altered the normal motion of her flukes in such a way as to stroke or lightly pat the head of the male as she swam. At times the posi- tions were reversed, and the male stroked the female’s head. The stroking motion was easily distinguished from the normal swimming mo- tion, and also from the lashing of the flukes given by another animal in an attempt to avoid contact or to get away. Such action on the part of the two animals was usually assumed when they were resting, and was the least vigorous and energetic of the sexual activities. At other times either male or female swam in such a way as to bring one or both flippers in contact with some part of the other’s body (PI. I, Fig. 1 ) , and by swimming along or across the other’s body produced a stroking effect. Another form of the same activity took place when one of the animals turned on its back and, swimming be- neath the other in the same direction and at the same speed, moved its flippers rhythmically back and forth across the other’s body. The other member of the couple responded by repeating the action at the some time so that the two sets of flippers stroked each other. Such actions could, of course, be continued only until it was necessary for one or the other to rise to breathe, but it was resumed immediately following this, and with interruptions for respiration was some- times continued for half an hour or more. Stroking was also carried out by gentle move- ments of the head against the body of the partner. Another variation took place when one animal placed itself beside or beneath the other. and stroked the genital area of the partner with the tip of one fluke. Rubbing. — This was a much more strenuous action in which one member, usually the male, swimming at a fast pace, advanced upon the female and rubbed part of his body vigorously against hers (PL I, Fig. 2). This advance took place from any direction relative to the two animals, but was most dramatic when the male approached the female head on, as if to engage in a head-on collision, and then turned aside at the right moment and just enough to produce vigorous contact of the two bodies along their lengths. This type of inter-animal contact rarely lasted for more than a few minutes, during which the male made repeated rushes at the female. Trumpeting and whistling sounds were sometimes heard during this time. The female sometimes turned after each rush and swam toward the male as he repeated the maneuver, enhancing the vigor of each contact, or she rushed away, leaping out of the water in a real or apparent attempt to escape. When the latter took place, the male usually rushed to the spot where she was to reenter the water and rubbed energetically against her as she did so. (PI. I, Fig. 3). Nuzzling. — Nuzzling took place when one animal applied its closed snout to some part of the body of the partner and moved it around in that area. Although this action took place in a number of other situations, such as the nuzzl- ing which goes on between mother and infant, the male often nuzzled at the genital area of the female, possibly as an investigative measure, and also possibly stimulating her thereby. The female, at times, nuzzled the male in the same way. Mouthing. — (PI. I, Fig. 4) Mouthing was an activity similar to nuzzling, but took place with the mouth open. Although used during precopu- latory activity in the same manner as nuzzling, it also took place in other ways. When in an appropriate position, one animal took a flipper, fluke or sometimes even the head or snout of the partner between the teeth (PI. I, Fig. 4) , and gently closed the mouth upon it, thereby holding the part. This grasp was a gentle one, and was released as the partner pulled away, without leaving tooth marks or gashes. This is in con- trast to the grips taken on each other by two fighting males, which result in deep cuts and gashes inflicted by the sharp teeth. Either nuz- zling or mouthing of the male’s genital region by the female sometimes resulted in an erection. Jaw Clapping. — At times during the com- panionship of the male and female, a second 14 Zoologica: New York Zoological Society [42:2 male interfered in the relationship between them, either by swimming between the two, separating them, or by nuzzling the female while she swam with the first male. In such a case, the male’s response was rapid and often violent. He is able, by clapping his jaws together forcefully, to produce a sharp loud sound, which can be heard at a distance of fifteen feet or more from the outer wall of the tank. As early as 1940, McBride recognized this sound as “the principal form of intimidation” when produced by a dominant animal. Often jaw clapping was sufficient to discourage the intruder, but if not, the first male resorted to violent lashes with his flukes, sometimes throwing the offender against the side of the tank with tremendous force; or he actively attacked the intruder with his teeth, inflicting serious gashes upon the other’s flukes, flippers, back and head. Jaw clapping was also used if the female swam away for a short time, or if another female interfered. In these cases, however, the final results were rarely violent, since the partner rejoined the male immediate- ly, and an interfering female hastily left. In other situations, jaw clapping was used by non- dominant animals as a sign of displeasure and not necessarily as intimidation. Yelping. — During the more intense stages of precopulatory behavior, another sound, char- acteristic of this situation, was often heard in the tank. This sound has been interpreted to be a type of mating call (Essapian, 1953; Wood, 1953). It was a high-pitched yelp, delivered as a series of short sharp cries. The number of yelps in a series varied from two or three to seven or eight, and the series might be repeated after an interval of about fifteen seconds, or might not be repeated for a much longer inter- val, depending upon the circumstances. There can be no doubt that the male produced these sounds, since when he was in the proper posi- tion relative to the observer, the flap of the blowhole and the area behind the blowhole could be seen to move as if they were being depressed at the time that the sound was heard. No bubbles escaped from the blowhole with the production of this sound. The male yelped very rarely when he and the female were swimming quietly together. On the other hand, if the fe- male strayed away from his vicinity, or stopped to investigate another animal or an object in the tank, in the majority of cases the male promptly produced a series of the yelping cries and con- tinued to do so at short intervals until the fe- male returned to him. If she did not respond at once, he continued to yelp until she did. Im- mediately upon her return to him the yelping ceased. If another female responded, she was ignored. Since, as far as the investigators could determine, yelping was heard only during the mating season, and the selected female re- sponded, it is felt that the interpretation of this sound as a mating call is valid. Although the male consistently swam with a single female during the earliest days of the spring season, the most intense of the precopu- latory activities did not take place at this time. Such activities seemed to begin shortly after the birth of the first infant of the year, and evidence is accumulating that true precopulatory activity may be triggered by this birth. In 1953, the first infant was born on February 7, and the first dis- play of intense sexual activity on the part of the male, accompanied by yelping, was seen three days later. Since no observations were taken on the two previous days, it may have taken place even earlier. In 1954 the first infant was born on March 4, and the beginning of yelping and intense sexual aggressiveness on the part of the male began the same afternoon. Since the March date is quite late in the year for the appearance of the first infant, and con- tinuous observations were carried out through all of February, it can be seen that there was a possible connection between the two events. However, more evidence is necessary to estab- lish a clear relationship between them. The pattern of the male’s sexual activities previous to copulation was not a consistent one. Any or all of the actions described above were involved, and any or all of them could take place within an hour or two. On one day the male engaged in each of them and periods of intense activity alternated with periods when the two animals swam quietly together for an hour or more. On another day he persistently engaged in one of them, rubbing for instance, though the entire day, and displayed none of the others, resting only a few minutes at a time before resuming his activity. Initiation of precopulatory behavior by the female.— The part of the female in a paired sit- uation was less energetic than that of the male. When she was the chosen companion, it was only rarely that she engaged in any but the mildest of precopulatory activities toward him. She swam close to him, perhaps maintaining contact with him by an extended flipper touch- ing his body, and she was occasionally seen to initiate mild stroking activity. She actively co- operated with him, however, when he initiated the more intense actions, and there was one occasion when she actively took the lead. As mentioned above, during the height of the rub- bing activity on the part of the male, she often 1957] Tavolga & Essapian: Behavior of the Bottle-nosed Dolphin 15 rushed away from him at great speed, and leaped far out of the water, perhaps to escape his attentions. On such occasions, he sometimes joined her in her leaps, but more often he swam rapidly to the spot where she was expected to enter the water, where he rubbed against her even more energetically as she came down. This sequence of events was often repeated a number of times before both animals reverted to swim- ming together quietly. Since after such activity the female immediately rejoined the male and continued to swim close to him, and because copulatory activity often followed such a dis- play, it is possible that the display aroused more intense activity on the part of the male. In the case of an unpaired female, however, the situation was quite different. At various times an unpaired female was seen engaging in somewhat unusual activities, such as scraping her ventral side along the bottom of the tank on the coquina gravel, and rubbing her genital area vigorously and repeatedly over the brushes, which are set up in the tank as back-scratchers for the animals. These actions have led the ob- servers to suspect that such a female was in a state of high sexual receptivity, and in some cases this opinion has been confirmed by her subsequent actions. An unpaired female in this state often postured before the male, and some- times initiated stroking and rubbing activities. She approached him from all angles and at- tempted to maintain physical contact with him. She was often very persistent in these activities even though the male paid no attention to her. She nuzzled and mouthed his genital region, and this latter behavior was sometimes followed by an erection on his part. If this happened, she often stationed herself within his line of vision and remained motionless in the water, her ven- tral side toward him as if inviting his attention. If he did not respond, she sometimes repeated some or all of these actions. In addition, she sometimes sought out the younger of the two males, who just reaching maturity, and initiated these activities with him. Since this younger male was often rejected as a sexual partner by the mature females, he usually responded to such attentions, and if he did so, the female al- lowed him to begin copulatory behavior. Such a female was seen to hang motionless and almost vertical at the surface of the water with only her blowhole exposed, while the younger male persistently engaged in copulatory behavior. Uusually he was not completely successful. At times, the mature and dominant male was at- tracted to the scene of this activiy, and par- icipated in it himself, in which case the two males alternately approached the female. Copulatory Behavior The males exhibited erections during the more vigorous phases of sexual activity. Erec- tion and the copulatory pattern often followed the leaping display of the female or a period of strong rubbing. The copulatory activities which have been observed may be divided as follows: erection, approach, intromission, ejaculation, and with- drawal. Erection.— The penis of the Bottle-nosed Dol- phin erects suddenly ( 1 to 2 seconds) , the body of the organ curving outward and ventrally from the genital slit in which it is ordinarily enclosed, followed by the tip which springs rap- idly away from the body. The erect penis is directed craniad, and deviates from the longi- tudinal axis of the animal an estimated 20 de- grees to the left. Approach.— Because of the sinistral deviation of the penis, the male usually approached the female on her right side, and partially rolling on his right side, brought the penis close to her genital opening. He approached from the rear and parallel to her, at an angle perpendicular to her path, or at any angle between these two extremes. He swam toward her, dorsal side up, rolling over as he approached her, or he turned on his back at a distance and swam toward her in that fashion. Sometimes he approached from her left side but this was unusual. At his approach, the female did any of three things. Sometimes she swam away from the male, or continued swimming at his side, with- out changing her position or speed. In both of these cases, entry was almost impossible, and the male usually desisted after a short time. At other times, she rolled over on her left side and presented her ventral side to the male, at the same time slowing her speed. This posi- tion allowed the male to effect penetration more easily. Intromission.— (Bl.l, Fig. 5) When the male was close to the female, with his ventral sur- face approximating hers, he brought the erect penis into contact with her ventral side in the region of the vagina. This contact was often repeated many times before intromission was achieved, since both animals were swimming slowly throughout, and other animals and water currents sometimes caused the copulat- ing pair to move apart. If there was no outside interference, and the female matched her speed to that of the male, intromission was effected. In most of the cases seen by the present observers, the penis was inserted no more than to half its length into the vagina. 16 Zoologica: New York Zoological Society [42:2 Pelvic thrusts were often observed, both be- fore and after intromission was effected. They usually began as soon as the ventral side of the erected penis came into contact with the ventral side of the female. They continued only as long as the contact was maintained, and at times did not occur until penetration took place. A series of these thrusts usually lasted for two or three seconds, but the series was sometimes repeated a number of times. The intromission lasted for not more than ten seconds, after which withdrawal took place. Ejaculation— observers do not know whether ejaculation took place during intro- mission, since the animals could not be cap- tured and sperm counts made. On one occa- sion, after a partial intromission of short duration such as those mentioned above, the animals broke apart suddenly, and a stream of milky fluid was ejected into the water from the penis of the male. It is thought that this was semen. Withdrawal— Atier an intromission lasting from two to ten seconds, the animals broke apart suddenly. Each righted itself and swam normally. The penis at this time was still erect. Often the male made a new ap- proach immediately, and the above activities were repeated. Sometimes the penis withdrew suddenly into the genital slit and the two an- imals swam off in different directions. At other times the copulatory pattern was re- peated at intervals of one to seven or eight minutes for a half hour or more. The greatest part of copulatory activity by the bull took place at night or in the early hours of the morning. It is not certain whether the reason for this lay in the diversions created dur- ing the day by feeding periods or whether it was the result of nocturnal preferences by the ani- mals. In the past when there were fewer feed- ing periods the animals also tended to copulate at night. Their behavior in the wild in this re- spect is, of course, unknown. Just after dawn on March 1, 1953, a member of the Marine Studios staff reported seeing a copulation between an adult male and a mature female. Although he did not observe the two coming together, he reported that the male swam on his back beneath the female. The ven- tral surfaces were closely pressed together so that the penis of the male was not visible. This position was retained for about thirty seconds. When the pair broke apart, the penis of the male was erect. No traces of the ejaculate were seen in the water. Since the female involved produced an infant the following spring, it is possible that it was at this time that conception took place. This, of course, is far from certain and such data can be used only to narrow the time of conception to within a week or two, since the female was paired with the male dur- ing that period of time. The single instance described above was of longer duration and there was complete pene- tration of the penis. Those seen by the present observers were of shorter duration and involved partial penetration. The observers were not able to determine whether the single instance was a typical copulation and the others incomplete, or whether a short copulation with partial pene- tration is typical and the single instance atypical. There were certain complicating factors in- volved in such observations, since other animals were attracted to the site of copulatory activities. Young males, particularly, exhibited erections and competed for the female by crowding and pushing one another aside. In addition to inter- fering with the copulating animals, they also effectively obscured the view of the observers. Behavior During Gestation Since the report of McBride & Kritzler (1951) five conceptions have taken place and the resulting pregnancies have been carried to completion, each terminating in the birth of a normal healthy infant. Four of these were rec- ognized during the latter months of the gesta- tion period. These resulted in the four births during the spring of 1953. The flfth (1954 birth) was recognized as a possibility from the time of conception, and its progress recorded throughout. “Spray,” a young female, was born at Marine Studios on February 26, 1947, and was the first normal infant porpoise born in captivity. She was, from the first, a healthy animal and despite bruises and lacerations inflicted on the day of her birth by sexually aroused males in the tank, she grew and progressed normally. She was a gregarious animal, and learned early to partici- pate in all of the activities of the other animals, as well as to respond to the attentions of human beings when members of the staff played with her. Although McBride & Kritzler believed that she might initiate her first pregnancy in the spring of her fourth year, it was not until the spring of 1953 (the beginning of her seventh year) that this event took place. On February 10, 1953, the only adult male dolphin in the tank began to show a persistent interest in Spray, and continued to seek her com- pany for the following 13 days without eliciting any interest on her part. He was heard yelping on several occasions, also without any response 1957] Tavolga & Essapian: Behavior of the Bottle-nosed Dolphin 17 from her. He continued his pursuit of her, and on February 23 they were seen swimming to- gether for prolonged periods for the first time. At this time the “courtship” took on what might be considered a typical pattern, including all the types of precopulatory behavior described in the previous section. Yelping by the male was often heard and Spray responded to it during this period. In addition. Spray was often seen leaping out of the water, followed by the male who persisted in keeping in close physical con- tact with her. No erections were seen during this activity. Early on the morning of March 1 , it was reported that the bull and Spray were seen copulating. No further such incidents were re- ported, although it is probable that others took place. The two kept up their close association until about March 9, when it was observed that they were no longer constantly together. Dur- ing the several days following this, the associ- ation was abandoned except for occasional short contacts, and for the remainder of the spring season the two animals were rarely seen together, each seeking the company of other dolphins. The male was seen either alone or in the company of other females. Spray reverted to her most con- sistent occupation, swimming with one or an- other of the new infants. In the spring of 1949, a male infant was born to another of the adult females. From the time of his birth. Spray was almost constantly in his company, and took on the role of playfellow, not only to him, but to all of the later infants born in the tank. She often remained with them when their mothers were at the feeding platform, and was often seen swimming about the tank with them when their mothers were otherwise occupied. It was natural, therefore, that she should continue this occupation with the infants born in 1953, and she continued in these acti- vities well into the summer. It has been reported in McBride & Kritzler (1951 ) that during the latter half of pregnancy, the female tends to withdraw from association with others. In the case reported in that paper, two females, both pregnant, spent most of their time together, apart from the rest of the tank population. In the present case there were no other pregnant females present. Although Spray was often sought out as a companion by other members of the tank community, she showed an obvious preference from her earliest days for the company of her mother, “Mona,” and this preference has continued to the present. There- fore, when she began to withdraw from the com- pany of the other animals, she sought her mother as a companion. The first sign of this association, which was to continue throughout her pregnan- cy, was seen on June 27. On that day it lasted for only a short time, but was repeated on July 1, July 3 and again on July 7. Thereafter it became a more regular companionship, and the time Spray spent with Mona gradually increased from a few minutes a day till, toward the end of her pregnancy, almost all the time she spent in com- pany was with her mother. The remainder of the time in company was spent with “Pudgy,” an- other mature female, and her recent infant, “Mitch” with both of whom Spray had spent much time since the infant’s birth in May, 1953. Although closely observed throughout, it was seldom after July 8 that she was seen in the company of any other animal, and when such instances occurred, they lasted for brief periods only. It was also noted that in most of these in- stances, the other animal involved approached Spray, while she displayed little interest in seek- ing out other company. At this time and there- after throughout her gestation, it was also noted that her activities decreased in vigor, and she adopted habits that were slower and more lei- surely. On June 30, the increasing roundness of her abdomen was noted for the first time, and al- though the animals have been known to gain and lose weight for various unknown reasons, at this time the possibility of her pregnancy be- came, in the minds of the observers, a prob- ability. PI. II, Figs. 6 & 7, show the difference in contour between a non-pregnant dolphin and one in a late stage of pregnancy. On October 2, it was noted for the first time that Spray’s mammary glands showed percep- tible growth, and the area just forward and on each side of the genital opening, which had heretofore been rather flat and undemarcated from the surrounding region, now became round- er and fuller, as the glands increased in size. This definition of the glands became greater as term approached, although the glands never reached the size of those of some of the other mature, lactating females. On December 9, straining or body flexion was observed for the first time in this pregnancy. Straining always occurred when the animal was in the horizontal position, often when she was swimming. If the animal was resting at the sur- face she remained there, with only the blowhole out of water. If she was swimming below the surface, she usually came to a complete halt. The flukes and caudal peduncle moved ventral- ly (PI. II, Fig. 8), and as the flexion reached its maximum, the tail approached an angle of 45 degrees from the horizontal, forming an angle with the body of about 135 degrees. Often the head also bent ventrally, and as the strain 18 I Zoologica: New York Zoological Society [42:2 reached its climax, the entire body formed an arch, and the muscles appeared to be taut. This position was held for a few seconds. When the animal relaxed, the tail passed the horizontal and was raised dorsally, where it was held up at an angle almost as great as in the other direction (PI. II, Fig. 9). This recurved position was also held for a few seconds, and then the animal re- laxed again and swam off in normal fashion. The arch might occur without the recurve but the recurve never took place alone. Although body flexion has been seen in animals other than pregnant females, it is most noticeable in such females, and is considered a characteristic ac- tion in the latter months of pregnancy because of its repeated appearance. It was not until January 12 that Spray was seen straining again, but after this date it was seen often. From the beginning of February, flexions were seen to oc- cur at the rate of about one a day, gradually in- creasing in frequency until the rate was two or three a day by the middle of the month. Toward the end of February, the total number of flex- ions per day increased, and the pattern changed substantially. During this period Spray might go through an entire day during which no flexi- ons were seen by the observers, while on another day she could be observed to strain almost con- tinuously for an hour or more, about two or three minutes elapsing between flexions. A period without flexions then followed, succeeded in turn by another period in which straining was frequent. This pattern continued until just be- fore the birth took place, and as term approach- ed, the periods of flexion succeeded each other with decreasing periods between until just be- fore term, when Spray was recorded as strain- ing almost continuously throughout several hours every day. On the morning of the day before she gave birth, she was seen to flex seventeen times in thirty-seven minutes. As term approached, there were also other activities in which Spray was seen to participate, and which were not usually seen in non-pregnant animals. In a number of instances, she came down over the brushes which were anchored to the floor of the tank and applied her ventral side to them, scratching vigorously. At times she was seen to actually insert some of the bristles into the genital opening and wriggle back and forth over them. She also glided down to the piles of rocks on the floor, and as she approached them, she lowered the posterior end of her body, so that she bumped her ventral side against them with considerable force. Twice she did this with such force that large rocks were displaced. The activities could, perhaps, be correlated with in- creased vaginal irritation or tension. During the latter months. Spray’s size in- creased considerably, especially in the abdomi- nal area. Although porpoises, which are exceed- ingly streamlined animals, do not show the dis- tortions in body shape exhibited by some other animals during late pregnancy, even casual visi- tors to the exhibit were able to observe her greater girth and often commented upon it. At this time her navel, which is normally a slightly depressed scar, became slightly elevated from the body surface. Formerly a fast and graceful swimmer, she became at times almost clumsy in her body movements. Several weeks before the birth Spray was ob- served to change her habits of respiration. She frequently rose to the surface, and hovering there, respired a number of times in succession, the total time elapsed being about two or three minutes. Thereafter, she resumed normal swim- ming. A minute or two later the behavior was repeated. This might be followed by an interval of normal breathing. As term approached the unusual breathing pattern became more frequent, and consequently she came to spend most of her time swimming at the upper level of the tank quite close to the surface. Only occasionally did she descend to the floor, usually to rub herself over the brushes or rocks. During this time she was also observed to open her mouth widely at intervals, as if yawning, and she sometimes pro- jected her tongue beyond the tip of her upper jaw, as if licking it. She did this while beneath the surface of the water, and its significance is unknown. At this time there was also an increase in the number of defecations per day. This number in- creased until during the last few days before birth defecation was almost continuous. Nor- mally the animals defecate after feeding, and occasionally between feedings. Since the feces dissipate rapidly into the water, it was difficult to estimate the amount, but it is believed that the total amount of feces was also increased. It is probable that the increased numbers of both respirations and defecations were caused by pressure of the growing uterus upon the other internal organs. Feeding continued normally throughout this time and through the day pre- ceding the birth. During the last few weeks, any exertion on Spray’s part produced another visible sign that parturition was not far off. During flexions, or during the spurts of vigorous swimming in which she occasionally indulged, the vaginal opening, ordinarily a closed slit, was seen to dilate to a width of about a half inch to an inch (PI. II, Fig. 10). As term drew near, these dilations became more frequent and also more pronounced. The 1957] Tavolga & Essapian: Behavior of the Bottle-nosed Dolphin 19 opening widened to approximately two inches and the dilation sometimes lasted as long as five to ten seconds before relaxation set in. During this last month the movements of the infant inside Spray’s abdomen were easily seen. At times, distinct bulges could be seen on various parts of her abdomen, bulges which shifted from place to place, sometimes in a matter of seconds. When this happened distinct movements could be observed which were not due to the action of the mother’s muscles. When a particularly violent movement of the foetus took place, she often halted abruptly and waited in mid-stroke until the movement had stopped and the foetus was quiet again. On February 4, a bulge ap- peared on her abdomen, posteriorly and ven- trolaterally on the right side. This was persist- ent, lasting for several days. On February 13, Spray’s contours changed again, and evidently the foetus was now situated more anteriorly than had previously been the case, since bulges were seen on both sides in the anterior part of the abdomen. Later the foetus shifted more to the left, and on February 23 an area on her left side just behind her flipper was noticeably protrud- ing. This condition persisted until the time of parturition. Parturition On the morning of March 4, both observers were called to the tank at 5 a.m. The birth was observed to begin at 4:55 a.m. by a member of the Marine Studios night crew, who had been previously alerted as to the imminence of the event. Spray was seen swimming slowly and nor- mally around the tank with her mother, Mona. The flukes of the infant were protruding from the vaginal opening and were lying in a horizon- tal plane. Most of the other animals in the tank were acting normally, and only when one of them came close enough to touch Spray or the baby did she speed up or swim evasively. At 5 : 1 5 a.m. the flukes turned at right angles to their former position, and the infant was seen to be lying on its left side. At 5:20 a.m. about three- quarters of the caudal peduncle had emerged. As she swam. Spray flexed mildly and the foetus emerged still further, but as she relaxed, it re- ceded again. The situation continued in this way until 5:40 a.m., when the entire length (about 14 inches) of the caudal peduncle had appeared. PI. II, Figs. 11 & 12, show two stages of birth. At this time Spray was still swimming slowly, but in a normal fashion, with Mona. At 5:43 a.m. there was suddenly a great deal of excitement in the tank. The animals gathered in the center amid whistling and a variety of other noises. Al- I though banks of lights had been set up in the I event of a birth at night, the excited racing of the animals and the flurry of bubbles in the water which this produced, plus the darkness of the hour, kept the observers from seeing what took place during the exact instant when the infant fell free from the mother. The infant was first seen at the surface a few seconds later, against the light at the far side of the tank. The total time elapsed was 48 minutes. A minute or two later the animals had again broken up into groups, and the infant, a female, was seen swim- ming uncertainly, but energetically, between the dorsal fins of Spray and Mona. The baby’s dor- sal was inclined to the left and her flippers and flukes were soft and flabby. A number of ver- tical creases without pigment were seen along her sides. This is characteristic of newborn dol- phins. A few minutes later, at 5:47 a.m., the adult male, possibly aroused by the birth, gave chase to Mona, and her place beside the baby was taken by Pudgy, the other female which had spent much time with Spray during her pregnancy. Shortly thereafter, however, the bull turned his attention to Pudgy, and Mona re- turned to swim with Spray and the infant. These three continued together, and the two adult animals guarded and guided the new infant throughout the first days of its life. As soon as it became light, the skin of Spray’s abdomen was seen to be patterned in a series of fine wrinkles, as in crepe paper. Probably this was the result of the strong contraction of the muscles beneath, which resumed their normal tonus before the skin could do so. The crepe paper effect lasted for several days before it disappeared. At 9:15 am. Spray, who had been swimming slowly with the infant and Mona, suddenly speeded up and raced across the tank. This was the first rapid swimming she had done for several days, and as she did so the placenta emerged from the vaginal opening as though forcibly ex- pelled. There was no blood visible as it was ejected, and Spray paid no further attention to it. Immediately after the placenta was expelled, Spray slowed down again, and swam off with the baby. She did not approach the placenta again. At 10:55 a.m. the new infant began to nuzzle her mother’s side. Spray neither slowed down nor rolled over at this time. At 11:18 the baby nuzzled again, this time more persistently, but without success. Again Spray did not cooperate. At 1:00 p.m., however. Spray rolled over on her side and adjusted her speed so that her mammary region was close to the baby’s snout. The in- fant displayed no reaction to this, and did not nurse. At 1:22 p.m. the baby nuzzled again. This time Spray rolled over, and after several 20 Zoologica: New York Zoological Society [42:2 thrusts of her snout, the infant found the nip- ples, and suckled successfully for the first time (PI. Ill, Fig. 13). Another nuzzle a minute or so later was unsuccessful, but others shortly thereafter and later in the afternoon showed that the young one was locating her objective more rapidly. Within 24 hours she was seen to suckle regularly without any preliminary searching. In 1953, the year previous to the parturition just described, four other births occurred. Each of these produced a normal, living infant. All the conceptions took place in the tank at Marine Studios. There were certain features in which each differed from the one just described. During the last few days of January, 1952, Mona was observed to be spending most of her time with the adult bull porpoise, and was re- ported in copulation with him at that time. Shortly thereafter, she ate very little for a period of almost a month, although still in the com- pany of the bull. She remained with the bull through most of February, although observa- tions indicate that during the latter half of Feb- ruary it was Mona who sought the bull’s com- pany, rather than the reverse. It was only in November, 1952, that her in- creased size attracted attention to her condition. On February 6, 1953, Mona’s abdomen showed a shallow, concave area ventrally between the region of the navel and that of the mammary glands, which could be seen only when viewed from the side. Anterior to this region the dis- tension of her abdomen was conspicuous. This concavity did not appear on Spray, and its signi- ficance is not known, though it may be con- cerned with the position of the foetus in the uterus. Some time during the night of February 6, a female infant was born to Mona, and when first seen the next morning, she was swimming between the dorsal fins of Mona and Spray. By that time the dorsal fin of the infant was already stiffened and upright, so that it could not be de- termined to which side it had inclined. Also, the placenta had been passed, and the baby was suckling regularly. The infant showed five deep transverse creases on the left side, six on the right. All of these were much lighter than the rest of the animal. The skin of Mona’s abdomen showed the fine wrinkles which are character- istic of the new mother. Although “Mrs. Jones,” a mature female, was not observed to spend more than a day or two with the bull in the spring of 1952, abdominal distension in November of that year made it evident that she, too, was pregnant. On Feb- ruary 14, 1953, she was reported to be strain- ing and dilating her vaginal opening. During this month she spent most of her time either alone or with Pudgy. On three occasions before the birth, Mrs. Jones showed flexion and on two of these dilations were also observed. She showed no concavity in the posterior abdominal area. On February 23, she gave birth to a female infant during the afternoon. When the infant was first seen, shortly after 4 p.m., its dorsal was inclined to the right. The mother showed the characteristic wrinkling of the abdominal skin. On February 11 and 18, 1952, the bull’s courtship activities were observed to be directed toward “Susie,” another female, and she re- sponded to his yelps. In November it was noted that Susie was pregnant. On February 6, 1953, the shallow concavity was seen on her abdomen. This persisted until she gave birth. During the latter months of her pregnancy, Susie was not accompanied by another female, but swam alone and avoided the company or contact of any other animal. She flexed intermittently during the month before parturition and her genital slit dilated when she strained or when she broke into a rapid swim. She was also seen to defecate often and copiously during the last few days be- fore birth. Early on the morning of February 27, she gave birth to a female infant, whose dorsal fin was inclined to the right. Susie was not accom- panied by any other female on the morning of the birth in spite of attempts made by Pudgy to remain near her, and she cared for the infant without the aid of any other animal. This was quite unusual, considering the conduct of all the other mothers on which data are available. Late in April, 1952, Pudgy was seen in the company of the bull for several days. It was not until April, 1953, that her pregnant condition was suspected. Although she was the stoutest animal in the group, her weight was well distrib- uted over her entire body, and the roundness of her abdominal region was not outstanding. Her previous pregnancy had gone unsuspected until May, 1950, when she gave birth to a male infant. Therefore, when her size increased in April, 1953, and especially when she avoided the company of other animals, she was again suspected of being pregnant, despite the absence of other symptoms. On May 5 and 6, Pudgy was seen in the com- pany of Spray, but with no other animal. On the 6th, she flexed a number of times. On the morning of the 7th, she strained almost continu- ously, and between flexions rose to the surface where she hung, respiring frequently in a pattern similar to that shown by Spray. At this time the shallow concave area was seen on her abdomen. During the afternoon of the 7th, Pudgy gave 1957] Tavolga & Essapian: Behavior of the Bottle-nosed Dolphin 21 birth to a male infant. The birth began at 2:30 p.m. when two or three inches of the baby’s flukes were seen to emerge immediately after a strong flexion. At this time Mrs. Jones was swim- ming beside Pudgy. The following outline, ex- tracted from the notes taken on the day of the birth, give a clear picture of the events during this parturition. 2:30 p.m. Pudgy arches and recurves. Suddenly there appear approximately 2-3 inches of the baby’s flukes. Pudgy contracts abdomi- nal muscles. 2:35 p.m. Pudgy swimming slowly in center of tank. 3-4 inches of flukes showing. Flexion. 2:37 p.m. Flukes completely visible. Horizon- tal, but tips curled ventrally. 2:38 p.m. Pudgy swimming close to floor of tank with Jones. She flexes and recurves. Other animals nosing around at distance of 2-3 feet. 2:40 p.m. Pudgy almost scrapes the bottom with her ventral side. Slows down till barely moving. 2:43 p.m. 2 inches of caudal peduncle showing. Pudgy maintains steady unhurried rate of speed. 2:45 p.m. 3 inches of caudal peduncle emerged. Jones with Pudgy. 2:47 p.m. Pudgy flexes. Spray swims just behind and below Pudgy, her head pointing up- ward. 2:50 p.m. 4-5 inches of caudal peduncle have emerged. All other animals quiet. Pudgy strains. 2:51 p.m. Foetus slips back an inch or two. 2:55 p.m. Foetus has turned, now lies on its right side. 2:57 p.m. Pudgy swimming close to surface and wall. 3:02 p.m. 5-6 inches of caudal peduncle have emerged. 3:07 p.m. Feeding show in progress. All animals rushing around. 3:08 p.m. Diver feeding underwater. All animals feeding except Pudgy, who swims in cir- cles away from points of disturbance. Pudgy alone. Infant’s tail hangs limp. 3:15 p.m. Still 5-6 inches of caudal peduncle showing. 3:22 p.m. 7-8 inches of caudal peduncle show- ing. Pudgy swims slowly, flexes. 3:27 p.m. Foetus slides partially out and then recedes as Pudgy flexes and relaxes. Pudgy is alone, swimming short circles in N. E. corner of the tank. 3 : 36 p.m. Pudgy now with Jones. Still 7-8 inches of caudal peduncle are all that can be seen of infant. 3:38 p.m. Pudgy flexes again, and recurves strongly, swimming partially on her right side. At this point, less than an inch of the baby’s dorsal becomes visible, and in the next 4-5 seconds, the foetus is expelled, followed by a stream of blood. Contrary to previous reports (McBride & Kritzler, 1951) Pudgy did not whirl as the infant was expelled, but continued swimming on and up toward the surface. The umbilical cord broke as it pulled taut. As in the case of Spray, all the dolphins gathered at the point where the birth took place, whistling and producing a clamor of other noises. Mrs. Jones pushed the baby out of the way of the other animals. Then she placed her snout beneath it and gave it one thrust toward the surface, where it took its first breath, 5 or 6 seconds after its birth. It is well known that infant dolphins are able to reach the sur- face without assistance immediately after birth (McBride & Kritzler, 1951), and there is no reason to suppose that this infant was an ex- ception, especially as it was seen swimming immediately after the nudge. As a normal infant, it would undoubtedly have reached the surface for its first breath unaided, and the push given by Mrs. Jones was of secondary importance. Spray positioned herself on the other side of Mrs. Jones and the infant swam between their dorsal fins. Pudgy trailing behind. This situation con- tinued for the next two or three minutes, when Pudgy moved into position beside the baby, and Mrs. Jones dropped out. The infant was ener- getic as it labored to keep pace with the adults. The baby’s dorsal was inclined to the left. The total time elapsed in this birth was 68 minutes. Behavior of Mother and Infant As stated by McBride & Kritzler (1951), and confirmed in later cases by the present observers, the newborn dolphin remains very close to its mother during the first month. During the first few days while its swimming is still uncertain, particularly in regard to direction, the infant will often stray slightly, but the mother invariably swims toward it and pushes it gently, guiding its direction until it is close to her side again, usually next to her dorsal fin. This position close to her dorsal fin is maintained by the infant during its first month or two, and gradually the baby then becomes accustomed to what the observers con- sider the typical rest position, under the mother’s tail, with the top of its head lightly touching her abdomen. During the first few weeks, at least, the infant sleeps in the position next to the mother’s dorsal fin. 22 Zoologica: New York Zoological Society [42:2 The mother does not allow the young infant to come into close contact with other animals, and when one approaches, will swiftly sweep the young one away with her, usually placing it on the side away from the intruder. After about two weeks the infant begins to swim about its mother and vary its position in relation to her. It may dart around her head, slip under her tail, or swim beside her at a dis- tance of a foot or two. However, if the young one increases the distance between them to as much as ten feet, the mother will immediately swim toward it and reorient it toward her. At feeding time, the new mother will guide her in- fant to a position a few feet from the feeding platform, where she leaves it swimming in a small circle. She will approach the feeding plat- form and secure one fish, then rush back to the baby. Often she will not approach the platform again, but spend her time with the infant on the side of the tank opposite to the feeding platform, maneuvering herself and her infant away from dolphins who race for food thrown on the sur- face of the water. Because of this situation, spe- cial feedings are often provided for new mothers, who will accept fish thrown to them while swim- ming with the offspring after the regular feedings are over. The four young ones born in 1953 prospered and grew quickly. Nursing began promptly and all the calves were able to find the nipples rapidly within 24 hours after birth. They suckled in the general pattern noted by McBride & Kritzler (1951), three to nine times within as many minutes, followed by an interval of 10 to about 25 or 30 minutes. At about two weeks of age all the babies began to move away from their mothers, and were re- peatedly retrieved. As mentioned under Spray’s behavior during pregnancy, she spent a good deal of her time in the company of the young dol- phins. From the time they were less than a month old, she was seen swimming with them and their mothers, and often swept away each of them with her in a rapid trip around the tank. If any other animal came close to one of the infants, the mother promptly showed her displeasure by slap- ping the intruder with her flukes, or by swiftly removing the calf from his vicinity. However, each mother soon allowed Spray to escort her infant without evidence of concern, and Spray usually returned the infant to its mother within a minute or two. When Spray had been accepted as a trusted escort for the infants, all the mothers occasion- ally left their calves in her company while they were occupied at the feeding platform, and she was often seen surrounded by two, three or four youngsters at this time. In addition, one mother could sometimes be seen guarding a group of infants while the others fed (PI. Ill, Pig. 14) . The mother who had received her food then returned and relieved the escort so that she could get her share. At the end of feeding time, each mother retrieved her own calf, and there was never the slightest hesitation about picking her own in- fant from the group. No mother was ever seen to make a mistake and choose the wrong infant. By the time they were six weeks old, the young dolphins were swimming away from their mothers regularly, and associating with each other or some of the adult animals. Their mothers, by this time, had relaxed their vigilance somewhat, and the infants were no longer kept within a few feet at all times, although they were still kept under close watch. A short excursion or play period was usually followed by a longer rest period spent under or close to the mother. The infants began to show an interest in the objects and fishes in the tank and were seen to investigate, and if small enough toss about, any new object which came to their attention. If the object was a familiar one in the tank, a bit of rock, a turtle, or a small fish, the mothers watched nearby while the calf investigated, and sometimes joined it. However, if the object was an unfamiliar one, such as a new animal, or an object dropped by a visitor, the mother displayed agitation and set up a loud and continuous whistling while steering the young one away, and only when she had removed the baby from the disturbing object did she resume her normal behavior. This extreme vigilance and care were typical of the dolphin mother during the first months of her infant’s life, and continued in somewhat lesser degree throughout at least the first year. . In direct contrast to the above situation. Spray 1 showed a distinct lack of care toward her infant born in 1954. At first the situation seemed nor- * mal. The baby began to nuzzle her mother’s ‘J side, as stated above, five hours after birth, and f it was 2 hours and 27 minutes after this initial ' | effort that she began to suckle. After this, nurs- J ing proceeded successfully, and the baby suckled in the same general pattern as the above infants. The baby seemed to be in good health and swam j strongly from the first. After a few days, however, it was seen that .( she was not growing as other infants had done. Newborn dolphins lose their thin appearance i quickly, and after a few days their necks become j less distinct and their abdomens become round- ed. This was not true in the case of Spray’s in- fant. In addition. Spray did not care for the baby 1957] Tavolga & Essapian: Behavior of the Bottle-nosed Dolphin 23 as vigilantly as other mothers had done. She remained at the feeding platform throughout almost the entire feeding, returning to the calf only once or twice during this five- or six-min- ute period. The infant was alone during this time, except for the momentary visits of her mother. On March 13, Spray’s infant attempted for the first time to swim away from her mother. Spray and Mona swam after her, but did not hurry to her side. This, too, was in contrast to the conduct of other dolphin mothers. On March 14, the baby persistently left Spray and swam among the other animals. If she strayed more than about twenty feet away. Spray I brought her back, but if she stayed within this distance, no attempt was made to restrict her movements. Twice, however, when she strayed to the other side of the tank. Spray went to her. Then Spray turned on her back, positioned the baby on her chest between her flippers, and rose to the surface, pushing the young one out of water. Other dolphin mothers have been seen ; to do this when their young infants swim eva- sively away. The effect is to temporarily immo- bilize the young one. On March 16, Spray left the baby entirely alone during the feeding. On this date, the in- fant nuzzled its mother’s side immediately after |i the feeding, but several minutes elapsed before Spray slowed down to allow her to suckle. On i' March 17, Spray allowed the infant to swim around alone for several ten-minute periods. On i that day, the baby appeared to be nursing almost [i continuously, only very short intervals elapsing ! between periods of suckling. On March 18, the baby left Spray several ? times, and sometimes entered another group of ' dolphins. Spray did not follow her, but allowed her to work her way out of the group alone. During the morning of March 19, the infant i alternately remained close to Spray and per- I sistently suckled, or wandered far away from her, while Spray was occupied with other ani- mals. At these times. Spray did not bring her back or swim after her. The baby was seen to suckle several times at 9:30 a.m. At 12:15 p.m. it was reported to the observers that the infant was in distress, and both hastened to the tank. Unfortunately, by the time they ar- rived, a minute or two later, the infant was dead, and it was necessary to gather reports from eye- witnesses, and to attempt to fit together what- ' ever facts could be supplied by them into a co- herent account. Early on the morning of the 19th, it was re- ported by an employee of Marine Studios that Spray and another unidentified dolphin, prob- ably Mona, pushed the baby down to the floor of the tank. By moving their bodies over her, they temporarily prevented her from rising to the surface. There was no apparent reason for this action, and its significance is not clear. This report was substantiated by two other employees wno happened to be looking into the tank at the time. Smce both observers had seen the infant following this incident, and she appeared to be in good condition, no particular importance was attached to it at the time. At about 12:06 p.m., the infant swam to the surface to breathe, apparently under her own power. Spray met her as she came down, and the two rose to the surface again. Spray pushing the young one up ahead of herself. As the baby sank, the second time hubbies of air streamed from her blowhole, and she did not swim again. Spray pushed her about the tank, nudging her up toward the surface every minute or so, but the infant did not breathe again. The baby’s mouth opened and her tongue hung out to the side. The above account came from another employee who was watching through the port- hole at the time, and it was at this moment that the authors arrived on the scene and it was ob- served that the baby was dead. A diver entered the tank immediately to bring her out. Spray actively avoided the diver and pushed the dead calf ahead of her, but the body was finally taken from her and removed from the tank. From the time the infant was first seen to be in distress to the time when she was removed from the tank, only 12 minues had elapsed. The body was taken to the laboratory and an autopsy was performed. This revealed that the calf was extremely thin, her stomach was completely empty, and she had suffered a fractured left mandible. Her lungs contained water, and it was therefore assumed that the immediate cause of death was drowning. Judging from the extremely fresh condition of the fracture, it is believed that the accident that caused it must have taken place within 24 hours of her death. At the time of her death at the age of 15 days, the infant measured 42.5 inches in total length, and weighed 30 pounds. Her weight may have been slightly less than normal for her age, since she was very thin. She pos- sessed eight hair follicles on the left side of her snout, and six on the right side, but only one vibrissa was visible, in the last hair follicle on the right side. No structural ahnormalities were found at the autopsy. Whatever the reason may have been. Spray’s infant was not receiving the usual amount of care that is given to newborn dolphins. Her mother left her for unusually long periods of time, and failed to remove her from situations in which she might have been injured inadvert- I 24 Zoologica: New York Zoological Society ently by the other animals. On March 13, the baby swam directly under the platform while the other animals were feeding, and was struck by at least one dolphin falling back into the water. On March 18, while Pudgy and the bull were engaged in some of the more vigorous courtship activities, the baby swam between them, and may have been lashed by a tail be- fore she found her way out. In neither of these instances did Spray make any attempt to rescue her young one, and it is possible that the latter incident was the cause of the infant’s fractured mandible. Two of the 1953 infants died before the end of their first year, one by accidental drowning, and the second because of an unknown infec- tion, on which an investigation is being carried out at the present time. The remaining two young dolphins born in February, 1953, were weighed and measured in September of that year, and their lengths at that time were 68 and 69 inches, their weights 146 and 141 pounds. At the pres- ent writing they are 16 months old. They still nurse regularly, usually after feeding time, al- though it is thought that they are about to end the nursing period, since they occasionally skip a nursing. They began to take small pieces of squid in August, 1953, and early in December they were observed to swallow fish for the first time. Since then both young dolphins have de- veloped excellent appetites for fish and eat them regularly at feeding times. The nursing period usually ends gradually over a period of months as the young animals show an increasing prefer- ence for fish. They race the adults for the fish even after their appetites have been satisfied. They have often been seen to rush for a fish thrown to them, then drop it to race after an- other on the other side of the tank. They seem to make a special effort to take fish from just in front of another animal, and often do so. They may play with the fish so taken in a num- ber of ways. They will drop a dead fish in front of the rocks where the small tank inhabitants hide, and just as they come out to eat, the young dolphin will snatch the fish away again, only to repeat the performance a minute later. The young animals indulge in this type of sport only after they have eaten and are satisfied, but often the fish are caught by a still hungry adult and the youngsters are deprived of their play- thing. Every activity of the young animals now in the tank shows them to be normal healthy dolphins. Discussion The dolphins are a specialized group of mam- mals adapted in both structure and function to a life lived completely in the water. Superficially, [42:2 they resemble fish. Their forelimbs are flippers, without use as grasping organs, and probably serve mostly as balancers, or in steering. They possess no hind limbs. Their tails, furnished with broad horizontal flukes, are excellent propulsive organs. Their streamlined form enables them to move gracefully and rapidly through the water. Since they were derived from land mammals, they bear their young alive in typical mammalian fashion, and suckle them for an extended period of time. They also display mammalian care for the young. Since they possess no limbs which can be used as manipulative organs, their mouths, which are provided with a set of efficient teeth for holding their prey, serve also to grasp ob- jects and to inflict wounds in their own defense. The powerful muscles of the tail also aid in defense. Together with these features, the dolphin pos- sesses a highly developed and exceptionally large brain, of which the largest part is cerebrum. The brain of one nine-months-old Tursiops weighed 1230.4 grams, and its volume was 1225 cc. The length of this male specimen at the time of its death was 64 inches, and its weight was 137 pounds. The large brain and the probable resulting high level of adaptiveness might be expected to lead, in animals with manipulative limbs, to complex varieties of behavior. The lack of these limbs results in the ingenious use of flukes and flippers. These can be delicately adjusted to pro- duce minute changes in movement and position. Such changes enable the animal to make the best possible use of the only grasping organ it possesses, the mouth. The remarkable use of these organs under the control of a well-devel- oped central nervous system leads to behavior patterns which are peculiar to these animals and their close relatives. No accurate observations on the precopulatory and copulatory behavior of dolphins were pos- sible before the tanks at Marine Studios were established, and therefore such observations as were available were recorded by whalers and transmitted to interested scientists, or such sci- entists reported their own observations from the decks of ships. Lillie (1910) reported that whal- ers saw two adult sperm whales (Physeter) rush together near the surface, then turn up ver- tically at the end of the rush. They believed that they had seen copulation, and said that the males were exhausted and easy to capture after these activities. Morch (1911) observed coition in blue whales during the summer, but gave no details of their behavior. Tomilin (1935) noted what he considered to be sexual attachment in various species, and differences between the species and 1957J Tavolga & Essapian: Behavior of the Bottle-nosed Dolphin 25 between individuals in the same species. He stated that this attachment sometimes reached great intensity and was strongly shown especially It one of a pair was wounded. He gave no details of copulation. Hamilton (1945) reported that while sailing off the coast of Brazil, he saw two Steno swimming on their sides with their genital areas pressed together. They were slightly sep- arated anteriorly, and one’s mouth was slightly open. This posture was maintained while they remained in sight, which, according to the au- thor, was “not for very long.” The ship was moving at 10 knots, and the animals maintained the speed of the ship while in the above position. Such fragmentary observations do not give any indication of the complex pattern of pre- copulatory and copulatory behavior in this group of animals and it was not until they could be observed under water for extended periods of time that details could be reported. Although many of the events leading to copulation are reported here, the sequence is variable, and it is by no means certain that copulation can be pre- dicted following any particular stage. The ob- servers have seen much copulatory behavior, some of it at least partially successful. In at least one instance an ejaculation of semen has been seen after withdrawal of the penis from the vagina of the female. Sudden erection and re- traction of the penis occur most often during the spring months both with and without true copulatory activity, so that retraction is no indi- cation that ejaculation has taken place during a partial intromission. The one instance in which full penetration took place occurred before seven in the morning and was not seen by the authors. The employee who witnessed it is a reliable ob- server who has made a hobby of watching the animals over a long period of time. It is not known whether the copulations involving partial penetration over a short time are as effective as those with full penetration extending over 30 seconds. Although the authors have spent a number of hours watching during the evening hours and also in the early morning, they have not seen the longer copulations. Since there is no provision for adequate lighting at night, ob- servations during this time are impossible. Per- haps long-continued early morning observations over a long period of time will provide more data. Although McBride, in 1940, recognized the jaw clapping of the bull as “the principal form of intimidation,” the entire use of this sound was not recognized until much later. As late as 1948, McBride & Hebb again stated that this sound was used by a dominant animal toward sub- ordinates. The observations of the present au- thors indicate that this is not always the case. and the jaw clap has, under direct observation, been made or attempted by almost every animal in the tank. It has occurred under a number of circumstances but these were always such that the jaw clapping animal appeared to be displeased, dissatisfied or annoyed. Pudgy clap- ped her jaws when the bull, who is dominant to her, made sexual advances to her and she was not responsive. All the mothers clapped their jaws when another animal interfered with or approached their young. Spray has been seen to clap her jaws when the young male swam between her and the infants in her company. The younger animals, such as Spray and the young male, produce a much less resounding clap than Pudgy and the bull, but their efforts were unmistakeable, and as they gained more experience and practice, the sound produced be- came louder. Apparently this evidence of dis- pleasure is limited to the adults, since the 1953 youngsters have not as yet attempted it. Townsend (1914) gives some evidence to show that Tursiops breeds yearly in the wild. Of 51 specimens taken in November, 34 were fe- males. All the females taken were in milk at that time. Although he does not state how many of the females were killed, all those that were killed were not only in milk but contained young. Un- fortunately, the age of the suckling infants could not be determined. Therefore, it is not known whether all infants in the wild suckle for 18 to 21 months or more, or whether there is a yearly reproductive cycle. Since young dolphins at Ma- rineland have been known to suckle for periods up to 20 months, it is possible that Townsend’s data would support the idea of a two-year cycle in the wild. At Marine Studios, no female has as yet pro- duced young two years in succession. Several animals have produced young at two-year in- tervals. There is in the tank only one fully adult male, and therefore no significant competition for females. There is the possibility that females unaccompanied by newborn young are prefer- red partners. It is to be hoped that some future study may show whether the two-year cycle seen at Marine Studios is unchanged from the wild condition. McBride & Kritzler (1951) believed that the female Bottle-nosed Dolphin becomes sexually mature at four years of age and may bear her first infant in the spring of her fifth year. Al- though this may be theoretically true, the only evidence now at hand indicates that the first in- fant may not be conceived until several years later. Spray, the only animal that was born and reached sexual maturity in captivity, did not conceive until she was six years of age, and 26 Zoologica: New York Zoological Society [42:2 delivered her first infant just after her seventh birthday, although she was known to be in ex- cellent condition throughout her life. It is, of course, not known whether she would have con- ceived earlier in the wild, and further evidence on the age at which sexual maturity is reached will depend on the time at which the young females now in the tanks first conceive. There has been much speculation on the length of the gestation period in these animals, and estimates which have run from 10 to 14 months have in the last few years narrowed to 11 to 13 months. It is now possible to state with a reason- able degree of accuracy that the gestation period is very close to 12 months. Spray’s association with the bull lasted for a period of two weeks, from February 23 to March 9, 1953, and her infant was born on March 4, 1954. Similarly, the associations of Mona and Susie with the bull were almost exactly a year before the birth of their young ones. Although Pudgy was observed to be in the company of the bull in late April, 1952, and her infant was born in early May of the following year, the authors believe that this discrepancy comes within the limit of error, and that the twelve-month period is the most valid one. McBride & Kritzler (1951) noted the with- drawal from association with other animals shown by pregnant females, especially during the latter part of pregnancy. They also mentioned the tendency of two pregnant females to associ- ate with each other. These tendencies have been confirmed in the present paper. In 1952, Pudgy and Mrs. Jones, although they conceived several months apart, spent most of their time together during their pregnancies, though they had not been close companions before this time. When Pudgy’s infant was born, Mrs. Jones was the animal which escorted the infant during the first few minutes of its life, and she and Spray kept the infant out of the way of other excited ani- mals, and accompanied Pudgy and the new baby for the first few weeks. When Spray’s infant was born it was Mona, her mother, and Pudgy who performed these functions alternately, so that there was almost always another mature female swimming beside the baby. Such protection proved to be important, since the records show that the birth of a young one seems to stimulate the adult male to begin or renew his courting activities. Often these are directed toward the new mother, whose evasive movements become frantic in her efforts to protect her newborn infant. In 1947, when Spray was born, there were several adult males in the tank, and in their attacks on the new mother and other females and each other. Spray’s skin was badly lacer- ated by their teeth. Since then, these competing animals have been removed, but the danger to the new infants still exists, though to a lesser degree. In one case, however, the new mother was not accompanied by another female. Susie is an ex- tremely retiring animal, and remained alone from the time she was first captured and placed in the tank. She conceived in February, 1952, and remained alone throughout her pregnancy, a situation which was unusual enough to war- rant the attention of the observers. When her infant was born in February, 1953, she was not accompanied by any other animal. Immediately after the birth Pudgy made persistent efforts to swim with her and remain near her, but Susie avoided Pudgy as diligently as she did all the other animals. Her behavior, though unusual, was consistent, since both before and after her pregnancy she showed no inclination to asso- ciate with any other dolphin. She was captured in 1949 and gave birth three days later to a male infant, an event which took place in the receiv- ing tank. This tank contains very shallow water, and the environment was quite unnatural to her. Her timidity may have stemmed from that time. The young male has now almost reached ma- turity, and the two ceased their association prior to the arrival of the new infant. It took more than a year of training before Susie took her food at the feeding platform. Either her inherent timidity or the shock of her arrival and the sub- sequent birth may account for her unusual conduct. The young of members of the Delphinidae are usually born tail first. In every birth at Marine Studios except one (Spring, 1955) the infant emerged in this manner. In the earliest papers describing the birth of small cetaceans, (James, 1914), the author as- sumed that the normal presentation was head first, and that it was mishandling or other im- proper treatment which caused the tail presen- tation. The birth described by James took place in Phocaena phocaena and was a stillbirth. James mentioned mishandling as the reason for the tail presentation, and the tail presentation as the reason for the stillbirth. It is now known that although mishandling my lead to stillbirth, the reasons are other than those described above. Caudal presentation has been described in de- tail by a number of authors (Wislocki & Enders, 1941; Slijper, 1949). McBride & Kritzler (1951) discussed it in connection with data available to them from Marine Studios, and all agree that the foetus is, or becomes, oriented in such a way that the head is pointed toward the tubal end of the uterine cornu, while the tail extends toward 1957] Tavolga & Essapian: Behavior of the Bottle-nosed Dolphin 27 the vaginal opening. Sleptzov (1940) observed 635 embryos of Delphinus delphis in situ during various stages of development, and came to the conclusion that during early embryology the foetus may lie in either direction. Later the head tends to become oriented toward the tubal end of the cornu, and by the time reversal is no longer possible, because of size, the vast majority of embryos are situated in this way, so that tail presentation is inevitable. The problem remains, however as to why the orientation takes place in this manner. Slijper (1949) , citing Williams & DeSnoo, discusses the dangers attendant upon breech presentation in mammals which produce foetuses with com- paratively large heads, and believes that the com- pression of the umbilical cord between the pel- vis of the mother and the head of the infant which occurs in breech presentation may cause asphyxia if the infant is not quickly extracted. Although Odontocetes possess no pelvic girdle other than two small pelvic bones, the pelvic region contains large masses of tough fibrous connective tissue which could exert considerable pressure, and it is possible that such compression may be exerted in the case of these animals. Whether or not this is true, it does not explain the orientation of the foetus in the uterus, but only illustrates an effect of such orientation. Slijper’s discussion of the role of uterine con- tractions in the orientation of the foetus is more pertinent to the question. Mammals other than primates show, during pregnancy, uterine con- tractions of a peristaltic nature which begin at the tubal end of the uterus, and Slijper believes that the head should be impelled in the direction of the cervix under the influence of these con- tractions, in mammals with a small head and long neck. Since this formation does not apply to cetaceans, he considers that the body form of the foetal cetacean might favor caudd presen- tation. Schumann (1914) discusses the possible causes of almost universal cephalic presentation. He states that in multiparous animals the foetuses lie indifferently, in either breech or cephalic presentation, whereas in unipara, the head pre- sents in 90 to 98 percent of the cases. In man the foetal head is heaviest and gravitates to the most dependent position, nearest the cervix. In uni- parous quadrupeds, however, the hindquarters and trunk of the foetus are its heaviest parts, and consequently sink to the lowest portion of the uterus, which with its cornua lies low in the abdominal cavity. The cervix rises to meet the vagina. In dolphins, which are uniparous, the uterus lies in a situation similar to that seen in the quadrupeds, so that the heaviest portion of the foetus would tend to gravitate to the lowest por- tion, the tubal end of the cornu. These animals, however, possess no hind limbs, only remnants of the pelvic girdle, and the tail is long and slender. The head, which is relatively very large, and the pectoral region are heavier than the hindquarters. Under these circumstances, the gravitation of the heaviest portion toward the tubal end of the uterus would result in tail presentation. The present authors believe that this situation may be of considerable influence in the determination of tail presentation in the dolphins.^ It is necessary, however, to consider the fold- ing of the uterine cornu in evaluating this ex- planation. The non-pregnant dolphin uterus is not folded. Presumably the fold occurs when the foetus is too long to extend to its full length in the abdomen of the mother. At this time it would also be too large to turn easily on its own axis. Examination of the pregnant uterus of many delphinids has shown that the foetus ap- pears, at first glance, to have its head oriented toward the vaginal end of the uterus. Closer ob- servation, however, has usually shown that the head is enclosed in a fold, which is oriented toward the caudal end of the mother, but that on following this fold in the direction in which the head is pointing, it is found that the uterus con- tinues from this point toward the Fallopian tubes. Following the body of the foetus in the opposite direction reveals that it folds close to the origin of the tail, and the tail occupies the lower end of the uterus, and extends toward the vaginal opening. These facts may seem to contradict the weight theory outlined above, but if we consider that the head of the foetus, as in most mammals, devel- ops more rapidly than the other parts, and be- comes the heaviest part of the foetus in an early stage of development, before the fold occurs, it seems possible that both events might occur in sequence, with the usual observed result of tail presentation. It is unfortunate that Sleptzov (1940), who had at his disposal more than six hundred foetuses in all stages of development, did not record more fully the sizes, weights and developmental characteristics of these animals, together with the condition and configuration of the uteri in which they were found. Although McBride & Kritzler (1951) stated that in all cases observed by them the dorsal fin of the newborn dolphin was folded to the right, ^Recently E. J. Slijper (1956. Some remarks on gestation and birth in cetacea and other aquatic mam- mals. Hvalradets Skrifter. Scientific Results of Marine Biological Research, No. 41, pp. 1-62) reported addi- tional data and conclusions concerning tail presenta- tion in cetacea which confirm those given above. 28 Zoologica: New York Zoological Society [42:2 at least two cases were seen by the present ob- servers in which the dorsal was folded to the left. In one of these cases the infant was seen to be lying on its left side when only partly ex- truded from the mother’s body. We do not know whether this correlation holds true in all cases. The symmetrically placed vertical creases on each side of the trunk at birth are not neces- sarily six in number, as stated by McBride & Kritzler. In the group of births recorded here, the creases have numbered five, six or seven, and there were sometimes more on one side than on the other. Also, one or more may be continuous over the dorsal part of the animal. They lacked pigment and were easily visible for several weeks after birth, when they gradually acquired pig- ment and disappeared. The young dolphin to which Spray gave birth, and which died at 15 days of age, showed two creases which were not continuous over the dorsal side but which ap- peared on both sides, three which were con- tinuous over the dorsal, and two ventral creases, which were not seen previous to her death be- cause of the lack of pigment on the ventral side of the body. It may be significant to note that the dorsal fin of the infant which died at 15 days was slightly inclined to the right when the body was removed from the tank. At birth it was distinctly folded to the left. It has been noted by many ob- servers that the dorsal fins of all the adults in- cline to the right, and it is thought that the swimming tendencies of the animals may be the cause of this phenomenon. Since the animals in the tank usually swim in a clockwise direction through a counterclockwise current, the force of the water against their dorsal fins may be the cause. It should be noted, however, that this inclination in no way hampers their swimming in any other direction. It seems significant that the tilt to the right should have taken place so quickly in an infant whose dorsal was inclined sharply in the opposite direction only 15 days before. The mother of a newborn dolphin has never been seen to eat the placenta, which is expelled several hours after birth. Since eating this organ appears to play a significant role in the behavior of so many other mammals, this point seemed to be of particular interest. In all the cases on record at Marine Studios no mother dolphin has even approached the placenta after expelling it, and unless removed by divers and preserved, the organ is torn to shreds by the turtles and fish in the tank. Although McBride & Kritzler (1951) reported that a mother dolphin, immediately after the infant’s body falls free, will whirl about and thus snap the umbilical cord, this does not ap- pear to be necessary in every case. At least one female in the present group spurted forward in a straight fast swim as the baby was released and the cord snapped quickly as it pulled taut. The survival value of the whirl, in case the infant fails to reach the surface, is obvious, but the pattern does not seem to be an invariable one. The problem of suckling in the dolphin has long been of interest to observers, and the mech- anism has been fairly well worked out. Milk is poured from the mammary glands into large sinusoids, which in turn open to the nipples, situated on each side of the genital slit. When the infant grasps the nipple, abdominal contrac- tions of the mother pour the secretions into the baby’s mouth. The question remained as to whether the stimulation of the nipple by the in- fant was necessary in order to produce a flow of milk, or whether the contraction of the ab- dominal muscles was under the voluntary con- trol of the mother. An instance that took place in the spring of 1954 shows that, at least in ex- traordinary circumstances, the flow of milk is entirely under the mother’s control. On February 22, 1954, a nine-months-old young dolphin, born the previous May, died at Marine Studios of an unknown skin infection. At the time of his death, he was still getting the main part of his food from his mother, and was only beginning to show an interest in other food. On the day before his death, he suckled very few times, and his mother’s nipples protruded from their slits, apparently pushed outward by the engorged sinusoids. On the day of his death and for sev- eral days afterwards this condition continued. Twice in the first several days after the young one’s death, the mother was seen to turn on her side in the water and visibly contract her ab- dominal muscles, raising her tail as she did so. As she did this, a stream of milk spurted from her nipples. This stream was ejected with a force strong enough to send it a distance of more than two feet before it dissipated into the water. It is not known whether such a feat can be accomplished in circumstances where the sinusoids are not filled to excess, but it shows that a female dolphin can eject milk without stimulation of the nipple under certain con- ditions. Several Russian authors, particularly Khvatov (1938) and Sleptzov (1940) have reported on births of Delphinidae (mainly Delphinus del- phis) which they observed in the Black Sea. In most of these births, the animals concerned were captured dolphins, and were seen while in seines being drawn toward shore. The rest were in shallow areas near shore. It is possible 1957] Tavolga & Essapian: Behavior of the Bottle-nosed Dolphin 29 that the births may have been abnormal in some cases, particularly in the case of the captured animals, which could have been shocked and frightened by the procedure of capture. Also the conditions for observing the births were un- favorable, since many animals were milling about in the seine, and in all cases the births were observed from above the surface of the water. In spite of these difficulties, there are descriptions of reverse flexions, leaps, dives, and in some instances, the moment of birth when the infant fell free. In all cases the infant was born tail first, but no description of the moments immediately following birth is recorded. In the most complete description, the birth took one hour and fifteen minutes from the time the author (Sleptzov, 1940) first saw the tail of the infant protruding from the mother, and the in- fant’s dorsal fin was folded to the right. The general description of the birth is similar to those seen at Marine Studios. Khvatov, after see- ing an apparently abnormal birth, observed an infant attached to its mother by the umbilical cord, and postulated that this was a normal con- dition. He thought that the infant suckled while still attached to the mother. In the specimens examined at Marine Studios, no umbilical cord was found to be long enough to permit such a procedure. Whether the umbilical cord of the species he studied was sufficient in length is not known, but Sleptzov, after seeing what he be- lieved to be the same species, thought Khvatov’s theory erroneous. He observed a larger number of births than did Khvatov, and in each instance the infant separated from the mother promptly with no apparent ill effects to either. All evi- dence from births in the wild indicates that they parallel closely those seen at Marine Studios. The reaction of Spray toward her first infant presents an interesting problem. She was seven years of age when she gave birth to this calf. It has been thought that dolphins mature sexually at an earlier age, and therefore it is doubtful that she was too young to care for it. She had been active for several years in escorting and protect- ing other young dolphins in the tank and it seems reasonable to assume that therefore she had sufficient general background experience. However, never before had she had the constant care of an infant, and she had not suckled an infant before. Although the immediate cause of the infant’s death was drowning, it is not certain whether the baby drowned because of a shock reaction to the fractured mandible, whether it starved until it was too weak to swim, or whether there was another reason, undiscernible to the investigators. The observers could find no ab- normalities in the digestive tract to account for its thinness or its inability to suckle. Young dolphins use their tongues, rather than their lower jaws, in suckling. The youngster was quite obviously hungry, judging from the number of times it approached its mother’s mammaries, and its persistent nuzzling there. It was im- possible to determine whether Spray produced sufficient milk as she could not be removed from the tank for examination. Her mammary glands remained distended for several days and gradu- ally receded over a period of several weeks. It is possible that more evidence may be found when Spray gives birth to a second infant and her reactions to it can be observed. Although this report contains many data that have not been previously presented, it is ex- tremely difficult to quantify, since records of the earlier births, both by McBride and his co- workers and by the present investigators, did not contain all the details which are now con- sidered essential to an analysis of the behavior of these animals. In addition it is probable that further details, up to now unrecorded, will later present themselves. Further observation by ex- perienced investigators and the recording of seemingly non-essential details will produce a more complete picture, and one that will allow a more accurate comparison with other mammals. Summary The group of captive Bottle-nosed Dolphins (Tursiops truncatus) at Marine Studios is an actively reproducing colony. Mating and par- turition take place from February to May, usually in a two-year cycle. The report com- prises the data on five pregnancies and births, one of which was observed from conception. Precopulatory behavior consisted of a pro- longed companionship between a male and fe- male. There followed such activities as postur- ing, stroking, rubbing, nuzzling, mouthing, jaw clapping and yelping on the part of the male. These occurred in any sequence and if the fe- male responded, copulatory behavior followed. This included erection, approach, intromission, ejaculation and withdrawal. Most of the copula- tory activity took place at night or in the early morning, although some of the activities were seen at all hours during the day. The gestation period was about twelve months. After about four or five months, the pregnant female tended to withdraw from the society of other dolphins, and usually estab- lished at this time the one or two associations which she maintained throughout the remainder of her pregnancy. The rest of her time was spent alone. Her activities gradually diminished in vigor, and as term approached she became slow and sometimes clumsy. Toward the end of preg- 30 I [42:2 Zoologica: New York Zoological Society nancy a labored respiratory pattern was seen, yawning and body flexion appeared, and defeca- tion increased in frequency. The infants were born tail first. The mother either whirled or swam rapidly ahead, and the umbilical cord broke as it pulled taut. The un- pigmented creases on each side of the body of the infant were not symmetrical in every case, and sometimes there were more on one side than on the other. The infants swam from birth. The dorsal fin was folded sharply either to the right or left, but stiffened in an upright position within a few hours. The placenta was expelled several hours after birth. The mother did not eat it, nor did she approach it again. Dolphin mothers kept close to their infants at all times for the first months, and removed them promptly from dangerous situations. Dur- ing the first weeks, the infants were not allowed to stray more than ten feet from the mother. Nursing was established during the first 24 hours. The infants were weaned at approxi- mately 18 months. One female dolphin, born and raised at Ma- rine Studios, did not care for her infant as care- fully as the other mothers. She left the baby alone at feeding time and for other short periods. She did not remove it from contact with other animals or from dangerous situations. The infant died at 15 days of age. Literature Cited Essapian, F. S. 1953. The birth and growth of a porpoise. Natural History, 62 (9): 392-399. Hamilton, J. E. 1945. Two short notes on Cetacea. 1— Coitus— Steno rostratus. 2— Function of the flippers and tail— Balaenoptera and Orcinus. Proc. Zool. Soc. London, 114: 549-550. James, L. H. 1914. Birth of a porpoise at the Brighton Aquarium. Proc. Zool. Soc. London, 1914: 1061-1062. Khvatov, J. P. 1938. New information on the durability of re- tention of the corpus luteum in the dol- phin. Biol. Med. Exper. U.R.S.S., 5: 27. Lillie, D. G. 1910. Observations on the anatomy and general biology of some members of the larger cetacea. Proc. Zool. Soc. London, 1910: 769-792. Lutken, C. 1888. Was die Gronlander von der Geburt der Wale wissen wollen. Zool. Jahrb., Abth. Syst., 3: 802. McBride, A. F. 1940. Meet Mr. Porpoise. Natural History, 45: 16-29. McBride, A. F. & D. O. Herb 1948. Behavior of the captive bottle-nose dol- phin, Tursiops truncatus. Journ. Comp, and Physiol. Psychology, 41(2): 111-123. McBride, A. F. & H. Kritzler 1951. Observations on pregnancy, parturition, and post-natal behavior in the bottlenose dolphin. Journ. Mammal., 32: 251-266. Morch, j. a. 1911. On the natural history of the whalebone whales. Proc. Zool. Soc. London, 1911: 661-670. Pedersen, A. 1931. Fortgesetzte Beitrage zur Kenntnis der Saugetier- und Vogelfauna der Ostkiiste Gronlands. Ergebnisse einer zweijahrigen zoologischen Untersuchungreise in Ost- gronland. Medd. Gronland, 77: 343-507. Schumann, E. A. 1914. The mechanism of labor from the stand- point of comparative anatomy. Amer. Journ. of Obstetrics, 69: 637-658. Sleptzov, M. M. 1940. On some particularities of birth and nutri- tion of the young of the Black Sea por- poise Delphinus delphis. Zoologich Zhur- nal, 19: 297-305. Slijper, E. j. 1936. Die Cetaceen — vergleichend-anatomisch und systematisch. Capita Zoologica— vol. VII. Die Lage des Embryo bei das Geburt — Kapitel 16, ss. 455-465. (Published by Martinus Nijhoff, The Hague, 1936). 1949. On some phenomena concerning preg- nancy and parturition of the Cetacea. Beijdragen tot de Dierkunde, 28: 416-446. Tomilin, a. G. 1935. Maternal instinct and sexual attachment in whales. Bull. Soc. Nat. Moscau, 44: 351-361. Townsend, C. H. 1914. The porpoise in captivity. Zoologica, 1 (16): 289-299. WiSLocKi, G. B. & R. K. Enders 1941. The placentation of the bottle-nose por- poise {Tursiops truncatus). Amer. Journ. Anat., 68: 97-125. Wood, F. G., Jr. 1953. Underwater sound production and con- current behavior of captive porpoises Tursiops truncatus and Stenella plagiodon. Bull. Marine Sci. of Gulf and Caribbean, 3: 120-133. 19571 Tavolga & Essapian: Behavior of the Bottle-nosed Dolphin 31 EXPLANATION OF THE PLATES* Plate I Fig. 1. Young male dolphin (Algie) above, and adult female (Mona) below. The male strokes the back of the female with his flippers. Fig. 2. Young male and female during rubbing activity. The two have approached each other head-on and their bodies will rub together (Algie and Spray). Fig. 3. Young male and female during rubbing activity (Algie and Spray). Fig. 4. Male and female during mouthing. Fig. 5. Adult male (Happy) and young female (Spray) in copulation. Note partial inser- tion of penis into vaginal opening. Plate II Fig. 6. Adult female (Susie) in non-pregnant state. Fig. 7. Adult female (Susie) in late pregnancy. Fig. 8. Adult female (Mrs. Jones) in late preg- nancy, in body flexion. Fig. 9. Adult female (Mona) in late pregnancy. Recurve of body flexion. Fig. 10. Female in late pregnancy. Recurve of body flexion while rubbing over rocks. Note vaginal distension and protruding nipples. Fig. 11. Birth in progress. Flukes and part of caudal peduncle are seen projecting from vagina of mother. Fig. 12. Birth in progress. Taken just before com- plete emergence. Note umbilical cord, also presence of companion female on the left. Plate III Fig. 13. Suckling. Young female (Spray) and in- fant (Peggy) a few days after birth. Fig. 14. A new mother (Pudgy) escorts her own infant and two others during feeding pe- riod while mothers of the other two feed. ♦Photographs by F. S. Essapian. 9 II liTAVOLGA a ESSAPIAN PLATE I FIG. 5 THE BEHAVIOR OF THE BOTTLE-NOSED DOLPHIN (TURSIOPS TRUNCATUS); MATING, PREGNANCY, PARTURITION AND MOTHER-INFANT BEHAVIOR FIG. 12 THE BEHAVIOR OF THE BOTTLE-NOSED DOLPHIN (TURSIOPS TRUNCATUS): MATING, PREGNANCY, PARTURITION AND MOTHER-INFANT BEHAVIOR TAVOLGA a ESSAPIAN PLATE III FIG, 13 FIG. 14 THE BEHAVIOR OF THE BOTTLE-NOSED DOLPHIN (TURSIOPS TRUNCATUS); MATING, PREGNANCY, PARTURITION AND MOTHER-INFANT BEHAVIOR 3 A Study of the Relationship Between Certain Internal and External Morphological Changes Occurring During Induced and Natural Metamorphosis in Rana pipiens and Rana catesbeiana Beulah Howatt McGovern^ & Harry A. Charipper Department of Biology, Graduate School of Arts and Science, New York University (Text-figures 1-12) The dramatic metamorphosis of the tailless Amphibia is the subject of an extensive literature resulting in large part from Gudernatsch’s (1912, 1914) accelera- tion of the process by thyroid feeding, and from the subsequent discovery of the thyroid-pitui- tary relationship by Adler (1914), Allen (1916) and Smith (1916a). A chronology of the ex- ternal metamorphic events was established by Etkin (1932), who also showed (1935) that thyroxine treatment did not change the order of events but did change their spacing. Knowledge of internal metamorphosis both normal and ac- celerated is quite complete, with many excellent and detailed studies of one, two, or more in- ternal organs. Elowever, there is no single analy- sis of internal changes comparable to Etkin’s for the external pattern. Disharmonies in de- velopment during accelerated metamorphosis have frequently been noted, and several theories of the control of the sequence and spacing of metamorphic events have been reported (see the review of Lynn & Wachowski, 1951 ) . During normal metamorphosis, the rate and extent of change in the several organs are so correlated that a given degree of change in an external organ, such as the developing hind limb, is a measure of the change to be expected in- ternally (Allen, 1929). Although the available evidence indicates that this is not true during accelerated metamorphosis, more information as to the degree of deviation from the normal corre- lation between internal and external changes in accelerated metamorphosis seemed desirable. 1 Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Graduate School of Arts and Sciences, New York University, New York, N. Y. Therefore, a comparison of the development of several internal organs in normal and thyroxine- treated animals at similar external metamorphic stages was undertaken. Rana pipiens and Rana catesbeiana were both studied in order to obtain differing lengths of larval period for the purpose of comparison. Grateful acknowledgement is made to Drs. E. G. Reinhard and W. Gardner Lynn of The Catholic University of America; to the former for interest in the problem and generosity in ex- tending facilities for research, and to the latter for valued discussion and advice. Materials and Methods 1 . Plan of the Study Rana pipiens and Rana catesbeiana were the species chosen for study because of the difference in the lengths of their larval periods. To compare the integration of internal and external changes in normal metamorphosis with that in accelerated metamorphosis, larval stages were selected at which the external morphology of the normal and treated animals was considered comparable. Normal and accelerated animals were killed at each of these stages and certain of their internal organs compared. Because of the occasional rather substantial variation in age between normal and accelerated tadpoles at ap- parently the same external metamorphic stage, untreated tadpoles of the same age as the ac- celerated were also killed and studied con- currently. To select accelerated stages most comparable morphologically to normal stages, the effects of the various thyroxine concentrations were ob- served. It was concluded that only the forelimb emergence stage was recognizably similar to the 33 34 Zoologica: New York Zoological Society [42:3 normal in all series, and therefore a more com- prehensive study was made of this stage. Earlier accelerated stages in certain respects overlapped several normal stages. Although accelerated ani- mals were studied at a stage later than forelimb emergence, this later stage has been named “terminal stage” rather than “taU stub” stage as in the normal series. The following stages were studied: (1) . Initial stage (normal animals, 12 and 21 days old) (2) . Two mm. hindlimb stage {R. pipiens only) (3) . Forelimb emergence stage (both fore- limbs emerged) (4) . Tail stub (terminal stage) The effects of rate of acceleration on meta- morphic integration were studied by the use of two aqueous thyroxine concentrations: 1 ; 10,000,000 and 1 : 100,000,000. Each has been used previously (e.g., Allen, 1932; Etkin, 1935). In addition, two age groups of tadpoles were treated with each concentration. One group was 12 days old at the beginning of treatment; the other 21 days old. Treatment by immersion was continuous. The following groups for each stage in both species were studied and compared: (1) . Animals metamorphosing normally in the laboratory; (2) . Animals accelerated by 1:10,000,000 thyroxine from 12 days of age; (3) . Untreated animals of the same age as were those in (2) when killed for ex- amination; (4) . Animals accelerated by 1:10,000,000 thyroxine from 21 days of age; (5) . Untreated animals of the same age as were those in (4) when killed for ex- amination; (6) . Animals accelerated by 1:100,000,000 thyroxine from 12 days of age; (7) . Untreated animals of the same age as were those taken in (6) when killed for examination; (8) . Animals treated with 1:100,000,000 thyroxine from 21 days of age; (9) . Untreated animals of the same age as were those in (8) when killed for exam- ination. At the selected stages, internal metamorphosis was studied morphologically following dissec- tion, and the forelimb stage of R. pipiens by microscopic examination of serial section. The internal organs studied were: (1) . Tongue: length, development, presence or absence or premetamorphic papillae; (2) . Intestine: length, histology (in R. pi- piens) at the forelimb stage; (3) . Gall bladder: color, size; (4) . Urinary bladder: presence, degree of development; (5) . Pancreas: size, histology (in R. pipiens, forelimb stage) ; (6) . Pituitary: (in /?. p/p/cni, forelimb stage) size, development; (7) . Thyroid: (in /?. pip/eni', forelimb stage) follicle number, mitotic activity, posi- tion. In addition, body length, tail length and hind- limb length were recorded throughout the course of normal and accelerated metamorphosis in both species. Because of R. catesbeiana’s long tadpole per- iod, 17 late tadpoles were collected from the Kenilworth Lily Ponds, Washington, D. C., March 24, and brought to the laboratory to serve as the normal series for the stages of fore- limb emergence and tail resorption. The rest of the R. catesbeiana tadpoles used for the study of younger normal stages and for the experi- mental series were laboratory raised from an egg film collected at the same place, July 2. 2. Method of Rearing and Feeding the Tadpoles The tadpoles were reared in flat, white enam- eled pans (10" X 15" X Wz"). For R. pipiens, 20 animals were placed in each pan with 1,000 cc. of water or thyroxine solution. For R. cates- beiana, 1 0 animals were placed in each pan with 2,000 cc. of water or thyroxine solution. The thyroxine solutions and water were changed daily. Food (chopped spinach and Pablum) was provided in excess. The experiments termi- nated after forelimb emergence when the ani- mals’ tails were resorbed to dark stubs or when deaths had reduced the number of animals in each series to 10 for R. pipiens and to 5 for R. catesbeiana. 3. The Normal and Experimental Series The experimental animals in each species were divided into four series. In the table below are listed the designation, the treatment and the number of animals in each series. Naturally metamorphosing animals consti- tuted the series designated as “SC” (Stage Con- trols). Separate series of untreated tadpoles were raised and observed, and individuals from these series were fixed concurrently with those from the experimental series. These constituted the “Age Controls.” For R. pipiens there was a total of 100 tadpoles in the Age and Stage Con- 1957] McGovern & Charipper: Metamorphosis in Rana pipiens and Rana catesbeiana 35 trol series; for R. catesbeiana, a total of 96 tad- poles. Number of ANIMALS Desig- R. pi- R. cates- nation Treatment piens beiana A-1 1:10,000,000 thyroxine from age 12 days 60 30 A-2 1:10,000,000 thyroxine from age 21 days 60 30 B-1 1:100,000,000 thyroxine from age 12 days 60 30 B-2 1 : 100,000,000 thyroxine from age 21 days 60 30 4. External Measurements Measurements were made as often as changes in rate of development seemed to require. In both species, measurements were made daily during the course of the experiments using 1:10,000,000 thyroxine. For the 1:100,000,000 thyroxine series, generally daily measurements were made for R. pipiens and less frequently for R. catesbeiana (once a week or ten days). At such times, 10 animals in each series were meas- ured. To facilitate the process of measuring, the normal and experimental animals of both species were anesthetized with MS 222. 5. Dissection Two specimens of R. pipiens and three speci- mens of R. catesbeiana for each normal stage and each experimental stage, along with an equal number of age controls, were dissected. Gall bladder color changes and presence or absence of the urinary bladder were investigated in addi- tional dissections made so that these observa- tions were based on at least five dissections for each stage and series in both species. Dissection was done using 7X and 17 X magnifications. The entire digestive tract was removed from the body cavity by freeing the cloaca from the ureters, cutting the mesentery and lifting the entire mass from the dorsal body wall anteriorly at the glottal level of the pharynx. Then the color of the gall bladder and its size relative to the liver were recorded. The shape of the ventral pancreas was sketched and its dimensions taken with dividers recorded on the sketch. The liver and pancreas were removed, the bile duct sev- ered and the intestinal mesentery cut so that the entire intestine could be uncoiled. Measurement was accomplished by cutting the intestine into convenient lengths and straightening these lengths on a millimeter rule. Finally the presence or absence of the urinary bladder and its relative degree of development were recorded. The shape of the tongue was sketched and (for R. catesbeiana) the anterior-posterior di- mension as taken with dividers was recorded. The number and disposition of the premeta- morphic papillae were also recorded. 6. Histological Study Two animals in the forelimb emergence stage in each of the R. pipiens series were serially sec- tioned and prepared for histological study. Age controls were also serially sectioned. Specimens were fixed in Bouin’s solution, sectioned in par- affin at 10 micra and stained with Masson’s stain. Observations 1. Growth Measurements It has been noted already that the normal series of R. catesbeiana represents animals ob- tained at different times; (first) mature tadpoles collected in the field and raised in the laboratory through the rest of the tadpole period; and (sec- ond) tadpoles hatched in the laboratory from an egg film collected later in the season. The tad- poles from the egg film provided both the experi- mental series and their normal Age Controls. The Age Controls were observed during a period of 173 days and were fixed at the end of that time with the remaining experimental. By this time, a number of the Age Controls were as large and as well developed as the late tadpoles (used as the Stage Controls) had been when collected. Therefore the growth data for the two normal groups were plotted as a single curve (Text-fig. 1) broken by a time lapse equal to that between December (when the Age Controls were fixed) and March (when the Stage Con- trols were collected). The growth curves and the time required for response (Text-fig. 2) were approximately the same in both species of organisms treated with 1 : 10,000,000 t^roxine. The experiments termi- nated in death after 16 days for R. pipiens and 10 or 11 days for R. catesbeiana. With 1:100,- 000,000 thyroxine, the growth curves for the two species are very similar (Text-figs. 3 & 4) . The number of days required for response in R. catesbeiana was greater than for R. pipiens, but, considering the normal length of the re- spective tadpole periods, the effect of treatment appeared earlier in R. catesbeiana (see Text- fig. 5). It should be pointed out, however, that tail resorption was carried further in R. pipiens (Text-fig. 6) during this time. The age difference in the groups of tadpoles treated with the same thyroxine concentrations caused no appreciable difference in the growth curves in either species. The curves obtained for hindlimb growth of tadpoles treated with 1:100,000,000 thyroxine (Text-figs. 3 & 4) are almost identical with that published by Etkin (1935) for thyroidectomized R. cantabrigensis tadpoles. Relative to body length, the hindlimbs of the 1 : 1 00,000,000 thy- roxine-treated animals were longer than normal in both species (Text-fig. 5). I 36 Zoological New York Zoological Society [42:3 Text-fig. 1. Normal growth curves for (a) body length, (b) tail length, and (c) hindlimb length in R. pipiens and R. catesbeiana. One unit of age for R. pipiens has been made equal to five units for R. catesbeiana for easier comparison of the curves. Continuous tail growth of both species of ani- mals treated with 1:100,000,000 thyroxine (Text-fig. 6) was unexpected in view of Blacher’s ( 1928) observation that short exposure to dilute concentrations of thyroidin produced tail short- ening. 2. Dissection Results a. Intestine Kuntz (1922) found that the average reduc- tion in intestine length during the normal meta- morphosis of R. pipiens was 82.2%. In the pres- ent investigation the average reduction in the intestine length of control animals was found to be 86.2% for R. pipiens and 83.7% for R. catesbeiana. These averages are in very good agreement if it is remembered that dietary dif- ferences may affect intestine length by as much as 58% (Babak, 1905). Text-fig. 7a shows the changes in the ratio of intestine length to body length for the normal MM days MM 40 30 (b) / Body a L. Tail Body A- 1 R pipiens 10 20 30 40 A-2 R catesbeiana Legend control treated ACCELERATED GROWTH Text-fig. 2. Growth under treatment with 1:10,000,000 thyroxine. A-1 curves are for tadpoles treated from 12 days of age; A-2 are for tadpoles treated from 21 days of age; (a) and (b) show data from treated R. pipiens tadpoles; (c) and (d) from R. catesbeiana. and for the accelerated R. pipiens series; in Text- fig. 7b, comparable data are presented for R. catesbeiana. It can be seen that, during normal metamorphosis, the ratio for the tadpole stage is 13 to 14 and it decreases to a value of approxi- mately 2 at the tail resorption stage. By the time of forelimb emergence, one-half of the total shortening had occurred in the intestine of R. catesbeiana and three-quarters of the total short- ening of the intestine of R. pipiens had been effected. Treatment with 1 : 10,000,000 thyroxine produced shortening of the intestine, giving a ratio (intestine length to body length) of 2; however, shortening had been completed by the forelimb emergence stage. Under the influence of 1:100,000,000 thyroxine, little shortening had taken place by forelimb emergence. By the terminal stage, the intestine length-body length ratio in R. pipiens is almost normal as compared with the ratio for R. catebeiana in which the intestine remains considerably longer than twice the body length. 1957] McGovern & Charipper: Metamorphosis in Rana pipiens and Rana catesbeiana 37 ACCELERATED GROWTH R. PIPIENS Text-fig. 3. R. pipiens, growth of tadpoles under 1:100,000,000 thyroxine treatment. B-1 curves are for tadpoles treated from 12 days of age; B-2 for tadpoles treated from 21 days of age. Blacher (1928) and Allen (1932) established the order of sensitivity to thyroid treatment of the several tadpole organs. Both Blacher and Allen reported the intestine to be one of the more sensitive organs by virtue of its shortening after even brief periods of exposure to dilute concentrations. In Text-fig. 7, the curves show that the intestine of tadpoles treated with 1:100,000,000 thyroxine continues to grow for some time prior too the onset of shortening. The data can be reconciled if, instead of shortening, one postulates a period of slower intestinal growth in the treated tadpoles than in controls of the same age. b. Pancreas In the normal animal the pancreas progres- sively increases in size up to the time of meta- morphosis when regression begins. The shape and the size of the pancreas were the same in treated and in untreated individuals of both species of the same age. Because regression cut short the normal growth period, the experi- mental animals’ pancreas began regression at a MM B-l ACCELERATED GROWTH R. CATESBEIANA Text-fig. 4. Growth of R. catesbeiana tadpoles un- der 1:100,000,000 thyroxine treatment, (a) B-l curves are for tadpoles treated from 12 days of age; (b) B-2 curves for tadpoles treated from 21 days of age. smaller size than normal. Treatment with 1 : 10,000,000 thyroxine elicited a marked regres- sion at the forelimb emergence stage and further shrinkage was observed in the terminal stage. There was no apparent regression in the pan- creas of animals in the forelimb emergence stage which were treated with 1:100,000,000 thyroxine. At the onset of the terminal stage, regression was evident in both species. c. Gall Bladder Early in normal development the gall bladder was large as compared with the liver (about 1 /5 its size). Later, it was relatively smaller (1/8 to 1 /lO of the size of the liver). By the time of forelimb emergence, the gall bladder again ap- peared larger ( 1/5 to 1/4 the size of the liver) . No measurements were made to show whether these size changes were changes in the gall bladder or in the liver, though it was apparent that the liver first increased and then decreased in size. Hoskins & Hoskins (1919a&b) reported I 38 Zoological New York Zoological Society [42:3 /^ge ,n 50 100 150 200 250 300 350 400 Doys (b) R.catesbeiano RELATIVE HINDLIMB LENGTHS Text-fig. 5. Relative hindlimb lengths in normal tadpoles and tadpoles treated with 1:100,000,000 thyroxine from 21 days of age. (a) R. pipiens; (b) R. catesbeiana. Age in days is plotted against the ratio of body length to hindlimb up to the tail stub stage in normal animals and to the “terminal” stage in the treated. that the liver decreased in size during meta- morphosis, and Kuntz (1922) reported an 80% reduction in liver weight which took place when tail resorption and intestine reduction were well advanced. The gall bladders of animals treated with 1:10,000,000 thyroxine remained relatively large. At the terminal stage, they appeared rela- tively larger than in normal animals at the tail stub stage and the livers appeared to be more than normally reduced in size. The gall bladders of tadpoles treated with 1:100,000,000 thy- roxine were still 1/8-1/10 of the size of the liver at the forelimb emergence stage. Color change in the gall bladder during accel- erated metamorphosis was studied in Hyla versi- color, H. crucifer, R. clamitans, Bufo ameri- canus and Ambystoma maculatum (A. puncta- tum fide Speidel) by Speidel (1926). He noted MM 40 30 (o) 20 / '' 10 - 0 10 20 30 40 50 60 70 R. pipiens MM - 50 40 Legend (b) 30 / / Control 20 ]/ Treated (8-2) 10 L 1 1 1 1 1 1 1 1 Age in 50 100 150 200 250 300 350 R.catesbeiano TAIL GROWTH Normal and Accelerated Text-fig. 6. Tail growth in normal tadpoles and tadpoles treated with 1:100,000,000 thyroxine from 21 days of age. (a) R. pipiens; (b) R. catesbeiana. no exceptions in several hundred observations from the series: “yellow-green,” “brighter- green,” “emerald green” and a “very dark green.” In normal R. pipiens, the gall bladder was dark green at the time of forelimb emer- gence. In R. catesbeiana the gall bladder was emerald green at this time, and it was dark green at the tail stub stage. In most of the 1:10,000,- 000 thyroxine-treated R. pipiens tadpoles, the gall bladder was dark green at the forelimb emergence stage; in R. catesbeiana, at the ter- minal stage, the gall bladder was still emerald green as noted above. In the 1 : 100,000,000 thy- roxine-treated tadpoles of both species, the gall bladder was emerald green at the forelimb emer- gence stage and approximately half of these in- dividuals had dark green gall bladder at the terminal stage. d. Urinary Bladder The urinary bladder developed late in the normal premetamorphic climax period and was small but well developed in the tadpole with fully developed hindlimbs. At the forelimb emergence stage of both spe- cies which had been treated with 1:10,000,000 1957] McGovern & Charipper: Metamorphosis in Rana pipiens and Rana catesbeiana 39 Text-fig. 7. Relative intestine lengths. Age in days is plotted against the ratio of intestine length to body length for R. pipiens in (a) and for R. catesbeiana in (b). A-1 curves are for tadpoles treated with 1:10,000,000 thyroxine from 12 days of age; A-2 for those treated with 1:10,000,000 thyroxine from 21 days of age; B-1 for those treated with 1:100,000,000 thyroxine from 12 days of age; B-2 for those treated with 1:100,000,000 thyroxine from 21 days of age. thyroxine, approximately half had urinary blad- ders. At the terminal stage, all but one or two animals had small, poorly developed urinary bladders. In both species, the animals treated with 1:100,000,000 thyroxine had well-devel- oped urinary bladders by the time of the fore- limb emergence period. e. Tongue No measurements of the tongue were made for R. pipiens. In R. catesbeiana, normal develop- ment of the tongue began (when the tadpoles reached 21-27 mm. body length) as a white con- densation of tissue in the anterior part of the floor of the mouth. At a body length of 35-38 mm., the tongue was clearly outlined and the tissue mass was about 1 mm. long. At the time of forelimb emergence, the tongue was a well- formed organ, 2.5-3 mm. long, with the poste- rior edge free and bifurcated. Between the end of forelimb emergence and the onset of the tail stub stage, the tongue grew rapidly and almost doubled its length to 5 mm. All the experimental animals of both species showed more advanced tongue development than normal animals of the same age. In the groups treated with 1 : 10,000,000 which were examined at the terminal stage, the tongue was poorly differentiated, i.e., a free but not bifurcate posterior edge. In the groups treated with 1:100,000,000 thyroxine, the tongue appeared to be the same as in corresponding stages of normal animals. The amount of tongue growth (in length) between the forelimb emergence and the terminal stages for thyroxine-treated R. catesbeiana is compared with that in the normal in the table below. Also included is the time in days between the two stages. (The A-1 series, treated with 1:10,000,000 thyroxine from 12 days of age, is omitted because the tongue meas- ured less than 0.5 mm in length at the forelimb emergence stage). The slower growth rate of the tongue in or- ganisms treated with 1:100,000,000 thyroxine is marked (See B-1 and B-2). In the normal R. pipiens tadpole there are two premetamorphic tongue papillae; in R. cates- beiana there are four. These begin to be resorbed in the late tadpole stages and disappear by the onset of tail stub stage. In all the R. pipiens series, the papillae were gone by the beginning of the forelimb emergence, but in R. catesbeiana. 40 Zoological New York Zoological Society [42:3 Tongue Growth (in mm.) Stage Control .4-2* B-1** B-2** Forelimb Emergence 2.7 0.5 1.3 1.3 Tail Stub (Terminal) 5.0 0.7 2.2 1.8 % Increase 85% 40% 69% 38% Days between Stages 7 2 82 90 * Treated from 21 days of age with 1:10,000,000 thyroxine. ** Treated from 12 days of age with 1:100,000,000 thyroxine. ***Treated from 21 days of age with 1:100,000,000 thyroxine. papillae were still present at the terminal stage of animals treated with 1:100,000,000 thy- roxine. The histology of the premetamorphic tongue papillae of R. sylvatica was studied by Helff & Mellicker (1941) who found four papillae, the characteristic number for that species. In their search of the literature, they found that one previous investigator had described these struc- tures. Schulze (1870, cited in Helff & Mellicker, 1941) found two papillae in Pleobates fusca, R. esculenta, R. temporaria and Bufo cinereus. R. catesbeiana thus resembles R. sylvatica in pos- sessing four papillae, whereas R. pipiens re- sembles the species studied by Schulze. 3. Histological Studies: Forelimb Stages, R. pipiens a. Pituitary Gland i. Anatomy In Text-fig. 8 are reproduced Atwell’s (1918) reconstructions of the development of the anuran pituitary gland. It can be seen that two changes take place: first, the two lobes of the pars lateralis come to lie anterior and lateral to and (in the adult) separate from the pars an- terior and pars intermedia; second, the mass of the gland grows posteriorly, the Anterior at a more rapid rate than the Intermedia (so that in the adult the former is the most posterior part of the gland. Also in Text-fig. 8, the length of each of the three components of the pituitaries of single normal and experimental animals is plotted in relationship to other parts. These dia- Text-fig. 8. The pituitary in normal and treated R. pipiens. The drawings numbered 9 through 14 above the block graphs are redrawn from Atwell’s (1918) reconstructions of pituitary development in Anura. The block graphs show the length and spatial relationships of the pituitary parts of normal and treated R. pipiens tadpoles obtained by counting the 10 micra sections each part occupied in each animal. A-1 is data from a forelimb stage animal 24 days old treated with 1:10,000,000 thyroxine from 12 days of age; A-l-C is from an untreated tadpole 24 days old; A-2 is from a forelimb stage animal 34 days old treated with 1:10,000,000 thyroxine from 21 days of age; A-2-C is from an untreated animal 34 days old; B-1 is from a forelimb stage animal 50 days old treated with 1:100,000,000 thyroxine from 12 days of age; B-l-C is from an untreated tadpole 50 days old; B-2 is from a forelimb stage animal 52 days old treated with 1:100,000,000 thyroxine from 21 days of age; SC is from an untreated fore- limb stage animal 67 days old. 1957] McGovern & Charipper: Metamorphosis in Rana pipiens and Rana catesbeiana 41 grams result from plotting the number of 10 micra sections constituting each part of the pituitary. In the 12-day-old, 1:10,000,000 thyroxine- treated forelimb stage (A-1), pituitary relation- ships correspond to those in Atwell’s youngest animal (stage 9), while the untreated animal of the same age (A-l-C) had progressed to Atwell’s stage 10. In the 21 -day-old, 1:10,000,000 thy- roxine-treated forelimb stage animal (A-2), these appear to be the same as in the untreated animal of the same age. Both correspond to Atwell’s stage 10 or 11. The higher concentra- tion of thyroxine did not cause precocious ana- tomical development of the pituitary. The pituitaries of the animals treated with 1:100,000,000 thyroxine from 12 days of age (B-1) and those treated from 21 days of age (B-2) were smaller than those of either the normal tadpole of the same age (B-l-C) or of the normal forelimb stage animal (SC). The block graphs do not show the progressive change in the connection between the two lobes of the pars lateralis with the pars anterior illus- trated by Atwell’s figures 12, 13 and 14. Exami- nation of the slides showed that the normal tad- pole (B-l-C), on the basis of lateralis position, had reached a stage between Atwell’s 12 and 13 while the normal animal (SC) was between stages 13 and 14. The forelimb animals treated with 1:100,000,000 thyroxine (B-2 and B-4) appear to resemble the normal tadpole more than the normal forelimb emergence stage. The lower concentration of thyroxine, moreover, showed no accelerating effect on the anatomical development of the pituitary. ii. Relative size Text-fig. 8 also shows that the pituitaries of PITUITARY PART LENGTHS AND BODY LENGTHS AS 0/0's OF NORMAL FORELIMB STAGE VALUES B Experimental (H Control Text-fig. 9. Pituitary part lengths and body lengths of thyroxine-treated forelimb stages compared with those of the normal forelimb stages in R. pipiens. The length of each pituitary part and the body length of the experimental forelimb stages and of normal tadpoles of the same age as the treated are shown as percents of the normal forelimb stage values (SC, last set of four 100% blocks). In A-1, data from a forelimb stage animal treated with 1:10,000,000 thyroxine from 12 days of age are shown in black blocks superimposed upon that of an untreated animal of the same age in white. In A-2 the data from a forelimb stage animal treated with 1:10,000,000 thyroxine from 21 days of age are shown in the same way against the data from an untreated tadpole of the same age; B-1 are data from a forelimb stage animal treated from 12 days of age with 1:100,000,000 thyroxine shown against that for an un- treated tadpole of the same age; B-2 data from a forelimb stage animal treated from 21 days of age with 1:100,000,000 thyroxine are shown against the data for an untreated tadpole of the same age. 42 Zoologica: New York Zoological Society [42:3 all of the experimental forelimb-emergence ani- mals were shorter than those of untreated ani- mals of the same age. To determine whether the reduced size of the pituitary was proportional to the reduced body size of the thyroxine-treated animals and to determine whether or not all pituitary parts were equally affected, the length of each pituitary part and the body length of the normal forelimb stage animal were taken as 100%. The same measurements for each ex- perimental forelimb animal and for its age con- trol were recalculated as percents of the lengths of the corresponding parts in the normal fore- limb animal. In Text-fig. 9, the profiles produced by graphing these values for the experimental animals were superimposed in black upon the profiles of their respective age controls. The last profile to the right (a block of four 100% col- umns) represents the values for the normal fore- limb stage animal. It can be seen that the bodies of those treated with 1:10,000,000 thyroxine and those treated with 1:100,000,000 thyroxine appeared to have bodies and pituitaries about equally inhibited (with regard to length) at the forelimb stage, and in these animals the pars anterior appeared to be the pituitary component most inhibited, iii. Cellular Differentiation Distribution and differential counts of the pituitary cell types were not considered because of the thickness of the sections. The cytoplasm in the cells of all the lobes was less abundant in the experimental animals than in either the age controls or the normal forelimb stage. Pigment granules which were large and coarse in young normal stages were more evident in all the ex- perimentals than in the stage controls. b. Thyroid Gland i. Size of the Gland Conclusions as to thyroid size were based on the length of the glands determined by counting the number of 1 micra sections which the two halves of the gland occupied in each animal, and on a relative area per section derived from count- ing the number of follicles containing visible colloid. To show how the thyroids of the thy- roxine-treated forelimb stage animals differed from those of the normal forelimb stage animal and from the untreated tadpoles of the same age, body length, length of thyroid and average num- ber of colloid-filled follicles per section in the normal forelimb stage animal were arbitrarily rated as 100%. The same data for the treated animals (forelimb stage) and for their respective age controls were recalculated as percents of these norms. In Text-fig. 10 are five profiles re- sulting from this treatment of the data. At the extreme right is the normal forelimb stage pro- ®/ 4> H Experimental □ Control THYROID LENGTHS Text-fig. 10. Thyroid size and body size in normal and thyroxine-treated R. pipiens. Body length, length of thyroid determined by counting the number of 10 micra sections the gland occupied and the aver- age number of follicles per section for the accele- rated forelimb stages and for untreated animals of the same ages as the treated are shown as per- cents of the normal forelimb stage values (SC, the last set of three 100% blocks). The values for each treated forelimb stage animal are shown in black blocks superimposed on the same values for a normal tadpole of the same age shown by white blocks. A-1 shows the values for a forelimb stage animal treated from 12 days of age with 1:10,000,000 thyroxine and for a normal animal of the same age; A-2, for a forelimb stage animal treated from 21 days of age with 1:10,000,000 thyroxine and for a normal tadpole of the same age; B-1 for a forelimb stage animal treated from 12 days of age with 1: 100,000,000 thyroxine and for a normal tad- pole of the same age; B-2 for a treated forelimb stage animal treated from 21 days of age with 1:100,000,000 thyroxine and for a normal tadpole of the same age. file. To the left are the profiles for the untreated tadpoles. Superimposed upon these are the pro- files of the treated forelimb stage animals of the same age. Text-fig. 10 shows that the thyroids of the treated animals at the forelimb emergence stage were inhibited as reflected by their size when compared with the normal forelimb stage and 1957] McGovern & Charipper: Metamorphosis in Rana pipiens and Rana catesbeiana 43 ... untreated animals of the same age. Inhibition as indicated by area (determined by the number of follicles) was greater than that indicated by length. ii. Mitotic Activity In the following table are given the number of 10 micra sections occupied by both the right and left halves of the thyroid glands, and the number of mitoses in the entire glands. Sections Mitoses A-1 30 3 A-l-C 39 47 A-2 49 12 A-2-C 62 120 B-1 95 14 B-l-C 130 406 B-2 131 24 SC 211 861 The normal series- -A-l-C, A-2-C, B-l-C and SC— gave evidence of higher mitotic activity than did the treated animals. iii. Glandular Activity A thyroid gland made up of vacuolate col- umnar cells, with vacuolate colloid, and of many collapsed follicles, possesses the generally ac- cepted histological criteria of a physiologically active gland. The forelimb stage animals treated from 12 days of age with 1:10,000,000 thy- roxine (A-1) had thyroid glands made up of cuboidal cells with scant cytoplasm, with few follicles, and with no vacuolation. In those treated with 1:100,000,000 thyroxine from 21 days of age (A-2), the thyroid cells were cuboidal. The cells had more cytoplasm than those of the A-1 animals but less than those in animals of the same age. As in the age controls, a few follicles showed vacuolation of the colloid. The animals from both age groups treated with 1 : 1 00,000,000 thyroxine had thyroids in which the cuboidal cells were approaching a columnar architecture. The thyroids of the normal fore- limb stage animals bore all of the criteria of high activity. iv. Pigment In the younger normal animals (as was noted by D’Angelo & Charipper, 1939) large aggre- gates of pigment are conspicuous features of the thyroid gland. By the time of forelimb emer- gence, however, such aggregates were rare. Moreover, the granules are distributed and the pigmentation is not conspicuous. All of the ex- perimental animals showed pigmentation either to the same degree or to a greater degree than their respective age controls. V. Thyroid Position Text-fig. 11 shows outline drawings from Text-fig. 11. Thyroid position relative to the gill skeleton in the normal forelimb stage and in the forelimb stage produced by treatment with 1:100,000,000 thyroxine from 12 days of age. Graphic reconstruction from serial sections. Normal stage magnified 5X; treated 11.25X. The thyroid is shaded. graphic reconstruction of the thyroid gland and its relationship to the ventral aspect of the hyobranchial cartilages. Text-fig. 11a shows the normal forelimb emergence stage (5X). Text- fig. 11b is the forelimb emergence stage after 1:100,000,000 thyroxine treatment from 12 days of age (enlarged 11.25X). The thyroid gland assumed a more anterior position under the gill-bearing skeleton in the treated animals than in the normal. D’Angelo & Charipper’s (1939) Figure 1 , illustrating the normal changes in thyroid position up to and including meta- morphosis, indicates that the more anterior posi- tion is characteristic of normal tadpoles of the same age as the treated. The position of the thyroid in the normal forelimb stage agrees with their observations from the same stage. Thus 1:10,000,000 thyroxine treatment did not effect a precocious migration of the thyroid. c. Small Intestine In the normal forelimb stage animal, the histology of the small intestine is well-repre- sented by Kuntz’s (1922) figures for the same stage of R. pipiens and by lane’s (1934) illus- tration for this stage in thyroid-fed tadpoles of R. clamitans. There were many mitoses in the basal cells of the new mucosal layer. The old mucosa was evidenced by conspicuous necrotic masses. The submucosa and muscularis had 44 Zoological New York Zoological Society [42:3 thickened considerably. The forelimb stage ani- mals treated with 1:10,000,000 thyroxine had the intestinal histology of the normal forelimb stage. However, the cytoplasm of the cells in all layers was very scant and the resulting histology was superficially quite different from the normal. Forelimb stage animals treated with 1:100,- 000,000 thyroxine had the same intestinal his- tology as that of the untreated tadpoles of the same age. d. Pancreas In the normal forelimb stage animal regres- sion of the pancreas was far advanced. The his- tological picture represented a condition inter- mediate to Kaywin’s (1936) stages 3 and 4 of thyroxine-treated R. catesbeiana. The acini showed three stages of metamorphic changes: ( 1 ) some were necrotic, (2) some were made up of syncytial cells, and (3) some were made up of cells with large nuclei and sparse, deeply- staining cytoplasm. There were pronounced con- nective tissue spaces. In the forelimb stage of animals treated with 1 : 10,000,000 thyroxine, the pancreas was com- parable to the normal forelimb stage pancreas as regards the degree of metamorphic change. However, pigment was very conspicuous. Fore- limb stage animals treated with 1:100,000,000 thyroxine failed to show metamorphic change in the pancreas. The glands were smaller than those of the untreated tadpoles of the same age, but there were no signs of loss of definition of cell boundaries or necrosis in the acinar cells. e. Summary of Observations on the Histology of the Forelimb Stages of R. pipiens Kahn (1916) found that the pituitaries of tadpoles treated with horse thyroid were much larger than normal. Schliefer (1935) found that thyroid extract had no effect on the pituitary although the development of the gland was ac- celerated along with the rest of the body. Under the conditions of the present study the anatom- ical development, at the forelimb emergence stage, of the pituitary of thyroxine-treated ani- mals was not found to be more advanced than that of untreated animals of the same age. In fact, the glands were smaller. Treatment with 1:10,000,000 thyroxine was less inhibitory to pituitary growth than it was to body growth. Thus, in agreement with Kahn, the pituitaries were larger than those of untreated tadpoles of the same age. After treatment with 1:100,000,- 000, body size and pituitary size were in a more normal proportion; with this concentration, the pars anterior appeared to be more inhibited than the other pituitary components. With both con- centrations, the cytoplasm of the cells of all parts of the pituitary gland was sparse and pig- ment remained conspicuous in the pars anterior and pars lateralis. The thyroids of treated animals showed evi- dence of inhibition in every respect considered at forelimb emergence. They were much smaller, had a lower mitotic rate and give histological evidence of a lower degree of differentiation and physiological activity than the glands of either the normal forelimb stage or of the untreated tadpoles of the same age. These findings agree with those of Clements (1932), Etkin (1935) and Brinks (1936). The small intestine and pancreas in animals treated with 1 : 10,000,000 thyroxine showed the same type of histological change at the time of forelimb emergence as that of normal animals at the same stage. On the other hand, the intestines and pancreases of those treated with 1 : 100,000,- 000 thyroxine did not show metamorphic change. 4. General Summary of Observations A comparison of the condition of seven in- ternal organs, in normal metamorphic stages and in comparable stages produced by treat- ment with two concentrations of thyroxine, has shown that at each stage, the accelerated ani- mals are different not only from the normal but also from each other. This is summarized, in the table below, for the forelimb stage. Difference in the length of the tadpole period in R. pipiens and in R. catesbeiana did not greatly affect the results of thyroxine treatment as studied by growth measurements and dissec- tion. R. catesbeiana reacted earlier, considering its normal tadpole period, to both concentra- tions of thyroxine than did R. pipiens. In days, R. catesbeiana reacted more quickly to 1:10,- 000,000 thyroxine and less quickly to 1:100,- 000,000 thyroxine than R. pipiens. As judged from the degree of response, R. catesbeiana was less sensitive than was R. pipiens, as shown by the fact that resorption of the tail and shorten- ing of the intestine after 1:100,000,000 thyrox- ine treatment was not as advanced. In compari- son with their respective normal stages, never- theless, the animals showing accelerated devel- opment behaved similarly in both species. Discussion The thyroid glands themselves of animals un- dergoing thyroxine-accelerated metamorphosis apparently do not modify the effects of thyroxine treatment. Etkin (1935) reached this conclu- sion by comparing the effects of thyroxine on thyroidectomized and partially thyroidectomized R. cantabrigensis tadpoles. The present study supports that conclusion by the detailed simi- larity of the external metamorphosis of acceler- Normal Development in Normal and Treated Forelimb Stage Animals 1957] McGovern & Charipper: Metamorphosis in Rana pipiens and Rana catesbeiana 45 o H Z O U w o < o o o o" o o o o o o o o o' .4 < o a o z o o 'S. 6 o U bO .a a 1 t: : o : ^ : w 4 ? C3 o Cu *0 C3 o 73 D 73 cd o Cl, 73 cd H 3 73 cd *o cu 73 cd H o -S- "o Is a o Z *> (U o Z X W o Oh 73 O CU 73 3 73 cd cy 'o Oh > w o Oh 73 Cd H 0 fl cd Pu 01 o S .2 gsl S o .C, 2i o Cub c« si I I I C 0 o « S .a C3 o K o wO •o J-l >«TN o ^ 13 Oh.S d T3 TJ cd s u, cd .a CJ o cd c3 .a 73 73 fl 2 O g S O 3 s, ^ fti pi d .g a d 5 (D 2 ^ M Jg ^ g C/5 pH CLi H Cd o u 73 a }H Cd O O 0) CUO CiO a cd a ^ 3 <3 Os £. b 2 3 3 .tS 3 2 < U 03 2 ^ CL Li W o «'» 'U 3 3 “ 2 Li a CO S C/D 73 3 i2 > CO PC (i. 6-W C/5 73 ^ IH .2 2 S > P2 [L, 00 .a 3 Lh 00 a o Z 00 X d 3 3 £?.§ S “ C3 C3 =2 S tB o ' •S- S- s S 2 0*^ O 5^ CO oo VO CO fs CO r4 O cc ps: o r- o 00 Os On VO 00 CO CO CO O' u o ' j q d CO q •o o d a +1 rt o\ o\ O +1 0\ N C/3 g C3 o s o\ *o d Tf C4 cn C<3 00 d 00 CT3 •o q ■rf Ov VO 00 m o\ '=d- d C<3 cn vd 00 d d d CN a 00 00 »o cs J0 o q q o q q CN O ri 00 cf\ '4" o ps! u q ra q q c^ q d (N q »o q 0\ q 00 q 3 L, O d c +1 o r-. +1 rJ 00 »o O ri d •o CN 00 00 vd ON c/3 s fcJO s CO d .a 00 o 'd* o\ *o VO VO VO Ov a cd CO c/3 CO (L) ’5 Oh CO O *Oh o VWIXVUI VJtC y/fl crepitans eptodactylus typhonius iipemphix pustiilosus rostherapis trinitatis emperate and boreal specie ana clamitans ana sylvatica -S ufo boreas halophilus ufo boreas boreas H ai a i-J a. H ac ac a: oa CQ 1957] Tashian & Ray: Oxygen Consumption and Temperature in Amirans 67 significantly reflected by over-all oxygen con- sumption in the tropical and arctic insects and spiders that they tested at 0° and 20° C. Prosser (1955) also concludes that metabolic differences in populations are to be found in aquatic but not in terrestrial poikilotherms. If we consider the amphibians studied here as semi-terrestrial, we should then expect to find a significant, but not pronounced, difference in climatic adapta- tion between the tropical and northern forms. This appears to be the case. It would seem that a similar study of tropical and temperate rep- tiles would be of particular interest in this respect. It is also generally accepted (Bullock, 1955) that cold-adapted poikilotherms tend to show a lower Qio when compared with the same or closely related warm-adapted species. This con- cept becomes apparent among the anurans of this study only in the toad Bufo boreas. The other frogs show no significant variation in Qm between tropical and temperate forms (Table 1). This seemingly inherent low Qio for Bufo boreas could have been a contributing factor in its successful northern extension into south- ern Alaska. It is interesting to note, however, that in Rana sylvatica, which ranges even fur- ther north than Bufo boreas, we find no in- herently lower temperature coefficient. These results can probably be partially explained by the limited number of determinations made as well as by the fact that we are here grossly com- paring different species. Moreover, familial dif- ferences such as the more terrestrial habits of Bufo boreas could conceivably explain the dis- similarities. Breeding condition, size and sea- sonal fluctuations may also have had an influence. However, with the exception of Hyla crucifer, all northern animals were collected between June 1 and July 4, and only adults of similar weight within a species were used in the determination of the temperature coefficients. Bufo boreas boreas was the only form in breed- ing condition at the time of collecting. It is sur- prising that individuals of Bufo boreas exhibited such a persistently similar metabolic picture, coming as they did from two places so widely separated in latitude and altitude (sea level to 7,500 ft.). There is good evidence in support of the con- clusion (for references see Dehnel, 1955) that at their normal environmental temperatures, ac- tivity rates of closely related northern and southern animals are usually similar. The annual range of mean monthly temperatures for Juneau vary from -1.6° C to 13.0° C (1921-1950) with an annual mean of about 5.8° C. compared with a range of 25.6° to 27.3° C. and an annual mean of 26.5° C. for Trinidad (Port-of -Spain) . If we take the mean monthly temperature for June at Juneau (12.0° C.) and the average January to March monthly temperature for Simla (ca. 24° C.), we can then use the lower testing tem- perature (14° C.) for Bufo boreas boreas and the upper testing temperature (25° C.) for the tropical species as being reasonably close to their normal mean habitat temperature at the time of collecting. In examining the two extreme lati- tudinal forms of similar size (Table 1), we find Bufo boreas boreas with a mean oxygen coeffi- cient (QO2) of .152 at 14° C., which is higher than that of all but one of the tropical frogs tested at 25° C. At their normal habitat tempera- tures, then, the northern species showed the more active metabolism. Other studies (Scho- lander et ah, 1953, and Tashian, 1956) showed a higher QO2 for the tropical forms at their normal environmental temperatures. Summary 1 . The oxygen consumption of five species of tropical frogs from Trinidad, B.W.I., was de- termined and compared with that of four species of temperate and boreal anurans from Vermont, New York, California and Alaska at two tem- perature levels. 2. The respiratory metabolism for the north- ern forms averaged higher than that of the tropical animals at both temperatures. This higher metabolism for the cold-adapted animals is more evident in the smaller species, and is reflected in a steepening of the weight regression slope for the tropical species at both tempera- ture levels. 3. With the exception of a lower Qio for both races of Bufo boreas when compared with any of the other forms tested, no significant variation could be detected in the Qio of northern and southern forms. 4. At their normal habitat temperatures, the northern species had a higher rate of respiration than the tropical species. Literature Cited Bullock, T. H. 1955. Compensation for temperature in the metabolism and activity of poikilotherms. Biol. Revs., 30: 311-342. Davison, J. 1955. Body weight, cell surface, and metabolic rate in anuran amphibia. Biol. Bull., 109: 407-419. Dehnel, P. A. 1955. Rates of growth of gastropods as a func- tion of latitude. Physiol. Zool., 28: 115- 144. 68 Zoologica: New York Zoological Society [42: 5: 1957] Flemister, L. J. & S. C. Flemister 1951. Chloride ion regulation and oxygen con- sumption in the crab Ocypode albicans (Bose). Biol. Bull., 101: 259-273. Moore, J. A. 1949. Geographic variation of adaptive char- acters in Rana pipiens Schreber. Evolu- tion, 3 : 1-24. Prosser, C. L. 1955. Physiological variation in animals. Biol. Revs., 30: 229-262. Rubner, M. 1924. Aus den Leben des Kaltbluters, 11 Teil, Amphlbien und Reptilien. Bioch. Zeitschr., 148: 268-307. SCHOLANDER, P. F., W. FLAGG, V. WALTERS & L. Irving 1953. Climatic adaptation in arctic and tropical poikllotherms. Physiol. Zool., 26: 67-92. Tashian, R. E. 1956. Geographic variation in the respiratory metabolism and temperature coefficient in tropical and temperate forms of the fiddler crab, Uca pugnax. Zoologica, 41: 39-47. VOLPE, E. P. 1954. Hybrid inviability between Rana pipiens from Wisconsin and Mexico. Tulane stud. Zool., 1: 111. 6 Basic Patterns of Display in Fiddler Crabs (Ocypodidae, Genus Uca) Jocelyn Crane Department of Tropical Research, New York Zoological Society, New York 60, N. Y. (Plate I; Text-figures 1-4) Contents I. Introduction 69 II. General Form of Display in Fiddler Crabs 70 III. Basic Wave Patterns 70 IV. Advanced Stages of Display 73 V. Ecological Differences in Distribution of Display Types 74 VI. Geographical Distribution in Relation to Display 76 VII. Evolutionary Trends 76 VIII. Summary 81 IX. Referenees 81 I. Introduction This contribution is a preliminary result of a general study of ocypodid crabs. The elaborate and varied displays characteris- tic of the fiddler crabs (genus Uca) appear to contribute substantially to an understanding of the phytogeny of the group. An interim report is presented because the increasing use of fiddler crabs in biological research makes their relation- ships of timely interest. Specific differences in display behavior were described in an earlier study (Crane, 1941) where it was found that in a group of sympatric species of Uca on the west coast of Central America every species could be distinguished from every other species by the form of the dis- play. Even species which could not be otherwise distinguished without a lens were easily identi- fied at a distance by the characteristic form and tempo of the “waving” of the large cheliped. It 1 This study has been supported by a grant from the National Science Foundation. 2 Contribution No. 975, Department of Tropical Re- search, New York Zoological Society. was found further that closely related species had similar types of display which were divis- ible into a number of groups. During subsequent years other American spe- cies of Uca have been studied in the field on both sides of the continent, from California to Guayaquil, Ecuador, and from Massachusetts to Rio de Janeiro, Brazil (Crane, 1943, 1944 and unpublished). Now, through a grant from the National Sci- ence Foundation, the work has been expanded into a world-wide program. In 1955 and 1956 the crabs were studied in Pakistan, Ceylon, Ma- laya, Singapore, Sarawak, the Philippines, north and northeast Australia, Fiji, Samoa, Tahiti and adjacent islands. Early in 1957 Panama was revisited, and previously known species reviewed in the light of the recent studies of Indo-Pacific forms. The behavior of the crabs, particularly of displaying individuals, was recorded in color on more than 9,000 feet of 16 mm. motion pic- ture film. This material, along with the correlated and supplementary field notes, forms a perma- nent record for comparative laboratory analysis. The displays of about 19 good Indo-Pacific spe- cies of Uca have been recorded in this manner and 29 American forms. Although extensive field work remains to be done, particularly in Africa, the principal outlines of the development of dis- play in the group appear now to be clear. All detailed supporting data, discussion on the functions and releasing mechanisms of display, specific correlation of display with morphology and discussion of display in relation to infra- specific categories and speciation are reserved for a monographic report on the genus, now in preparation. The following subjects more or less closely associated with display will also be treated only in the later publication : color change, sound 69 70 Zoologica: New York Zoological Society [42; 6 production, territoriality, physical combat be- tween males and the erection of structures near the burrows. Among the many people who have facilitated the field work I wish at this time to express my appreciation particularly to Dr. William Beebe, Director Emeritus of the Department of Tropical Research, New York Zoological Society, and to Dr. Waldo L. Schmitt, Chief Curator of Zoology at the United States National Museum. Both were instrumental in inaugurating the study and, through the years, have given most helpful ad- vice and encouragement. II. General Form of Display in Fiddler Crabs The outstanding characteristic of displaying fiddler crabs is a rhythmic elevation and lower- ing of the hypertrophied major cheliped of the male. In the few species usually familiar to western observers in the field this takes the form of “waving” or “beckoning,” and is character- istic of the breeding season. The significance of the gesture has been described by various work- ers as non-sexually territorial, sexually territorial, a sex attractant, a challenge to other males and as various combinations of all of these possi- bilities. Although it is not proposed at this time to contribute further to the discussion, it may be said that field observations and motion pic- ture records prove conclusively two points., First, all the suggested functions occur, sometimes all in a single species. Furthermore, in some ad- vanced species there are distinct differences separating territorial, male-to-male and male- to-female displays, the latter being the most highly developed. The motion of the cheliped is by no means always wavelike; in some species it is the merest slight raising of the flexed manus and chelae in front of the buccal region. In others it is a vio- lent, very rapid shaking of the same parts at the level of the eye-stalks; in others the base of the manus is rapped against the ground, the waving is in jerks, or the cheliped is held high and re- volved in circles. Among these extremes there are all degrees. Tempo is exceedingly various and specific, although unrelated crabs often wave at about the same rate of speed. Single waves range from one wave lasting 13 seconds to five waves being crowded into a single second. The first extreme of timing is found in certain South American populations of U. pugnax rapax under certain conditions and the second in an appar- ently undescribed species from Port Darwin, Australia. The waving rate of most species falls between one-half and two seconds per wave. Motions associated with waving include ele- vation of the carapace on the ambulatories, re- volving, moving from side to side, and various “bounces” and “curtsies” accomplished by rapid- ly lowering and raising the body on one or both sides. In the following pages the display motion of the large cheliped will for convenience be termed “wave,” whether or not it bears an anthropo- morphic resemblance to such a gesture. III. Basic Wave Patterns Most of the species of Uca may be clearly divided into two groups depending on a basic difference in the direction of the first part of the motion of the major cheliped. These two behavioral complexes are characteristic of the majority of species in, respectively, the “narrow- fronted” and “broad-fronted” groups, as they have been termed in most systematic treatments.^ 3 Bolt, 1954, basing his decision on a collection of specialized Central American species of Uca, gave the narrow-fronted and broad-fronted groups each generic standing, reserving Uca for the narrow-fronts and pro- posing Mimica for the broad-fronts. Peters (1955), working on ecological and behavioral aspects of the same material, reduced Minuca to subgeneric rank. The present writer currently feels that Minuca should be given at most subgeneric status. This view is dependent on the extensive series of intermediate species distributed in other parts of the world, and on the distinctness and homogeneity of Uca, sensu lata, in comparison with all other genera in the family. Until further comparative morphological work is completed, it seems preferable to use simply the non-technical terms, “narrow-fronts” and “broad-fronts,” as a practical division which for the majority of species appears to be phylogenetically justified. Vertical Waves. (Text-figs. 1; 4, A-B). In the great majority of narrow-fronts, the cheliped at the beginning of the wave remains flexed in front of the buccal region and is raised up and slightly forward from there, without unflexing, until it reaches the level of the eyeballs. In the simplest displays it reaches no farther and therefore is never unflexed. In more advanced displays its elevation continues obliquely above the eyes, through the unflexing of the manus and chelae. Regardless of the amount of elevation or degree of unflexing, the cheliped is lowered back into rest position in the same plane in which it was elevated. This type of display will be called a “vertical wave.” Lateral Waves. (Text-figs. 3; 4, D-F). By con- trast, the characteristic wave form of almost all species of broad-fronted Uca commences with a sweep to the side, rather than with a vertical elevation of the cheliped; this kind of gesture will be called a “lateral wave.” In moderately intensive display, in which differences among species are most apparent, the flexed cheliped is pushed away from the body at the beginning 1957] Crane: Basic Patterns of Display in Fiddler Crabs 71 of the wave and more or less unflexed toward the side; it then sweeps up to a completely un- flexed, high-reaching position. Finally it is flexed once more in front of the buccal region. Since it is often lowered in a direct path, close to the body, a more or less circular motion is com- pleted. The amount of deviation from the path of the first part of the wave differs among species, and within species depending on the degree of display intensity; therefore the circularity varies widely; sometimes, in fact, lateral displays are altogether single plane. In displays of the lowest intensity of all, even laterality is often absent, the cheliped motion closely resembling the slight, flexed, single-plane display of the vertically wav- ing narrow-fronts, as described in the preceding paragraph (Text-fig. 4 C) . Intermediate Waves. The waving pattern of a few species is intermediate between vertical and lateral types. Morphologically these forms also show some characters intermediate between those of narrow-fronts and broad-fronts. Position of Body during Waving. In addition to the form of the wave there are several funda- mental differences in the position of the body during waving. In some species it is scarcely raised at all; in many it is raised on the ambula- tories and lowered with every wave; in still others it is raised and held erect during a series of waves. The amount of elevation varies some- what within species in accordance with the in- tensity of waving. Systematic Distribution of Basic Wave Pat- terns. An arrangement of species characterized by various types of wave is presented in Table 2. Authorities for the species, geographical distri- butions and localities where display has been observed are listed in Table I. As will be seen from Table 2, the vertical type of wave is typical of Indo-Pacific narrow-fronts. These species in turn include those forms in which Uca display is simplest and, it seems unquestion- ably, most primitive. These uncomplicated dis- plays are found in manii, rhizophorae (Text-fig. 1, A-B), rosea, ischnodactyla and rathbunae; in these species the body is scarcely if at all raised above the ground during display. U. dussumieri is somewhat further advanced. Although specific differences in the waving pattern of all of these are clear-cut, the group similarity is obvious to the observer. In all of these species the cheliped is not raised high overhead, and the amount of time devoted to waving is negligible in compari- son with that so spent in the more advanced species. The waving of still other narrow-fronts from the Indo-Pacific is considerably more advanced, as in marionis, signata (Text-fig. 1, C-D) and zamboangana (Text-fig. 1, E-F). Although the pattern is still vertical and single plane rather than lateral in character, the reach is higher, the tempo tends to be swifter, the proportion of time devoted to waving greater and the eleva- tion of the body conspicuous. Neotropical members of the narrow-fronted group are aberrant as well as intermediate, both morphologically and behavioristically. The dis- play is characterized by a lateral, not vertical, type of wave. Except in low intensity display the gesture is spiral, rather than merely circular, since the cheliped, after its initial lateral eleva- tion, is revolved throughout a series of waves without being lowered to the ground. This late- ral, spirally circular display has been observed in heteropleura, heterochelos, stylifera, princeps, maracoani and insignis (Text-fig. 2). It reaches its maximum development in the two latter species, which are allopatric counterparts in the Atlantic and Pacific. In both of them, which attain large size and a correlated remark- able development of the heterogonic major cheli- ped, the chelae are directed upward, spread wide apart and then rotated without pause, often for several minutes at a time. Even in their low in- tensity displays these two species are of the lateral type; in the others, however, low intensity waving shows single-plane, vertical-wave affini- ties. Most of the group hold the carapace high during a series of waves, although in stylifera and princeps the hind legs bend during the high- est reach of the cheliped, tilting the posterior part of the carapace downward. U. heteropleura, heterochelos and young princeps raise and lower the carapace with each wave. Two narrow-fronted crabs, the Australian longidigitum and the Indo-Pacific tetragonon, as well as the Indo-Pacific broad-fronted gai- mardi, all show wave characteristics intermediate between the vertical and lateral types. The dis- plays are very distinct, however, and the three species are not morphologically closely related to one another. Alone among the known displays of broad- fronted crabs the neotropical Atlantic thayeri has a vertical, single-plane type of wave similar to those of the Indo-Pacific narrow-fronts. Mor- phologically, also, thayeri shows affinities with those species. The remaining broad-fronts, in which display has been observed, are all characterized by a strongly lateral type of wave, usually with a variable degree of circularity (Text-fig. 3). In a number of the very rapidly waving neo- tropical broad-fronts, such as beebei and salti- tanta, the last half of the wave is so swift that the eye does not record it. Hence although the motion was often described in field notes as a single-plane wave, subsequent examination of 72 Zoologica: New York Zoological Society [42: 6 Text-fig. 1. Examples of vertical types of wave in three species of narrow-fronted Uca from the Indo- Pacific region. Illustrations on left show rest positions between waves, those on right the maximum eleva- tion of the cheliped, which is raised and lowered in a single plane. Note in the series, reading from the top down, the progressively higher reach of the cheliped and greater elevation of the carapace. See text, p. 70 ff.. Table 2 and Text-fig. 4, A-B. A, B, U. rhizophorae (photographed in Singapore); C, D, signata (Philippine Is.); E, F, zamboan- gana (Philippine Is.). Drawings by Dorothy F. Warren, after motion picture frames and mounted specimens. motion picture frames proved that circularity was clearly evident. Examples of a strongly circular type of lateral wave include the Indo-Pacific annulipes and lac- tea (Text-fig. 3, A-D) and the eastern Pacific latimanus. In a few species, such as U. terpsichores, the rest position of the cheliped during display is with the chelae directed forward. An aberrant group of broad-fronts, of which pugnax rapax (Text-fig. 3, E-F) is typical, is an important element in neotropical Uca; two At- lantic representatives extend even into the north temperate region {pugnax, minax). They are all characterized by exceedingly broad fronts and by a jerking, obliquely-lateral wave which dur- ing moderate intensity is circular. They have no Indo-Pacific representatives. The lateral wave of the broad-fronts, in the progressively specialized species, shows increas- ing speed, maximum unflexing of the cheliped and more prolonged periods of time devoted to waving. It has already been noted that in species in which waving is poorly developed {rhizophorae, manii, etc.) the body is raised scarcely or not at all during waving. In the majority of species, among both narrow-fronts and less advanced broad-fronts, the carapace is raised on the am- bulatories and lowered with every wave. In some 1957] Crane: Basic Patterns of Display in Fiddler Crabs 73 lateral wavers, the body is only slightly elevated, but is held in this position throughout the series. (In a few of this group, for example in beebei, an impression of raising-and-lowering is given by the tendency of the crab to bend the hind legs during the peak of the cheliped reach, thus tilting the carapace down posteriorly; this is es- pecially true when the crab is displaying on soft mud; c.f. Peters, 1955, who reported that this form raises and lowers the carapace with every wave). In advanced species among both narrow-fronts and broad-fronts, the body is held high on the ambulatories throughout a series of waves. IV. Advanced Stages of Display One behavioral distinction between the nar- row-fronts and the broad-fronts is found in all the species so far observed except for a few instances. This concerns the final stages of pre- mating behavior. Advanced Display in Narrow-fronts. In the narrow-fronts the male pursues a female or ap- proaches her at the mouth of her burrow; this behavior does not necessarily immediately follow display, although the male always has been dis- playing during the same low-tide period. The approach, however, often does follow display directed more or less toward a female, although this individual is not necessarily the one which is eventually approached. There is no marked increase in tempo of waving during the approach, and additional elements of display are lacking, except for a few special steps in the aberrant neotropical stylifera, insignis and maracoani. When within reach, the male seizes the female and, after brief tapping or stroking of her cara- pace with his ambulatories, attempts copulation at or near the mouth of her burrow. This has been observed by Altevogt (1955) in marionis in India; and by the present author as follows: marionis in Fiji, the Philippines, Australia and Singapore; dussumieri in the Philippines, Sara- wak and Singapore; manii in Singapore and Penang; rathbunae in the Philippines; two unde- scribed new species in Australia; heteropleura, stylifera and insignis in Ecuador and Pacific Panama; and maracoani in Trinidad, northern South America and Brazil. Actual copulation was seen in marionis (four times), rhizophorae (twice), manii (twice), dussumieri (once), Australian new species (once) and stylifera (once). The last stage of courtship, involving stroking, has been observed probably 75 times at a conservative estimate, although these latter episodes did not, after display, end in copulation. A few instances of apparently atypical be- havior is known in the narrow-fronts. Females of stylifera (Crane, 1941, p. 172) insignis and Text-fig. 2. Lateral circular type of wave in the neotropical crab, Uca insignis (photographed in Panama). Carapace is held high on the ambulatories throughout a series of high-reaching circular waves, during which the cheliped is never brought down into the flexed position of rest. A, maximum reach of cheliped; B, “low” position, between waves. See text, p. 71, and Table 2. Drawings by Dorothy F. Warren after motion picture frames and mounted specimens. signata were seen on one occasion each to follow a displaying male into his burrow. Two small princeps {ibid, p. 170) seized females. Each tried unsuccessfully to drag the female down 74 Zoologica: New York Zoological Society [42: 6 his own burrow, grasping her with the ambula- tories of the minor side. During all of the above instances where I have observed final stages of display and copulation above ground I have never once seen the male seize and hold the female with the major cheli- ped. Therefore the definite holding observed by Altevogt near Bombay in copulating marionis (1955.2, p. 518) apparently is of rather excep- tional occurrence. Advanced Display in Broad-fronts. In the broad-fronted group, in addition to an increased tempo of display at the approach of a female, there are specialized steps, curtsies or rappings which are only elicited at this time. Another group distinction is apparent at the end of courtship. In the broad-fronts the usual procedure is for the male, after high intensity display, to precede the female down his own burrow; the female may or may not follow, and may or may not stay below, presumably copu- lating, for any length of time. I have seen the female actually follow the male into the burrow in the Indo-Pacific annulipes in Karachi, Singa- pore, Sarawak and Davao; in lactea in Pakistan, Singapore, the Philippines, and Fiji; in the Paci- fic neotropics in stenodactyla, beebei, batuenta, saltitanta and latimanus; and, finally, in the western Atlantic in pugnax, pugnax rapax, pugi- lator, cumulanta, leptodactyla and in two appar- ently undescribed new species. In only two broad-fronts, stenodactyla and beebei in Panama, was copulation seen; this was at the surface by the female’s burrow as in nar- row-fronts (Crane, 1941, pp. 193, 197). A num- ber of females in each of these two species, however, have been seen to follow the male in the usual broad-front fashion. Throughout the broad-fronted group, therefore, underground mating in the male’s burrow is presumably the rule and the female, instead of being pursued toward her own burrow by the male and often seized more or less forcibly, is attracted by rapid and characteristic display into the burrow of the male which she enters after he has disappeared. In the broad-fronted gaimardi (= pulchella Stimpson) in Tahiti both the first and the last stages of display appear almost perfectly inter- mediate betwen those typical of the primitive narrow-fronts and of the broad-fronts. The in- termediate character of the wave has already been mentioned; advanced stages of display were lacking and the male pursued the female as in narrow-fronted crabs. Even during high intensity display the carapace was scarcely elevated. Mor- phologically, as usual with crabs having inter- mediate displays, the species shows intermediate characters. Another broad-fronted species of particular interest is saltitanta, from the Pacific neotropics. As noted previously (1941, p. 166) this form represents the extreme in development of the rapping type of display, which occurs in a num- ber of related species on both coasts of tropical America. In these species a crab, at the end of a wave, may bounce the lower edge of the major manus and pollex on the ground (Text-fig. 4 F). Usually this bouncing or rapping occurs only when the crab is excited, often in the final stages of courtship. It is typical immediately before the male’s descent into his own burrow, after display has been directed toward a particular female. In saltitanta the rapping has been pushed forward to become an integral part of routine waving, whether or not the crab is particularly excited and whether or not display has been directed toward a female. The basic characteristics of the lateral wave itself may be superceded during the specialized motions of advanced display. Depending on the species, the cheliped may be held motionless overhead, or in front, whUe the crab bobs and races to and fro. Again, the cheliped, having passed through stages of low-intensity single- plane and high-intensity circular display, reverts once more to a single-plane return; now, how- ever, the cheliped is opened laterally and raised to maximum height while the waving tempo is greatly increased. In a number of species the cheliped is not returned to its usual rest position between waves, but performs an aerial circle, somewhat similar to that found in the aberrant neotropical narrow-fronts. In Table 2 only a slight indication is given of the types of advanced display occurring among the broad-fronts. In spite of group simi- larities, the display of each species is so distinct that, if seen in moderate intensity and advanced stages, none could ever be confused in the field with that of any other species, even from other parts of the world. As in studies of the complex displays of cer- tain vertebrates, it is obviously essential, both for accuracy of description and for an approach to understanding, that the displays be observed for prolonged periods and under a variety of conditions. V. Ecological Differences in the Distribution of Display Types In 1941 (p. 160) the present investigator noted that the most highly developed displays in west- ern Central American crabs were found in spe- cies living on the shores of bays and estuaries which, while protected from waves, were freely confluent with the open sea and consequently of 1957] Crane: Basic Patterns of Display in Fiddler Crabs 75 Text-fig. 3. Examples of lateral types of wave in broad-fronted Uca. A-D, U. lactea (photographed in Fiji Is.), showing maximum development of the lateral circular wave, in which the cheliped starting from the flexed position (A) is unflexed outward (B), then raised (C), and Anally returned (D) to the start- ing point. This wave is best developed in displays of moderate intensity; at low intensity or at high inten- sity during advanced display, the wave may be of a vertical or lateral single plane type. See text, p. 71, Table 2 and Text-fig. 4, C-E. E, F, U. piignax rapax (Venezuela). Rest position and maximum cheliped reach of lateral circular wave, characteristic of moderate intensity display. Cheliped is unflexed outward, raised and lowered in a series of jerks. See text, p. 72, and Table 2. Drawings by Dorothy F. Warren, after motion picture frames and mounted specimens. relatively high salinity. Recent observations in the Indo-Pacific have shown agreement. In both the narrow-fronted and broad-fronted groups the most actively displaying species, which ap- parently are also most morphologically ad- vanced, are those in similar localities. Examples of Indo-Pacific crabs with well developed dis- plays are the narrow-fronts tnarionis and tetra- gonon and the broad-fronts annulipes and lactea; all live typically on rather open shores or inlets of protected bays, as do their neotropical coun- terparts including insignis and maracoani among 76 Zoologica: New York Zoological Society [42: 6 the narrow-fronts and stenodactyla, deichmanni, terpsichores, latimanus and leptodactyla among the broad-fronts. Those with the least developed displays, the Indo-Pacific rhizophorae, manii, etc. (p. 71), are also those morphologically the least specialized for an amphibious life; all inhabit less saline situations, sometimes living in almost fresh water, on more sheltered mudflats and streambanks. An interesting point is that the extremely broad-fronted American group, including pug- nax and mordax, has spread into sheltered locali- ties extending even, in mordax, far up tropical rivers. The tempo of their displays is slow, as in sheltered, unrelated species in the Indo-Paciflc, but it is clearly of the lateral, highly developed broad-fronted type, and the females, as usual in the broad-fronts, are attracted down the bur- rows of the males. VI. Geographical Distribution in Relation TO Display Those narrow-fronted species with the sim- plest and apparently most primitive display {manii, etc.) are all confined to the Indo-Malay- an region, which undoubtedly is the center of distribution for the family. The intermediate narrow-front, tetragonon, which currently reaches Tahiti and was formerly found also in Hawaii, shows in its display as in its morphology certain intermediate elements which seem to form a link with the American narrow-fronts. The displays of the latter are more highly evolved than any yet known in the group in the Indo- Pacific, and show affinities to those of the broad- fronts; the species are also specialized morpho- logically, being characterized for example by extreme narrowness of the front, by unique peculiarities of the spoon-tipped hairs of the second maxilliped, and by male abdominal ap- pendages more similar to those of broad-fronts than of Indo-Pacific narrow-fronts. The broad-fronted displays are represented by both simple and advanced examples in both hemispheres. There are, however, many more American than Indo-Pacific species, and most of the American species are more specialized in both behavior and morphology than any of the observed Indo-Pacific broad-fronts with the ex- ception of lactea. U. gaimardi, the only Indo-Pacific broad-front reaching as far east as Tahiti, appears to be, as is tetragonon among the narrow-fronts, an intermediary between Indo-Pacific narrow-fronts and American broad-fronts. Three Indo-Pacific species have been observed in display over a horizontal distribution of 5,000 to 8,000 miles. These are marionis, annulipes Text-fig. 4. Ground projection of path of various types of wave in Uca. In each path the beginning is at upper right boundary of the black line. In A, B and C, the length of the line represents the extent of the obliquely forward thrust of the cheliped dur- ing its elevation. A, U. rhizophorae, a low vertical wave, single plane; B, z^mboangana, a high vertical wave, single plane; C, lactea, a vertical, single plane wave occurring in low intensity display; D, same, a lateral circular wave, typical of display of moderate intensity; E, same, a lateral, single-plane wave found during advanced stages of display; F, saltitanta, a lateral circular wave, with straight line marking route as cheliped retraces beginning of path, rapping ground in front of crab as the claw bounces back into rest position. Cf. text, p. 70, and Text-figs. 1 and 3. These projections were obtained by placing a pre- served, fiexible specimen of each species on a thick sheet of modelling clay and fastening a needle to the tips of the major chelae with scotch tape in such a way that, regardless of the elevation of the cheli- ped, the needle cut partially through the modelling clay. Starting from the flexed, rest position the cheliped was then moved as though in natural dis- play, in accordance with motion picture records, the tip of the needle marking the projected path. and lactea. No gross variations in display be- havior have been observed among the various populations within each species, although de- tailed analyses of motion picture film are ex- pected to show minor distinctions. The one major difference, noted among the populations of lac- tea, was that no “shelters” (Crane, 1941, p. 157) were built by the species even at the height of display in Singapore, the Philippines or Fiji. They were, however, built by about one-fourth of the adult male lactea in Karachi, Pakistan, in mid- June. VII. Evolutionary Trends The varying patterns of Uca display, from the simplest to the most complex, show a trend prev- alent in many branches of the animal kingdom in groups where vision is well developed. In Uca, as in numbers of birds, reptiles, fishes, spiders and mantids, for example, it is axiomatic that the display increases the conspicuousness of the animal— regardless of the function of the par- ticular display under consideration. 1957] Crane: Basic Patterns of Display in Fiddler Crabs 11 A primary component of the conspicuousness is usually a temporary or apparent increase in size. In fiddler crabs this increase is attained in three ways; first by a progressively higher reach with the cheliped; second by the broadening of the wave into an expansive lateral gesture re- placing a close-to-the-body motion; and, third, by the higher and more prolonged elevation of the body on the increasingly straightened walk- ing legs. In some advanced laterally-waving species, however, the body is not held very high, the walking legs being braced far out at the sides; this position seems to be an aid in balanc- ing. It is probably a necessary stance in these species in which both heterogony and the lateral extension of the cheliped are extreme. There are other display characteristics which contribute to the conspicuousness of the crab and which are attained most fully in those species in which waving is highly developed. These char- acteristics include faster tempo of the wave, a rhythm accented by jerks or other specializations and an increase in time devoted daily to waving. The trend toward the attraction of females to the male burrow, typical of the broad-fronts, may well have adaptive value, since pairs copu- lating at the surface are fully exposed to preda- tion by birds. A probable factor in the evolution of this pattern is the fact that among broad- fronts the male is often considerably larger than the female. Because of this size discrepancy males cannot easily descend into the narrow burrows of the females. Many species with highly developed displays resemble primitive forms when waving at low intensity. This is especially well shown when a crab is feeding and waving at the same time, when tidal or meteorological conditions are un- favorable, or when the crab is not ph3'siologically in full display condition. Under any of these circumstances even species as far apart phylo- genetically as maracoani,stenodactyla and lactea may raise the cheliped only slightly and lower it in the same, nearly vertical plane, scarcely elevating the body. This low intensity display therefore is very similar to the full display of rhizophorae, manii and their relations. There appear to be no structural reasons pre- venting vertical wavers from displaying in the lateral fashion. The form of the condyles, the proximal concavities of articulating segments and the elasticity of the connections are similar throughout, although Peters (1955, pp. 489 ff.) illustrates minor differences in the degrees of possible motion in two closely related species. Throughout the genus the cheliped is often wide- ly extended, regardless of the type of wave per- formed during display, in combat between males when the chelae of the combatants are actually engaged. Altevogt (1955.2, p. 514) recorded differences in the display of very large and moderate-sized individuals of marionis, the giants having a less out-reaching form. He thinks this is probably associated with the altered center of gravity. It seems to the present investigator that the evolu- tion of the form of display in the gigantic species of neotropical narrow-fronts, maracoani and insignis, may have been guided by similar prob- lems of weight and balance. Evidence has recently been found that dis- placement behavior sometimes occurs during Uca display, and it seems likely that such behavior varies among the species roughly in accordance with their phylogenetic position. Gordon (1955) first reported displacement feed- ing (in the sense in which the term has been used by Armstrong, 1950, and Tinbergen, 1952) in a number of African species. I fully agree with her interpretation, having often observed similar types of activity not only in Uca but in other genera of displaying ocypodids, notably Ilyoplax. Gordon’s observations were of feeding, often in a token, non-functional fashion, usually after fighting or after high intensity display and a thwarting of mating behavior. The current studies present evidence of two more aspects of the subject. First, displacement behavior is most prevalent in species character- ized by highly developed display. Second, in these species there occurs not only displacement feeding, as described by Gordon, but also dis- placement claw-cleaning. The large manus and chelae are buffed and polished by the small, or at least the requisite motion is made, although the terminal brush may not actually touch the large chelae. The gesture is repeated frequently even though the crab has been above ground for many minutes and the cheliped is altogether free of mud. This activity seems to occur espe- cially when there is conflict between the urge to display and that to escape, such as when the camera or observer is very close to the crab. Current observations indicate furthermore that there will prove to be specific differences in the tendency toward one type or another of displacement behavior. For example, festae in Ecuador, deichmanni in Panama and cumulanta in Trinidad and Venezuela all appear to be especially prone to displacement claw-cleaning. U. lactea in the Indo-Pacific and leptodactyla in Brazil, on the other hand, seem to have a stronger tendency to displacement feeding. In contrast, in the primitive Indo-Pacific narrow-fronts {rhi- zophorae et al.) no displacement behavior has been noticed; it appears likely that in these 78 Zoologica: New York Zoological Society [42: 6 Table 1. Species of Uca in which Display Has Been Studied During the Present Investigation Note: The majority of these forms are undoubtedly good species. It is expected that some will be reduced to subspecific rank, especially certain narrow-fronted crabs with narrow ranges in the Indo-Pacific. This however does not alter the fact that the displays of all the forms listed below are distinct. The 43 species are arranged alphabetically since a phylogenetic presentation must await the publica- tion of correlated morphological evidence. Display has also been observed in at least ten additional species; since these are apparently new and undescribed, they have been omitted from the list. Motion picture records have been obtained of more than three-quarters of all the species observed. In the second column are given references to descriptions of waving published after 1940. An account of earlier work on Uca display is given in Crane, 1941, p. 152. Since these early reports lacked a compara- tive ethological viewpoint they remain, for present purposes, chiefly of historical interest and will not be repeated here. The only references included below, therefore, are recent wave descriptions which are to some degree specifically diagnostic. Published records which are extremely incomplete (e.g. of minax, Crane, 1944) are omitted, as are displays not described but mentioned in connection with another subject (e.g. Gordon, 1955, on displacement behavior). In addition to the displays listed below which Peters (1955) has also recorded, he has described (loc. cit.) those of macrodactyla glabromana Bott and leptochela Bott, neither of which species has been observed by the present author. Hediger (1934) gives a helpful account of tangeri in a reference inadvert- ently omitted by Crane, 1941. Species Description of waving since 1940 General Range Localities where display was observed by author annulipes (Latreille) — Indo-Pacific Pakistan (Karachi), Ceylon, Penang, Singapore, Sarawak, Philippines batuenta Crane Crane, 1941 Peters, 1955 E. Pacific Panama, Ecuador beebei Crane Crane, 1941 Peters, 1955 E. Pacific Panama, Ecuador cumidanta Crane Crane, 1943 W. Atlantic Venezuela, Trinidad deichmanni Rathbun Crane, 1941 E. Pacific Panama dussumieri (Milne Edwards) — Indo-Pacific Singapore, Sarawak, Philippines festae Nobili — Ecuador Ecuador gaimardi (Milne Edwards) — Western & Central Pacific French Oceania (Tahiti, Raiateia, Bora Bora) galapagensis Rathbun — E. Pacific Ecuador heterochelos (Lamarck) — W. Atlantic Venezuela heteropleura (Smith) Crane, 1941 E. Pacific Panama, Ecuador inaequalis Rathbun Crane, 1941 Peters, 1955 E. Pacific Panama, Ecuador insignis (Milne Edwards) — E. Pacific Panama, Ecuador inversa (Hoffmann) — Indo-Pacific Pakistan (Karachi) ischnodactyla Nemec — Fiji Fiji lactea (de Haan) Indo-Pacific Pakistan (Karachi), Singapore, Philippines, Fiji, Samoa latimana (Rathbun) Crane, 1941 Peters, 1955 E. Pacific Panama, Ecuador leptodactyla (Guerin) — W. Atlantic Venezuela, Brazil limicola Crane — E. Pacific Panama longidigitum (Kingsley) — E. Australia Nr. Brisbane 1957] Crane: Basic Patterns of Display in Fiddler Crabs 79 Table 1. Species of Uca in WHICH Display Has Been Studied During THE Present Investigation ( Continued) Species Description of waving since 1940 General Range Localities where display was observed by author manii Rathbun — Indo-Pacific Singapore maracoani (Latreille) Crane, 1943 W. Atlantic Venezuela, Trinidad, British Guiana, Surinam, Brazil marionis (Desmarest) (incl. nitida (Dana)) Altevogt, 1955.1, 1955.2 Indo-Pacific Singapore, Sarawak, East & North Australia, Philippines, Fiji minax (Le Conte) — W. Atlantic New Jersey mordax (Smith) Beebe, 1928 Crane, 1943 W. Atlantic Guatemala, Venezuela, Trinidad, Brazil oerstedi Rathbun Crane, 1941 E. Pacific Panama olympioi Oliveira — S. Brazil Rio de Janeiro panamensis (Stimpson) — E. Pacific Panama princeps (Smith) Crane, 1941 Peters, 1955 E. Pacific Panama, Ecuador pugilator (Bose) Crane, 1944 Burkenroad, 1947 W. Atlantic Connecticut, New York, Florida pugnax (Smith) Crane, 1944 W. Atlantic Massachusetts, Connecticut, New York, Florida piignax rapax (Smith) Crane, 1943 W. Atlantic S. Florida, Guatemala, W. Indies, Colombia, Venezuela, British Guiana, Surinam, Brazil rathbunae Pearse — Philippines Manila, G. of Davao rhizophorae (Tweedie) — Malaya & Borneo Singapore, Sarawak rosea (Tweedie) — Malaya Penang saltitanta Crane Crane, 1941 Peters, 1955 E. Pacific Panama si gnat a (Hess) — Indo-Pacific Australia (Gladstone), Philippines speciosa (Ives) — W. Atlantic S. Florida stenodactyla (Milne Edwards & Lucas) Crane, 1941 Peters, 1955 E. Pacific Costa Rica, Panama, Ecuador stylifera (Milne Edwards) Crane, 1941 E. Pacific Panama, Ecuador terpsichores Crane Crane, 1941 E. Pacific Panama, Ecuador tetragonon (Herbst) — Indo-Pacific French Oceania (Bora Bora) thayeri Rathbun — W. Atlantic Trinidad, Brazil zamboangana Rathbun — Philippines Gulf of Davao species display is so feebly developed that, when any conflict or frustrating situation arises, the crab simply stops waving. No comment can yet be made on the role displacement behavior, through ritualization. may have played in the evolution of display motions. Studies on heliconiid butterflies in Trinidad (Crane, 1955 and in ms.) apparently show con- clusively the frequent occurrence of displace- 80 Zoologica: New York Zoological Society [42: 6 Table 2. Arrangement of Uca spp. According to General Type of Wave. Key: Species names in italics', narrow-fronts. Species names in bold-face: broad-fronts. IP : Indo-Pacific. curtsy ^ general type of EP: Eastern Pacific. rapping V outstanding characteristic WA: Western Atlantic. special steps \ in advanced display. No AD: No special advanced display characteristics. (Where distribution symbol is not followed by one of these key designations, advanced display is incom- pletely known. See text). Wave Vertical Wave Lateral Body position at moderate display intensity (Cheliped raised up- ward at beginning of wave; always re- turned to position in same plane) (Cheliped extended to side at beginning of wave; often returned to position via a circular route) Body raised throughout a series of waves Body conspicuous- ly raised and lowered with each wave. Body not raised, at all, or mini- mally raised and lowered with each wave. zamboangana (IP) dussumieri (IP) (No AD) marionis (IP) (No AD) signata (IP) (No AD) thayeri (WA) (No AD) annulipes (IP) (curtsy) beebei (EP) (special steps) festae (EP) (special steps) galapagensis (EP) inversa (IP) lacfea (IP) (curtsy) Eatimanus (EP) (curtsy) leptodactyla (WA) (curtsy) lEinicola (EP) minax (WA) mordax (WA) (curtsy) olympioi (WA) (curtsy) pugnax (WA) (curtsy) stenodactyla (EP) (special steps) terpsichores (EP) (special steps) insignis (EP) (special steps) maracoani (WA) (special steps) princeps (EP) stylifera (EP) (special steps) batuenta (EP) (rapping) cumulanta (WA) (rapping) deichmanni (EP) (rapping) inaequalis (EP) (rapping) cerstedi (EP) panamensis (EP) pugilafor (WA) (rapping) saltitanta (EP) (rapping) speciosa (WA) (rapping) heterochelos (WA) heteropleura (EP) (No AD) gaimardi (IP) (No AD) longidigitum (Australia) tetragonon (IP) (No AD) ischnodactyla (IP) (No AD) manii (IP) (No AD) tathbunae (IP) (No AD) rhizophorae (IP) (No AD) rosea (IP) (No AD) 1957] Crane: Basic Patterns of Display in Fiddler Crabs 81 ment behavior in these insects. Its probable occurrence has been mentioned in salticid spid- ers {idem, 1948, p. 202), mantids {idem, 1952, p. 288) and Drosophila (Bastock & Manning, 1955, p. 104) . It now seems likely that displace- ment behavior will prove to be a noteworthy factor in the ethological study of many higher invertebrates. VIII. Summary Two basic patterns of display have been dis- tinguished during field studies of more than fifty species of fiddler crabs {Uca). The first pattern is characteristic of a group of species with narrow fronts. It is distinguished by a simple, more or less vertical gesture (“wave”) made with the major cheliped of the male, and by the male’s pursuit of the female toward her burrow; copulation in the known instances takes place on the surface of the ground. The second pattern is typical of broad-fronted species in the genus. It is characterized as follows; the cheliped is unflexed laterally, rather than verti- cally elevated, and sometimes completes a cir- cular motion in returning to rest position; there is in addition a distinct second stage of display which is usually elicited by the approach of a female and which depends both on special move- ments of the various appendages and on an increased tempo of waving; finally, in the last stage of display the male attracts the female down his own burrow, which he enters first. A few species with intermediate types of be- havior have been observed, especially in the Indo-Pacific. The simplest and most primitive of the nar- row-fronted displays are found in a group of closely related Indo-Pacific species of narrow distribution which inhabit protected estuaries and tidal streams; manii and rathbunae are ex- amples. Displays of higher development are found among species inhabiting more saline, exposed locations; typical of these are tetra- gonon and zamboangana, both from the Indo- Pacific area. The narrow-fronted crabs reach their highest display development, along with their greatest morphological specialization, in the neotropical representatives, culminating in insignis and maracoani. All favor relatively ex- posed habitats. The broad-fronted, laterally waving fiddler crabs are distributed in both the Indo-Pacific and American regions but, unlike the narrow-fronted species, they are poorly represented in the Indo- Pacific. As in the other groups, the highest dis- play development occurs among species living in more seaward niches. Representatives are found of both simple and highly evolved dis- plays. As examples, gaimardi, from the western and central Pacific, illustrates the simplest known type of lateral display, while complex specializa- tions are shown variously by lactea from the Indo-Pacific, saltitanta and terpsichores from the eastern Pacific and pugnax from the western Atlantic. Three wide-ranging Indo-Pacific species have been observed in display over distances ranging from 5,000 to 8,000 miles. No gross intraspecific differences were noted on the peripheries or elsewhere in their ranges, except that some dis- playing lactea built small structures of sandy mud in Karachi, but not in either the Philippines or the Fiji Islands, at least during the period of observation. As is usual in a number of animals, the gen- eral trend in display evolution in Uca is toward increased conspicuousness. This increase is attained principally by higher speed in waving, by greater complexity in the rhythms and forms of display motions and by increased apparent size, through extension of the appendages either vertically or horizontally. The species in which display is highly evolved also spend a greater portion of their time in display than do other members of the genus. Displacement behavior during Uca display is briefly discussed, including its possible role in the systematics of the group. IX. References Altevogt, a. 1955.1 Some studies on two species of Indian fid- dler crabs, Uca marionis nitidus (Dana) and U. annulipes (Latr.). Jour. Bombay Natural History Soc., 52:700-716. 1955.2 Beobachtungen und untersuchungen an indischen winkerkrabben. Z. Morph, u. Okol. Tiere, 43:501-522. Armstrong, E. A. 1950. The nature and function of displacement activities. Symposia of the Society for Experimental Biology, No. 4. Physiologi- cal mechanisms in animal behavior. Aca- demic Press, N. Y. 361-384. Bastock, M. & A. Manning 1955. The courtship of Drosophila melanogaster. Behaviour, 8:85-111. Bott, R. 1954. Dekapoden (Crustacea) aus El Salvador. 1. Winkerkrabben {Uca). Senck. Biol., 35:155-180. Burkenroad, M. D. 1947. Production of sound by the fiddler crab, Uca pugilator Bose, with remarks on its nocturnal and mating behavior. Ecology, 28:458-461. 82 Zoologica: New York Zoological Society [42: 6: 1957] 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, 26:145-208. 1943.1 Crabs of the genus Uca from Venezuela. Zoologica, 28:33-44. 1943.2 Display, breeding and relationships of fid- dler crabs (Brachyura, genus Uca) in the northeastern United States. Zoologica, 28:217-223. 1952. A comparative study of innate defensive behavior in Trinidad mantids (Orthoptera, Mantoidea) . Zoologica, 37:259-293. 1955. Imaginal behavior of a Trinidad butterfly, Heliconius erato hydara Hewitson, with special reference to the social use of color. Zoologica, 40:167-196. Gordon, H. R. S. 1955. Displacement activities in fiddler crabs. Nature, 176 (4477) :356-357. EXPLANATION Plate I Oblique views of different types of wave in fiddler crabs. For comparison with frontal views. Text-figs. 1 and 3. Fig. 1. Vertical wave: Uca dussumieri displaying near Sasa, Gulf of Davao, Philippine Islands. Maximum elevation of cheliped. Fig. 2. Vertical wave: Uca marionis in Singapore. Hediger, H. 1934. Zur biologie und psychologic der flucht bei tieren. Biol. Zentralbl. Leipzig, 54:21- 40. Peters, H. M. 1955. Die winkgebarde von Uca und Minuca (Brachyura) in vergleichend-ethologi- scher, okologischer und morphologisch- anatomischer hetrachtung. Z. Morph, u. Okol. Tiere, 43:425-500. Tinbergen, N. 1951. The study of instinct. Oxford, at the Clar- endon Press. 228 pp. 1952. “Derived” activities; their causation, bio- logical significance, origin, and emancipa- tion during evolution. Quart. Rev. Biol., 27: 1-32. Vervey, J. 1930. Einiges fiber die biologie ost-indischer mangrovekrabben. Treubia, 12:169-261. OF THE PLATE Major cheliped partially elevated. Note for- ward direction of merus and carpus, and compare their position in Fig. 4, below. Fig. 3. Lateral wave: Uca latimanus near Panama City, Panama. Maximum elevation of cheli- ped. Fig. 4. Lateral wave: Uca annulipes near Santo- bong, Sarawak. Cheliped partially elevated. Note lateral direction of merus and carpus. CRANE PLATE I FIG. 1 FIG. 3 FIG. 2 FIG. 4 BASIC PATTERNS OF DISPLAY IN FIDDLER CRABS (OCYPODIDAE, GENUS UCA) NEW YORK ZOOLOGICAL SOCIETY GENERAL OFFICE 30 East Fortieth Street, New York 16, N, Y. PUBLICATION OFFICE The Zoological Park, New York 60, N. Y. OFFICERS -PRESIDENT VICE-PRESIDENTS SECRETARY TREASURER Fairfield Osborn Alfred Ely Harold J. O’Connell David H. McAlpin Laurance S. Rockefeller SCIENTIFIC STAFF: Zoological Park and Aquarium John Tee- Van General Director ZOOLOGICAL PARK Leonard J. Goss Assistant Director and Veterinarian William G. Conway. .Associate Curator of Birds Grace Davall. ...... .Assistant Curator, Mammals and Birds James A. Oliver Curator of Reptiles Charles P. Gandal. . .Associate Veterinarian Lee S. Crandall General Curator Emeritus William Beebe Honorary Curator, Birds AQUARIUM Christopher W. Coates . Director James W. Atz Associate Curator Carleton Ray Assistant to the Director Ross F. Nigrelli Pathologist Myron Gordon Geneticist C. M. Breder, Jr Research Associate in Ichthyology Harry A. Charipper. . .Research Associate in Histology Homer W. Smith Research Associate in Physiology GENERAL William Bridges Editor & Curator, Publications Sam Dunton Photographer Henry M. Lester . . . Photographic Consultant DEPARTMENT OF TROPICAL RESEARCH William Beebe .... Jocelyn Crane .... David W. Snow .... Henry Fleming . . . John Tee-Van .... William K. Gregory AFFILIATES C. R. Carpenter Co-ordinator, Animal Behavior Research Programs L. Floyd Clarke Director, Jackson Hole Research Station Director Emeritus Assistant Director . Resident Naturalist Entomologist Associate Associate SCIENTIFIC ADVISORY COUNCIL A. Raymond Dochez Alfred E. Emerson W. A. Hagan Caryl P. Haskins K. S. Lashley John S. Nicholas EDITORIAL COMMITTEE Fairfield Osborn, Chairman James W. Atz Lee S. CrandaU William Beebe Leonard J. Goss William Bridges James A. Oliver Christopher W. Coates John Tee-Van ZOOLOGICA SCIENTIFIC CONTRIBUTIONS OF THE NEW YORK ZOOLOGICAL SOCIETY VOLUME 42 • PARTS • NOVEMBER 25, 1957 • NUMBERS 7 TO 10 PUBLISHED BY THE SOCIETY The ZOOLOGICAL PARK, New York Contents 7. Studies on the Lizard Family Xantusiidae. III. A New Genus for Xantusia river siana Cope, 1883. By Jay M. Savage. Text-figures 1-3 8. Nesting Behavior of the Crested Oropendola (Psarocolius decumanus) in Northern Trinidad, B.W.I. By Richard E. Tashian. Plates I & II; Text- figures 1-3 9. Changes in the Cytological Structure of the Adenohypophysis and Gonads in Juvenile Bathygobius soporator after Pituitary Implantation. By Louise M. Stoll. Plates I & II 10. The Ctenuchidae (Moths) of Trinidad, B.W.I. Part I. Euchromiinae. By Henry Fleming. Plates I-III PAGE 83 87 99 105 7 Studies on the Lizard Family Xantusiidae. III. A New Genus for Xantusia riversiana Cope, 1883 Jay M. Savage Department of Biology, University of Southern California, Los Angeles 7, California (Text-figures 1-3) IN the course of investigations leading toward a complete revision of the night lizard fam- ily Xantusiidae, it has become apparent that Xantusia riversiana Cope, 1883, is so dif- ferent from other members of the group that a new genus is required for its reception. The decision to place riversiana in a separate genus is based upon an analysis of the scutellation, body proportions, coloration, life history and habits of all known xantusiids, combined with an extensive study of the osteology of the family. Detailed consideration of these features, char- acterization of the genera and species of xantu- siids and discussion of the evolutionary patterns within the family are reserved for the revisional report. However, it seems advisable to propose the new genus in advance of the appearance of the larger work in order to make the name avail- able to others investigating various aspects of the biology of the Xantusiidae. To increase the usefulness of the present paper an artificial key to the genera of night lizards is appended to the description of the new genus. There has been considerable confusion in the past concerning the status and relationships of several genera and species of Xantusiidae be- cause of the lack of a standardized terminology for the squamation of these lizards. In order to analyze the scutellational differences and simi- larities among members of the family it has been necessary to revise completely the existing system employed for the scales on these lizards. A de- tailed discussion and definition of scutellational terms will be included in my revision of the fam- ily, but for the sake of consistency the new sys- tem of scale terminology is used throughout the present report. The head shields of riversiana are illustrated in Text-figures 1-3. The terms radials and femorals refer to the scales along the anterior surface of the forearm and thigh, respectively. Walker (1955) has utilized my system of scale nomenclature in his descriptions of new Lepi- dophyma from Mexico. It is a great pleasure to have the opportunity of naming what may well be the last new genus of recent reptiles from western North America in honor of Dr. Laurence M. Klauber of San Diego, who has contributed so magnificently to our understanding of the herpetofauna of this region. Klauberina, new genus Type of Genus— Xantusia riversiana Cope, 1883. Definition.— External characteristics: (1) one frontonasal; (2) a large median; (3) two front- als; (4) two parietals; (5) no pretemporals; (6) temporals a series of enlarged plates; (7) two rows of supraoculars; (8) nostrils pierced at juncture between nasal, postnasal, rostral and first supralabial; (9) pretympanics granular; (10) postmentals and infralabials distinct; (11) anterior postmentals paired; (12) anterior pre- gulars granular; (13) gulars enlarged rectangular plates, much larger than pregulars; (14) dorsal scales granular; (15) dorsals essentially homo- geneous in size and shape; (16) ventrals in 16 longitudinal rows; (17) radials and femorals forming enlarged plates; (18) caudals faintly keeled; (19) all scales in caudal whorls of same size. Osteological characteristics: (20) nasals meeting one another for most of their length; 83 84 Zoologica: New York Zoological Society [42: 7: 1957] Text-fig. 1. Diagram of dorsal head scutellation of Klauberina riversiana (Cope, 1883). R = rostral; FN - frontonasal; M = median; F - frontal; P =: parietal; O = supraocular series. Text-fig. 2. Diagram of lateral head scutellation of Klauberina riversiana (Cope, 1883). N =: nasal; PN - postnasal; PT = pre- tympanics; S = supralabial series; T = temporals. (21) premaxillary not reaching to frontals; (22) frontals paired; (23) frontals not roofing over orbits above, outer margins concave; (24) pari- etals paired; (25) squamosal touching parietal; (26) prefrontal extending onto top of skull, ex- cluding a portion of frontal from orbit; (27) jugal moderately broad; (28) postfronto-orbital not expanded medially, gradually decreasing in width from anterior to posterior; (29) anterior bony palate of paleochoanate type; (30) fenes- tra vomeronasalis externa bordered by prevomer and maxillary; (31) epipterygoid process from prootic; (32) ectopterygoid and palatine in con- tact or narrowly separated; (33) basisphenoid and basioccipital distinct; (34) teeth on both jaws strongly triconodont; (35) clavicles per- forate; (36) sternum without posterior projec- tion; (37) phalangeal formula of hand 2-3-4-S-3. Diagnosis.— new genus may be readily sep- arated from Cricosaura Gundlach & Peters, 1863, and Lepidophyma Auguste Dumeril, 1851 (in- cluding Gaigeia H. M. Smith, 1939), by the characters presented in the accompanying key. In addition Klauberina is distinct from these genera in a number of the features mentioned in the definition above (from Cricosaura in char- acters 1-5, 8-9, 11-12, 14-16, 18-22, 24-25, 27- 32, 34-37, and from Lepidophyma in 5, 7, 8, 10-11, 13, 15-17, 23, 26-28, 34), but a complete summary of the differences between the various genera will be given elsewhere. From its nearest ally, Xantusia Baird, 1859, Klauberina is distinguished by having (charac- ters for Xantusia given in parentheses) : external: no pretemporal scales (pretemporals present); temporals formed of a series of enlarged plates (temporals formed of small granules) ; two rows of supraoculars (one row of supraoculars) ; ven- tral scales in 16 longitudinal rows (ventrals in 12-14 rows) ; caudal scales slightly keeled (caud- als smooth) ; osteological: jugal moderately broadened (jugal reduced to a narrow sliver); ectopterygoid and palatine in contact or nearly meeting (ectopterygoid and palatine widely 1957] Savage: Studies on the Lizard Family Xantusiidae 85 Text-fig. 3. Diagram of squamation on underside of head of Klauberina riversiana (Cope, 1883). I = infralabials; PM = postmentals. The pregular scales are the small scales lying median and pos- terior from the postmentals and anterior to the gular fold; the gulars are the scales on the gular fold. separated); teeth strongly triconodont (teeth simple) . Klauberina riversiana is markedly different from the recently described Eocene fossil xantu- siid, Palaeoxantusia fera Hecht, 1956, based upon a portion of a lower jaw from Elk Moun- tain, Wyoming (Bridger Formation), in having strongly triconodont teeth. The fossil genus has simple rounded teeth. The nominal night lizard genus Impensodens erected on the basis of an incomplete lower jaw from Post-Pleistocene deposits in Yucatan, Mex- ico (Langebartel, 1953), appears to be synony- mous with Lepidophyma according to Hecht (1956, p. 3). In any event this mandible is dis- tinct from that of Klauberina in having weakly triconodont teeth with the secondary cusps lo- cated on the inner surface of the teeth and not in the same longitudinal plane as the primary median cusp.. Klauberina is distinctive within the family in having strongly triconodont teeth with all the cusps on a given tooth located along the same longitudinal plane. Included Species. — Klauberina riversiana (Cope, 1883) of the California Channel Islands, Santa Barbara, San Clemente and San Nicholas, is the unique member of the genus. Notes on the Family.— In a recent discussion of the position of the Xantusiidae within the suborder Sauria, McDowell & Bogert (1954, pp. 94-98) acknowledge the aid of my unpublished study on the family as a source for certain of their remarks. Unfortunately the information presented by them cannot be derived from the data provided by me. Corrections therefore seem in order. McDowell & Bogert state that among other features the family Xantusiidae is characterized by paired parietal bones, absence of a parietal foramen and ovulation involving but two eggs at one time (with rare exceptions). Contrary to these statements, one xantusiid, Cricosaura typica Gundlach & Peters, 1863, has the parietal bones fused to form a single element. Neither can the family be characterized as lacking a parietal foramen. Members of the genus Xan- tusia always have this aperture and Klauberina also consistently has the foramen (as clearly shown on figure 25 in McDowell & Bogert), although it may be partially or entirely covered by osteoderms on the outer surface of the skull in adult examples. The published studies of Brattstrom (1951) and Shaw (1949) conclu- sively show that Klauberina may produce four to ten ova at one time and my own investiga- tions indicate that Lepidophyma usually has about six eggs formed at one period of ovulation. The species of Xantusia produce one to three ova but the usual number is two. In addition to these points, it may be noted that postanal bones and sacs are present in males of both Xantusia ari- zonae Klauber, 1931, and Xantusia vigilis Baird, 1 859, not in vigilis alone as stated by McDowell & Bogert. An Artificial Key to the Recent Genera OF Night Lizards la. A single frontal scale; no parietal scales; a single anterior postmental scale; nasal bones completely separated from one another by nasal process of premaxillary; frontal bone single; parietal bone single; phalangeal formula of hand 2-3-4-4-3 Cricosaura lb. Two frontal scales; two parietal scales; a pair of anterior postmentals or postmentals fused with infralabials; nasal bones in contact with one another for most of length; two frontals; two parietals; phalangeal formula of hand 2-3-4-5-3. 2a. Supraoculars well-developed; postmentals and infralabials distinct; scales on back and sides essentially homogeneous in size and shape; radials and femorals enlarged into plates; orbit not completely roofed over above by frontal bone, outer margin of frontal concave; prefrontal bone ex- tending onto top of skull, separating an- terior portion of frontal from orbit; 86 Zoologica: New York Zoological Society [42: 7 postfronto-orbital not expanded medially, gradually decreasing in width posteriorly. 3a. No pretemporal scales; temporal scales enlarged plates; two rows of supraoculars; ventrals in 16 longi- tudinal rows; jugal expanded; ecto- pterygoid and palatine in contact or narrowly separated from one another; teeth triconodont Klauberina 3b. Pretemporals present; temporals small, granular; one row of supra- oculars; ventrals in 12-14 longitudinal rows; jugal reduced to a narrow sliv- er; ectopterygoid and palatine widely separated; teeth simple. . . .Xantusia 2b. Supraoculars reduced to a fleshy flap projecting from side of frontal scale; postmentals and infralabials fused into a single series; scales on back and sides a heterogeneous mixture of granules and enlarged scales; radials and femorals not enlarged, same size as other scales on limbs; orbit completely roofed over above by frontal bone, outer margin of frontal even, not concave; prefrontal bone not extending onto top of skull, not separating anterior portion of frontal from orbit; postfronto-orbital expanded medially, not gradually decreasing in width posteriorly Lepidophyma Literature Cited Brattstrom, Bayard Holmes 1951. The number of young of Xantusia. Her- petologica, vol. 7, pt. 3, pp. 143-144. Cope, Edward Drinker 1883. Notes on the geographical distribution of Batrachia and Reptilia in western North America. Proc. Acad. Nat. Sci. Philadel- phia, 1883, pp. 10-35. Hecht, Max Knobler 1956. A new xantusiid lizard from the Eocene of Wyoming. American Mus. Novitates, no. 1774, 8 pp., 2 text-figs. Langebartel, Dave A. 1954. Faunal and archeological researches in Yucatan caves. 4. The reptiles and am- phibians. Bull. 33, Cranbrook Inst. Sci., pp. 91-108, 2 text-figs. McDowell, Samuel Booker, & Charles Mitchell Bogert 1954. The systematic position of Lanthanotus and the affinities of the anguinomorphan lizards. Bull. American Mus. Nat. Hist., vol. 105, art. 1, pp. 1-142, pis. 1-16, text- figs. 1-43. Shaw, Charles Edward 1949. Notes on broods of two xantusiids. Her- petologica, vol. 5, pt. 2, pp. 23-26. Walker, Charles Frederick 1955. Two new lizards of the genus Lepidophy- ma from Tamaullpas. Occ. Paps. Mus. Zool. Univ. Michigan, no. 564, pp. 1-10. 8 Nesting Behavior of the Crested Oropendola (Psarocolius decumanus) in Northern Trinidad, B.W.I/ Richard E. Tashian^ Department of Tropical Research, New York Zoological Society, New York 60, New York (Plates I & II; Text-figures 1-3) [This paper is one of a series emanating from the tropical Field Station of the New York Zoological Society, at Simla, Arima Valley, Trinidad, British West Indies. This station was founded in 1950 by the Zoological Society’s Department of Tropical Research, under the direction of Dr. William Beebe. It comprises 200 acres in the middle of the Northern Range, which includes large stretches of undis- turbed government forest reserves. The laboratory of the station is intended for research in tropical ecology and in animal behavior. The altitude of the research area is 500 to 1,800 feet, with an annual rainfall of more than 100 inches.] Contents Introduction The Nests The Pre-incubation and Incubation Periods The Nestling Period Vocalization and Display Cowbird Activity The Nesting at Colony B Miscellaneous Observations Discussion and Summary Literature Cited Introduction The oropendolas or giant caciques con- stitute an assemblage of neotropical birds, within the family Icteridae, made up of five genera (Ocyalus, Zarhynchus, Clypicterus, Gymnostinops and Psarocolius) and 12 species. They are among the largest passerine birds of the New World tropics, reaching 21 inches in iContribution No. 977, Department of Tropical Re- search, New York Zoological Society. ^Present Address: University of Michigan, 1133 E. Catherine St., Ann Arbor, Michigan. Page . 87 . 88 . 89 . 90 . 92 . 94 . 95 . 95 . 96 . 96 length, and are characterized by similar social habits, colonial nesting, vocalizations and dis- play patterns. Their dominant colors are shades of black and brown with yellow or orange char- acteristically present in the tail and bill. Another closely related group of icterids are the true caciques (i.e., Cassiculus spp.) which, although quite similar in habits and appearance to the oropendolas, average smaller in size. Previous studies on the 12 species comprising the oro- pendolas have been limited to the life history studies of the Wagler Oropendola {Zarhynchus wagleri) by Chapman (1928) and the Monte- zuma Oropendola {Gymnostinops montezuma) by Skutch (1954). The Crested Oropendola or Giant Crested Cacique {Psarocolius decumanus insularis) is uniform brownish-black in general coloration, becoming chestnut on the rump and under tail coverts; the tail is lemon yellow with the middle pair of rectrices brownish-black; the bill pale greenish-yellow; the tarsi and toes black and the iris blue. The sexes are similar in color but differ in size, the female averaging 13 inches in length and the male 17 inches. Five subspecies of this oropendola are recognized, ranging from Panama across northern South America south to southern Brazil, Paraguay, northeastern Ar- gentina, eastern Peru and eastern Bolivia. The race insularis is restricted to the islands of Trinidad and Tobago and northeastern Vene- zuela. The genus Psarocolius also includes the Green Oropendola (F. viridis), the D’Orbigny Oropendola (F. atro-virens) and the Chestnut Oropendola (F. angustifrons) . Two nesting colonies of the Crested Oro- 87 88 Zoologica: New York Zoological Society [42: 8 pendola were discovered in early January, 1955, in the cocoa plantation and secondary forest near Simla about four miles north of the town of Arima in the Northern Range of Trinidad. This area is situated in the Arima Valley at an elevation of 800 feet, and is bordered by lower montane rain forest which was, in all proba- bility, the original forest type of the study area. For further ecological details of meteorology and biotic zones of this region see Beebe (1952). Both nesting sites were established in long-boled mountain immortelle or erythrina trees {Ery- thrina micropteryx) with the sack-like nests hanging some 60 to 80 feet from the ground. These immortelles are familiar nesting sites of the Crested Oropendola in the Arima valley, and are common trees on the cocoa plantations where they were originally introduced as cover for the cocoa. The birds were still in the process of nest- building when their sites were found. A blind was constructed on a slope opposite the more accessible colony (Colony A) at a level with most of the nests and about 100 feet away. This colony was visited daily from Jan. 12 to March 29 for a total of 303 hours. The second colony (Colony B), about a quarter of a mile from the first, was visited weekly from Feb. 6 to April 3 for a total of 15 hours. Observations were made with the aid of 7 X 50 binoculars, and moving pictures and photographs of special activities were taken through a six-inch or 300 mm. tele- photo lens. The majority of the observation periods took place between 9:00 A.M. and noon in the morning and 2:30 P.M. to 6:00 P.M. in the afternoon. However, a number of observa- tions were also made between dawn and 9:00 A.M. and from 6:00 P.M. to darkness. Since it was not possible to look into the in- accessible nests, it was especially important to keep them under frequent observation in order to determine any change in rhythm of the nest- ing visits. In this manner a good approximation of the dates of egg laying and hatching could be made. It was felt that the observation periods were of such regularity that determination of these dates is accurate to within 24 hours. Un- less otherwise stated, the figures given for the frequency of the attentive and inattentive pe- riods of the parent at the nest during the various nesting phases are included only if the period of observation exceeded three hours. A nest visit implies complete entry into the nest, except on those occasions during advanced feeding when the female fed the nestling without entering the nest. The average monthly maximum and mini- mum temperatures at Simla for Jan., Feb. and March, 1955, were 61-89, 66-88 and 66-88° F. respectively. The total monthly rainfall for the same months was: Jan., 2.20; Feb., 2.29; March, 2.68 inches. I am especially grateful to Dr. William Beebe for his many helpful suggestions during this study as well as for placing his highly informa- tive notes at my disposal. I also wish to thank Miss Rosemary Kenedy for assisting me so gen- erously in the photographic phases of this work. Dr. Edward J. Alexander of the New York Bo- tanical Garden kindly identified the nesting material. The Nests Twenty-three nests were begun at colony A (Plate I, Fig. 1), of which 11 were completed, and at colony B, 14 nests were begun, all of which were completed. The nests at colony A were fairly evenly distributed throughout the tree, whereas at colony B they were bunched together on one side of the tree. Incidentally, at another colony of Crested Oropendolas, found on March 14 in an adjacent valley at an eleva- tion of about 1,500 feet, the nests were sus- pended in an immortelle tree some 50 feet above the ground and were all arranged very closely together for a horizontal distance of only about 20-25 feet. Eighteen nests appeared to have been started, of which 14 seemed complete. During a half hour’s watch feeding of the nest- lings was observed at four separate nests. High winds apparently take a considerable toll of nests. Four nests (three occupied and one unoccupied) at colony B, and one unoccupied nest at colony A, fell during heavy winds. Two of these fallen nests seemed typical and measure- ments are given in Table 1. A great increase in the length of the nest open- ing and the dispersal of the leaf lining is ap- parent between incubation and the third week of feeding. Another unoccupied nest, obviously abnormal, measured seven feet in length and weighed 312 grams. Examination of these nests showed that they were composed largely of the stripped leaf fibers of bromeliads (Bromeliaceae) and heliconia (Musaceae) and stems of vines (Cucurbita- ceae). The leaves used in lining the bottom of the nests on the inside were almost exclusively those of the immortelle (^Erythrina microp- teryx). According to Goeldi (1897, 1903) the Crested Oropendola from southern Brazil em- ploys the fibers of the hromeliad Tillandsia usneoides almost entirely as a nesting material, whereas near the Amazon, orchid roots and elongated roots of the rhizomorph Marasimius sp. are used. Young (1929) describes a nest 1957] Tashian: Nesting Behavior of Crested Oropendola 89 Table 1. Measurements of Two Nests of Psarocolius decumanus, in Incubating AND Feeding Nesting Stages Length of opening Weight (gms.) Nesting phase Length Width Fiber Leaf lining Total Incubating 2T0" 8" 5" 93.1 87.7 180.8 Feeding 2' 11" 9" 1'5" 178.8 4.1 182.9 from British Guiana as made up of narrow strips of cane leaves, plantain bark, cocoanut fiber and vines. He also states that two or three nests are sometimes built by the same female. The color of some broken egg shells recov- ered from a fallen nest was a pale bluish-green or, by comparison with Ridgway (1912), Pale Glaucous Green. The Pre-incubation and Incubation Periods Dr. William Beebe (unpublished notes) noted nest building in the Crested Oropendola to be under way in the Arima valley as early as Dec. 5 (1952) and Dec. 23 (1953) and as late as March 4 (1950). Belcher & Smooker (1937) reported that in Trinidad and Tobago nest build- ing began in December and eggs were found from mid-January onwards, with May 24 as the latest date. The nesting colonies near Simla were not dis- covered until Jan. 9, 1955, at colony B and Jan. 11 at colony A, at which time nest building was well advanced at both sites. There had been no activity at either tree in late December, 1954, and nest building therefore must have begun either in late December or early January and lasted about two weeks. Sixteen of the 23 nests started at colony A were being actively worked on during the obser- vation period on Jan. 12. On the following day, however, with 12 nests still incomplete, only two nests were visited by the birds during almost eight and one-half hours of observation. Al- though most of the nests were no longer actively visited after Jan. 12, there was considerable ac- tivity in the nest tree between Jan. 13 and Jan. 18. Periodically small groups made up mainly of females, many of which carried nesting mate- rial, would fly into the top of the tree. Here they would remain for varying periods of time, clucking noisily, but with only a few exceptions they made no further attempts at nest building. Of the 16 nests being built on Jan. 12, only eight were observed to be revisited for the pur- pose of nest building on subsequent days; one was visited once, two were visited twice, one was visited five times and one was visited six times, and at only three were visits continued to egg laying and incubation. These last three nests (1, 2 and 3) showed no similarity in the length of time between the end of nest-building and the start of incubation. This pre-incubation activity is summarized in Table 2. The earliest onset of incubation took place at nest 1 on Jan. 20. In all probability, this nest was completed prior to Jan. 12. Nest 2, how- ever, was not completed until Jan. 14, and incubation did not begin until Feb. 5. Nest 3 appeared to be complete on Jan. 12, with incu- bation commencing on Feb. 6. Just why colony A was seemingly abandoned is difficult to explain, for it appeared to be thriving. The observation post was well camou- flaged with canvas and concealed by vegetation, in addition to being situated some 100 feet from the nest tree. Moreover, these birds are not par- ticularly known for their shyness, and active colonies are to be found in well-populated areas. In fact, colony B flourished despite its proximity to a rock quarry where frequent blasting was taking place. Dr. Beebe has noted that in the previous three years, as well as the following year (1956), this site always contained medium- sized colonies. As can be seen by Table 2, nests 2 and 3 were visited frequently and fairly regularly before in- cubation. It appeared that incubation was under way at nest 3 on the afternoon of Jan. 20, as the female remained in her nest for periods ranging from 23 to 112 minutes during three and a half hours of observation. However, from Jan. 21 to 30 no further activity was noted at nest 3. With this one exception, the length of time spent in the nests at each visit was usually less than a minute, and exceeded five minutes on only three occasions. Oftentimes, nests 2 and 3 were visited in seemingly unaccountable spurts over relatively short periods of time. The most outstanding example of this behavior took place at nest 2 on the morning of Jan. 29 where, within a period of 44 minutes, the nest was visited briefly 22 times. The significance of this pre-incubation activity 90 Zoologica: New York Zoological Society [42: 8 Table 2. Pre-incubation Activity by the Female Parents at Colony A Date Total observation time (min.) Number of visits per hour Nest Nest Nest 1 2 3 Average time in nest (min.) Nest Nest Nest 1 2 3 Jan. 12 173 0.6 0.3 1.3 <1.0 <1.0 <1.0 13 501 0 0.1 0 0 <1.0 0 14 540 0 0.4 0 0 <1.0 0 15 575 0 0 0.1 0 0 <1.0 16 285 0 0 0 0 0 0 17 410 0.8 0 1.1 <1.0 0 <1.0 18 444 0.9 2.0 1.8 3.7 1.3 <1.0 19 240 0.7 2.7 0.8 <1.0 2.0 9.0 20 210 * 1.1 1.1 0.8 58.6 21 180 0 0.6 0 <1.0 22 350 4.1 0 3.1 0 23 245 1.7 0 1.3 0 24 370 0 0 0 0 25 215 2.0 0 2.3 0 26 226 0.5 0 <1.0 0 27 361 4.1 0.1 4.5 <1.0 28 131 3.2 0.4 6.0 <1.0 29 131 11.9 0.9 0.8 <1.0 30 190 2.0 1.2 6.7 4.7 31 352 1.7 0.8 4.4 <1.0 Feb. 1 170 0.3 0.3 20.0 <1.0 2 350 0.6 1.3 7.5 <1.0 3 80 0 0 0 0 4 83 0.7 2.4 <1.0 <1.0 5 341 0 0 ^ ^Onset of incubation. is not clear. Since nests 1, 2 and 3 were com- plete by Jan. 14, after which no nesting mate- rial was brought to the nests, and no nest weav- ing detected, these facts appear to eliminate the possibility that the hens involved were giv- ing “finishing touches” to their nests. Activity ranged from the apparent pseudo-incubation behavior at nest 3 on Jan. 20 to the extremely frequent nesting visits at nest 2 on Jan. 29. The beginning of incubation was determined by observing the increase in the duration of the attentive periods by the female at her nest. In colony A, the laying of the first egg at nest 1 occurred on Jan. 20, at nest 2 on Feb. 5 and at nest 3 on Feb. 7. The incubation period at all three nests lasted approximately 15 days. During this time the average time spent on the eggs was 74.6% at nest 1, 72.0% at nest 2 and 67.9% at nest 3. The frequency and duration of the attentive periods are indicated in Table 3 and Text-figures 1-3. The percentage of time spent in the nest was determined by dividing the average length of the attentive periods by the sum of the average attentive periods and multiplying by 100. The number of visits per hour at all nests ranged from 0.5 to 1.9, averag- ing 1.0 at nests 1 and 3 and 0.8 at nest 2. For at least seven days prior to egg laying, the female parents at nests 2 and 3 were seen to return to their empty nests at the onset of darkness, and presumably spent the night there. The times of these final entries, as well as those noted during incubation proper, varied from about 5:55 to 6:20 P.M. The Nestling Period With the sudden increase in the number of nest visits per hour, the hatching of the first egg was considered to have taken place. This oc- curred at nest 1 on Feb. 4, at nest 2 on Feb. 21 and at nest 3 on Feb. 22. As with incubation, all parental care was undertaken by the females. The length of nestling life varied from 31 to 32 days at nest 1 and 34 or 35 days at nest 2 to 36 days at nest 3. The amount of time that the hen spent in the nest with the newly-hatched young decreased gradually to less than about 20%. We can refer to this as the brooding 1957] Tashian: Nesting Behavior of Crested Oropendola 91 period. At nest 1 this brooding period lasted for seven days, at nest 2 for 11 days and nest 3 for nine days. After the last day of brooding the frequency of visits ranged from 3.2 to 8.5 per hour, averaging 5.7, 5.6 and 5.9 at nests 1, 2 and 3 respectively. Time spent at the nest was usually less than 5% at nests 2 and 3 during the post-brooding period, the length of the aver- age visit being but a few seconds. At nest 1, however, it varied from 5% to 15%, a factor which might possibly have been due to the presence of more than one nestling. After the cessation of brooding, the females stayed with their nestlings all night for a varying number of days. At nest 1, this period extended for 11 days after hatching, or 34% of the nes- tling stage; at nest 2 for 26 days (74%), and at nest 3 for 23 days (63%). The exact nature of the food brought to the netslings could not be determined; however, large orthopterous insects or lepidopterous larvae were frequently noted. Excrement packets were periodically removed from the nests. By the 18th or 19th day of feeding, the whining of the nestlings, as well as their move- ments within the nests, could be detected. The young birds are either able to see through the nest wall, or can hear the wing sound of the ap- proaching parent, as often this whining began even before the hen had reached the nest. It usually continued on through the feeding and for a few minutes after she had left. The male can also bring about this whining, for on one occasion a male flying to a branch near an occu- pied nest caused the young to whine. This, in- cidentally, induced the male to explore the nest curiously, but he did not enter. During the early days of feeding the hen en- tered the nest entirely. But as the nest opening became enlarged by her frequent passage and the nestling increased in size, she fed it either while perching at the nest entrance (Plate I, Fig. 2) , or while clinging to the side of the nest. In the last days of nestling life, the young bird at nest 3 was sometimes seen to climb entirely out of the nest. Then, while clinging to the side and flapping its wings, it would at times com- pletely circle the nest before re-entering. Nest-leaving took place at nest 1 sometime between 11:30 A.M. on March 7 and 3:30 P.M. on March 8 (31 or 32 days after hatching), at nest 2 between 3:49 P.M. on March 27 and 8:09 A.M. on March 28 (34 or 35 days after hatching) and at nest 3 at 9:07 A.M. on March 29 (36 days after hatching). The following notes were taken on the morning of nest-leaving at nest 3: 7:49 A.M. Observer entered blind. 7:50-8:53 Nestling fed on 5 separate occa- sions. 9:02 Nestling had head out of the nest opening before hen reached nest; after being fed it climbed com- pletely out of the nest, perching at the entrance and peering about for a few minutes before re-entering. 9 : 07 After being fed, the nestling perched at the nest entrance for a few sec- onds, then exercised its wings mo- mentarily and flew off in a fairly strong, slightly descending flight, alighting in a tree some 75 feet away. 9:08 Fledgling whined and vibrated wings while being fed. Table 3. Per cent, of Time Spent in Their Nests BY THE Female Parents at Colony A during the Incubation and Brooding Periods Number of days after egg laying Nest 1 Jan. 20* Nest 2 Feb. 5* Nest 3 Feb. 7* Incubation 1 81.8 81.0 2 98.8 — 71.8 3 41.6 81.3 4 81.6 79.8 73.0 5 78.1 — 58.5 6 59.4 63.9 66.1 7 60.7 75.1 8 68.1 84.5 75.4 10 60.9 — 73.9 1 1 60.9 — 71.2 12 74.0 69.5 65.1 13 — 72.9 85.0 14 42.6 61.5 _ 15 — 83.2 72.0 Brooding 16 _ 56.0 17 31.9 27.1 59.0 18 — 52.3 23.0 19 22.8 38.3 34.1 20 21.5 44.3 34.5 21 — 25.0 22 31.6t — 24.7 23 42.1 28.4 24 — 23.7t 25 32.4 26 22.lt *Probable dates of egg laying. tLast day brooding observed. Percentages not includ- ed if observation period was less than 3 hours. 92 Zoologica: New York Zoological Society [42: 8 Text-fig. 1. Frequency of nest visits and duration of attentiveness by female parent at nest 1 in colony A. 9:09 Fledgling flew to another perch about 25 feet away where it re- mained perching quietly. 9:25 Fledgling fed by hen. 9:30 Fledgling flew to another tree about 50 feet distant, again remaining quiet after landing. 9:43 Fledgling flew off again; no longer observable. No more than one nestling was seen at any of the nests at colony A. Vocalization and Display At first hearing, any description of the various calls of the male Crested Oropendola appears hopeless, as there does not seem to be any par- ticular pattern to his polyphonic and diverse utterances. However, after prolonged and care- ful listening, they were recognized as falling into four groups: PER CENT. TIME IN NEST 1957] Tashian: Nesting Behavior of Crested Oropendola 93 7 § O 3: Uj 3 ^ Text-fig. 3. Frequency of nest visits and duration of attentiveness by female parent at nest 3 in colony A. Call 1. An introductory whaaa cheeee with the second phrase higher and fol- lowed by a continuous, unmusical trUI; whaaa Call 2. Introduction a rising unmusical trill (rrrrrrrrrup) followed by cheow wow, eow woo woo or cheeeeeow, not unlike the meow of a cat: cheow wow Call 3. An introductory cheeeee (falling in pitch), or a rising brrrrrree, cheeeeeow or rrrrrrrrrup, usually diminishing in intensity, followed by lob lob baah or ko ko plo, repeated three or more commonly four times; the first two notes are given rapidly and the last is drawn out: ko ho plo ko ko pio ko ko plo Call 4. An introductory peeeeeow or cheeeeee (falling) or rrrrrrrup (ris- ing), diminishing in intensity, fol- lowed by a series of ko, plo or kok notes usually repeated five times on the same level: ko ko ko ko ko The most common was Call 1 which com- prised 45% of all of those heard. Call 2 was given 24% of the time, Call 3, 16%, and Call 4, 15%. On numerous occasions the male was heard to utter a confusion of stuttering, gurgling notes, containing some of the elements of a typical display call but wholly lacking in pattern: oooorrrrrup whup whaaaat upka brrrrrr woo woe whup wheel brrrrreee woo, etc. This was accompanied by much gurgling, feather rustling and wing flapping. It was almost as though a young male were attempting to imitate the call of an adult and failing badly. Every once in a while a male would rather half-heartedly begin a display, but never really finish. Both the abor- tive gurgling and the unfinished calls were heard during all nesting periods. Calls 1, 2 and 3 usually accompanied a dis- play, whereas Call 4 was apparently given in the absence of any display. The male was seen to go through his complete display not only in the presence of females, but also when alone or in the company of males. Dr. Beebe observed two males simultaneously going through a complete display while facing each other. They were some distance from the nest 94 Zoologica: New York Zoological Society [42: 8 tree, and no females were noted in the vicinity. In the typical display, the bird begins by bending forward in a deep bow until his head is well below the perch. At the same time he ruffles his body feathers and brings the extended wings together over the back. The wings are then vibrated rapidly with the primaries open. Almost immediately after the wings are ex- tended, the tail is brought forward quickly over the back until it makes an approximate 90° angle with the back (Plate II, Figs. 1 & 2), and remains in this position from 1.5 to 1.9 seconds. Then, as the tail begins to lower to its normal position, the wing beats become less rapid and take on a flapping or fanning quality which lasts from 1.1 to 2.6 seconds. Sometimes, however, the wing beat does not slacken until the tail has been completely lowered. Finally, the wing flap- ping stops and the bird comes out of his bow, the entire performance lasting about four to five seconds. These times are based on the analysis of moving pictures taken of five com- plete displays. The average duration of the four main phases of the display are as follows: 1. Tail elevated forward over back. 0.20 seconds 2. Wings brought together over back and vibrated. 1.74 seconds 3. Tail lowered to normal position. 0.36 seconds 4. Wing fanning or flapping 1.97 seconds The intervals between beats in the wing-fan- ning phase were measured at 0.03 seconds. Al- though not actually measured, many of the dis- play calls seemed to last around five to seven seconds. In all probability, the figures above would have averaged higher if more displays had been analyzed. Although the spreading of the tail feathers was not noted in the present study. Young (1929) reported it in Psarocolius decumanus from British Guinea.. This tail spreading during the display is probably characteristic of all oropendolas, as Fuertes (1916) observed it for the Montezuma Oropendola and Chapman (1928) for the Wagler Oropendola. Displays were frequently performed on the nests proper. In this type of display, the male flew directly to the nest and, clinging to its side, began the display almost immediately upon alighting. The head was always directed down- ward. In such a position only the initial bowing of the perching display was, of necessity, omit- ted. Twenty-eight per cent, of all displays at the nest tree were of this type. Crandall (1914) also described this type of nest display for the Monte- zuma Oropendola in Costa Rica. During the 303 hours of observation at colony A, 683 displays were counted either at the nest tree or from the immediately surrounding trees. The displays from the nest tree numbered 320, and those from the nearby trees, 363. The fre- quency of displays given on or near the nest tree ranged from 0.1 to 10.2 per hour, with an average of 2.2 per hour throughout all nesting periods. At no time of the day was the display observed to be given more frequently. Often a male would fly to a nest and perform no display. On the other hand, there were in- stances where a male displayed on as many as five different nests in succession. There seemed to be no particular preference in selecting the nests to display on, and oftentimes an occupied nest was flown to while a hen was incubating, whereupon she would usually vigorously chase the male just as she would a cowbird. Sometimes when a group of males were in the nest tree and one of them flew to a nest, another male would chase it off and, as though stimulated by this chase, then go through a complete nest display. On one occasion this type of behavior was re- peated three times in the course of about three minutes by the same pair in the same respective roles. It was a fairly common practice for males accompanying returning females to the nesting colony to display almost as soon as they arrived in the nest tree. The most common note of the female was a cluck which she sometimes uttered repeatedly for considerable periods of time while perch- ing, preening or flying about the nest tree. Often, when she first came to the nest tree after an inattentive period, during either incubation or feeding, she clucked continuously sometimes up to 15 minutes before finally entering her nest. Less frequently she carried on this clucking after leaving the nest. When annoyed or disturbed, the clucking became more rapid. Although not noted for certain, the male probably also gives a similar cluck note. Skutch (1953) reports this in the male Montezuma Oropendola. In chasing cowbirds, males or other females from her nest or its vicinity, the hen generally voiced a harsh, nasal mewing rising at times to a high-pitched whine. As is characteristic of aU oropendolas and caciques, a warning cack is given by the male at the approach of danger. This seemed to be most often brought on either by the sight or scream of the Gray Hawk {Buteo nitidis), or the sudden appearance of man. Frequently after the sounding of an alarm cack, the oropendolas present failed to leave the nest tree or its imme- diate area, and incubating hens rarely left their nests. Cowbird Activity The visitations of Giant Cowbirds (Psomo- 1957] Tashian: Nesting Behavior of Crested Oropendola 95 colax oryzivorus) to colony A were carried on with persistant regularity between Jan. 13 and March 4. On the latter date the nestlings were 26 to 43 days old. After March 4 only one further cowbird visit was noted, on March 23. On 34 of the 50 days between these dates, cow- birds were seen coming to the nest tree at one time or other on 52 separate visits. Cowbird activity was recorded on 77% or 29 of the 37 days in which the observation periods lasted longer than three hours. It was interesting to note that on 34% of their visits, the cowbirds came to the nest tree follow- ing returning female oropendolas. In the ma- jority of cases only one cowbird visited the colony at any one time. However, on five occa- sions two cowbirds came to the nest tree to- gether, and once a group of four came in together. Usually the cowbirds immediately began in- specting a series of nests. At colony A most of the nests were unoccupied. The process of the inspecting consisted in momentarily peering into the nest opening, and only on two occa- sions were cowbirds seen to enter the nests com- pletely. If the nest happened to be occupied by an incubating female, she would always give her scolding whine and usually give immediate chase. The cowbird was chased for only a short distance, after which the female oropendola would usually immediately return to her nest. Often the same cowbird would return immedi- ately, only to be chased again from the same or other nests. Only three instances were noted when an incubating or brooding female did not leave her nest to chase a cowbird, but instead merely voiced a scolding whine from within. On several occasions cowbirds were chased from unoccupied nests by females who happened to be either returning or leaving their nests. These chases were not especially energetic, however, and pursuit was not vigorous. Male oropendolas were rarely seen to chase cowbirds, and when both cowbirds and males were in the nest tree at the same time, neither seemed very disturbed by the other’s presence even when perched only a few feet apart. If cowbirds remained quietly in the nest tree they were never observed being chased even by females. Cowbirds were never noted to remain for very long in the nest tree during inspecting visits. The average length of time was only two to three minutes. From all indications, cowbirds appeared to have had little success in parasitizing any nests at colony A. Only two nests were ever observed to be entirely entered, and these for but a few seconds. The Nesting at Colony B The various nesting phases at both colonies were relatively comparable. On the first day of observation at colony B (Feb. 6) nestlings were present in three nests and incubation was under way in five of the 1 1 nests eventually oc- cupied. On the same date at colony A, one nest contained young and two were incubating. As shown in Table 4, three of the 14 original nests were never occupied, and of the remaining 11, three fell, leaving eight (73%) of the occupied nests to produce fledged young. Nest number 3, which the nestlings had left by March 6, was seen to be revisited on March 13. During observation periods in the following three weeks a female was seen to visit it, but the nature of the visits could not be determined with any certainty. This might possibly be an instance of the rearing of a second brood. In 1956 Dr. Beebe noted nest activity by females at colony A as late as June 4, again indicating the possibility of a second brood The general activity at colony B was much the same as that at colony A, but naturally on a larger scale. The times of the various nesting phases were similar, with incubation lasting about two weeks and feeding from about four to five weeks. Miscellaneous Observations Most of the time the females went about their various activities independently. However, it was not uncommon to see two hens leave or return to their respective nests at the same time. This was observed not only when they left the nests for the first time in the morning and at the final entry at the coming of darkness, but also during regular daytime activity. For ex- Table 4. Summary of the Nesting Phases for the Occupied Nests at Colony B. (I=incubat- ing; N=nestling period; F=nestlings fledged; X=nest fell; R=nest reoccupied). Nest February March 6 13 20 27 6 13 20 27 1 I X 2 N N N X 3 N N N N F R? 4 N N N N F 5 I I I N N X 6 I I N N N F 7 I I I N N N N F 8 — I I N N N N F 9 — I I N N N N F 10 I I I N N N N N 11 I I I N N N N N 96 Zoologica: New York Zoological Society [42: 8 ample, on March 4 the three females at colony A, who were feeding their nestlings, either en- tered or left their nests 218 times during six and a half hours of observation., On 23 (10%) of these nest entries or departures, two females were seen either to leave or arrive at their nests together. Copulation was seen on but one occasion at the nest tree; this took place in the late after- noon (6:00 P.M.) on Feb. 2. The male ap- proached one of a pair of females and displayed, after which he began to peck at her cloacal area. The female fluttered her wings but otherwise remained fairly still. The male then engaged in a brief copulatory act and followed this by peck- ing at her cloacal region once more, while she squatted and fluttered her wings again slightly. The male then went through another complete display. Other than the cowbirds, all species of birds that visited the nest tree at colony A remained unmolested. Once a Gray Hawk {Buteo nitidis) flew into the nest tree and perched undisturbed for a few minutes before flying off. It was ap- parently unseen by the oropendolas of the colony, as the recognized presence of a hawk in the vicinity would invariably elicit the sounding of the alarm cack. Once at colony B, when a Gray Hawk flew overhead, several oropendolas took off in pursuit. Discussion and Summary The nesting activity of two colonies of the Crested Oropendola (Psarocolius decumanus insularis) in the Northern Range of Trinidad, B. W.I., was studied from the end of nest build- ing to nest leaving by the young. Fourteen nests were completed at one colony all of which were later occupied, and 11 were completed at the other, of which three were occupied. The rela- tive length of the various nesting phases as well as the general activity were similar at both colonies. Nesting success at the colony with the three occupied nests was 100%, and at the colony with the 11 occupied nests, 73%. Nest building and care of the young were under- taken entirely by the females., The following discussion and summary are based on the obser- vations at the colony containing the three occu- pied nests. The period from the completion of nest build- ing to the beginning of incubation varied from 7 to 25 days with the incubation period proper lasting 15 or 16 days. This is but slightly longer than the incubation of 11 to 14 days reported in the literature for the family Icteridae. The average time spent on the eggs ranged from 68% to 75%. The nestling period varied from 31 to 36 days, and brooding from 7 to 1 1 days. These nestling periods do not vary significantly from those of 30 to 37 days given by Chapman (1928) and Skutch (1954) for two other oropendolas: Gymnostinops montezuma and Zarhynchus wagleri. After the cessation of brooding the females remained with the nestlings at night from 11 to 26 days, or 34% to 74% of the nestling period. The different phases of the male perch dis- play were analyzed with the aid of motion pic- tures. Comparisons with the descriptions in the literature of the display of other oropendolas {Psarocolius decumanus, Chapman, 1894, Young, 1929; Gymnostinops montezuma, Fuertes, 1916, Skutch, 1954; and Zarhynchus wagleri, Chapman, 1928) emphasize the marked similarity of the sequence of display phases among oropendolas. This similarity is brought out in Plate II, which compares the perch dis- plays of the Crested and Green Oropendola. Dis- plays were frequently performed on the nests proper, and 28% of all displays at the nest tree were of this type. The frequency of all dis- plays given on or near the nest tree averaged 2.2 per hour throughout all nesting periods. The various calls of the male were analyzed and found to fall into four separable groups. These calls, although of a similar quality to those reported for other oropendolas (op. cit.), appear to differ considerably, not only within the oropendola complex, but also within the same genus. Beebe (unpublished notes) records Psarocolius decumanus from Trinidad as calling wholly unlike Psarocolius angustifrons from Venezuela; and Friedman & Smith’s (1950) de- scription of the call of Psarocolius decumanus from northern Venezuela does not seem to re- semble closely any of the calls of the Trinidad birds. Persistant cowbird activity was noted from the termination of nest building on to advanced feeding, at which time the nestlings were 26 to 43 days old. During this period cowbirds were seen to visit the colony to inspect nests on about 77% of the observation days. This type of per- sistance by Giant Cowbirds was also noted by Skutch (1954) in a colony of Montezuma Oro- pendolas. Literature Cited Beebe, W. 1952. Introduction to the ecology of the Arima valley, Trinidad, B.W.I. Zoologica, 37: 157-184. Belcher, C. & G. D. Smooker 1937. Birds of the colony of Trinidad and To- bago. Ibis, pp. 226-550. I .1 I I ■ji ^ i; 1 1957] Tashian: Nesting Behavior of Crested Oropendola 97 Chapman, F. M. 1894. On the birds of the island of Trinidad. Bull. Am. Mus. Nat. Hist., 6: 1-86. 1928. The nesting habits of Wagler’s Oropen- dola (Zarhynchus wagleri) on Barro Colorado Is. Bull. Am. Mus. Nat. Hist., 58: 123-166. Crandall, L. S. 1914. Notes on Costa Rican birds. Zoologica, 1: 325-343. Friedmann, H. & F. D. Smith, Jr. 1950. A contribution to the ornithology of north- eastern Venezuela. Proc. U. S. Nat. Mus., 100: 411-538. Fuertes, L. a. 1916. Impressions of the voices of tropical birds. Smithsonian Report for 1915 (Pub. 2392), pp. 299-323. Goeldi, E. 1897. On the nesting of Cassicus persicus, Cassi- cus oryzivora, Gymnomystax melanicterus and Todirostrnm maculatus. Ibis, pp. 361- 370. 1903. On the nesting of O. decumanus near Para, and the materials of its nest. Ibis, pp. 630-632. Ridgway, R. 1912. Color standards and color nomenclature. Publ. by the author. Washington, D. C. Skutch, a. F. 1954. Life histories of Central American birds, families Fringilidae, Thraupidae, Icteri- dae, Parulldae and Coerebidae. Pacific Coast Avifauna No. 31, pp. 287-304. Young, C. B. 1929. A contribution to the ornithology of the coastland of British Guiana. Part III. Ibis, pp. 221-261. 98 Zoologica: New York Zoological Society [42: 8: 1957] EXPLANATION OF THE PLATES Plate I Fig. 1. View of colony A from observation blind (February). Fig. 2. Female parent feeding nestling in nest 3 at colony A (March). Plate II Figs. 1 & 2. Male Crested Oropendolas (Psarocolius decumamts) displaying in nest tree at colony A. Fig. 3. Beginning of display in the male Green Oropendola (Psarocolius virens) . New York Zoological Park. Fig. 4. End of display in the male Green Oro- pendola. Note bow is deeper than in P. decumanus. TASHIAN PLATE 1 NESTING BEHAVIOR OF THE CRESTED OROPENDOLA (PSAROCOLIUS DECUMANUS) IN NORTHERN TRINIDAD, B.W.I TASHIAN PLATE II FIG. 1 FIG. 3 FIG. 4 NESTING BEHAVIOR OF THE CRESTED OROPENDOLA IPSAROCOLIUS DECUMANUS) IN NORTHERN TRINIDAD, B.W.I. 9 Changes in the Cytological Structure of the Adenohypophysis and Gonads in Juvenile Bathygobius soporator after Pituitary Implantation^ Louise M. Stoll American Museum of Natural History, Department of Fishes and Aquatic Biology (Plates I & II) Introduction The relationship of the pituitary and the gonads has been studied and acknowl- edged in all vertebrate classes. General seasonal changes in the pituitary and gonads of the goldfish and carp were described by Scruggs (1951) and correlated with the previous liter- ature on the subject in teleosts. No report has been found on the histology of the the specific cell types of the pituitary correlated with the histology of the gonads after experimentally in- duced changes in the two tissues. The present report is a histological study of gonads and trans- itional lobes of pituitaries of juvenile fish after pituitary implantation. All fish used were collected in shallow water areas around North Bimini Island, Bahamas, B. W. I. Experimental work was done at the Lerner Marine Laboratory on Bimini. The help given during this study by Dr. C. M. Breder, Jr., and Miss Priscilla Rasquin is grate- fully acknowledged. Materials and Methods The marine teleost Bathygobius soporator (Cuvier & Valenciennes) was used. This species is very hardy and will survive the handling in- volved during implantation. Since these fish do not engage in active swimming to a large extent, the incision healed rapidly and no stitches were necessary. Twenty-five young Bathygobius soporator of an average standard length of 38 mm. were given intraperitoneal implants of whole pituitary ^This study was supported in part by a grant from The National Science Foundation. glands from larger fish of the same species. The average standard length of the donors was 66 mm. All 25 fish received three pituitaries each. Pituitaries were obtained by the following pro- cedure. The donor fish was heavily anesthetized in ethyl urethane dissolved in sea water. The lower jaw was removed and the floor of the cranium with its covering membranes was cut away with small bone clippers, exposing the pituitary. The area for cutting was determined by the position of the saccus vasculosus, which could be seen through the bone. Pituitaries were removed by grasping the infundibulum with watch-maker forceps and pulling gently. The pituitaries were held in physiological saline (0.6%) until all three were collected. The recipient fish was lightly anesthetized with ethyl urethane and placed on cotton sat- urated in sea water. A slit was made in the ab- dominal wall lateral to the midline with watch- maker forceps and the three pituitaries were introduced through the slit. A dissecting micro- scope was used white removing and implanting the glands. Immediately after implantation, the fish were placed in individual two-gallon tanks with running sea water. The gonads of the donor fish were removed and fixed in Bouin’s picro-formol solution, em- bedded in paraffin, sectioned at 10^, and stained with Harris’s haematoxylin and eosin. The fish receiving the implants were killed after varying lengths of time. Eighteen fish were killed at seven to eight days, and seven fish from two and one-half to three days post-operative. Their bodies, minus the heads, were fixed in Bouin’s, sectioned at lO/i, through the region of the gonads and stained with Harris’s haematoxylin and eos- in. Heads of the 25 experimental fish were fixed. 99 100 Zoologica: New York Zoological Society [42; 9 embedded in paraffin and sectioned at 5/u.. The modification of Halmi’s paraldehyde fuchsin stain by Sokol (1953) provided an excellent stain with which to observe the response of any specific cell group of the pituitary to a given experimental procedure. Bouin’s fixative and the paraldehyde fuchsin stain were used for the lieads of 18 of the experimental fish, and Baker’s fixative and acid haematin stain were used for the seven other experimental fish. The paralde- hyde fuchsin was used specifically for the study of basophils and the acid haematin for acidophils. A group of 11 normal fish were killed and prepared for histological study by the same methods used for experimental fish. The normal fish ranged from 80 mm. to 29 mm. in standard length. All stages of normal gonad development and maturation were represented in this series. Tavolga (1955) reported that B. soporator collected in the area of Marineland, Florida, were at the height of the spawning season in July and August., However, the largest gobies which were collected for the present study in Bimini waters were judged by gross dissection and histological section to be in or near spawn- ing condition in March and April. Testes were swollen, nearly white in color and had high epithelium; ovaries were bright yellow and thin walled as described by Vivien (1941) as charac- teristic of the spawning condition for the European goby, Gobius paganellus. The largest female collected by Tavolga (1954) was 65 mm. in standard length, while males of 90 mm. or more were found. In Bimini, the largest female found was 87 mm. and the largest male 108 mm. The mean, of course, was lower, the average of 43 females being 66 mm. and of 56 males, 74 mm. Testes of eight normal Bathy gobius were stained with Sudan Black B and by the Baker technique for phospholipids. The testes of one fish which had received three pituitary implants and maintained for one week was also stained with Sudan Black B. Results Implants of pituitaries of mature fish to juve- nile fish caused degranulation of the basophils of the adenohypophysis and stimulation of the gonads. The acidophils of the transitional lobe were not affected. In both experimental and con- trol fish, the acidophils were large, well-rounded and heavily stained with the Orange G of the paraldehyde fuchsin stain. The acidophils stained with acid haematin did not show degranu- lation. Degranulation of the basophils occurred in nine fish which had received implants and were maintained for one week before being killed. Little or no degranulation was seen in seven fish sacrificed three days after receiving pituitary im- plants, nor in two fish maintained for one week before killing. The latter had gonads in very early stages of maturation and it is thought that the tissue may be refractory at this early stage. Partially degranulated basophils of a three-day implant are shown in Plate I, Figure 1, and may be contrasted with the advanced degranulated condition found in a seven-day implant shown in Figure 2. Normal acidophils are shown in Figure 3. The degranulated basophils are found along the periphery of the pituitary and in groups along the region of juncture with the inter- mediate lobe., The centers of the cells are clear and the cells somewhat shrunken and collapsed. Dark-staining granular material may be seen between the cells. The basophils of the control and non-reactive experimental fish are large and have either very small clear areas or none. The basophils take a very heavy and dark stain. Atz (1953) described basophils similarly lo- cated in the fresh water teleost, Astyanax mex- icanus. The correlation of changes in the peri- pheral basophils with changes in the gonads led her to consider these cells as gonadotrophs. In all but one case basophils were reported to have increased in number near the time of onset of the spawning season. In no case was there a re- port of degranulation occurring at this time. In the present experiment no increase in the num- ber of basophils was noted., The possibility of basophilic degranulation occurring as a result of stress and loss of ACTH was considered. However, two fish which sur- vived seven days showed no degranulation and the seven fish which were killed after three days, an even shorter time to adjust to stress, showed little or no degranulation. The testes of Bathygobius soporator have an unusual amount of tissue which is not sperma- togenic. The functional germ cell tissue of the testis is spirally wrapped around a core of tissue of very different nature. The core tissue has some resemblance to interstitial tissue but the amount seems out of proportion to the need or to the amount of interstitial tissue found in other teleosts. The intimate relation of the germ cell tissue and the core tissue would indicate that the core is a functional part of the testis but no defi- nite function has been described. Gonads, especially the ovaries, were stimu- lated in the fish showing degranulation. The stimulation caused rapid maturation of about half the ova and a third of the spermatogonia but little increase in over-all size of the gonad. 1957] Stoll: Changes in Bathygobiiis separator after Pituitary Implantation 101 However, the core tissue of the testes hypertro- phied. Stimulation was greater when the pituitary donor had gonads in intermediate stages of maturation than when the gonads were very ad- vanced and approaching or in spawning condi- tion. An ovary from an untreated juvenile and an untreated mature fish are shown in Plate I, Figures 4 and 5. Figure 6 shows the response of an ovary to the experimental procedure. The ovary from the mature untreated fish shows more uniform graded development than that of the treated juvenile fish where several ova are highly developed while the rest are in a uniform early stage. Testicular and core tissue of an untreated juvenile and untreated mature fish are shown in Plate II, Figures 1 and 2. The normal condi- tion of the core tissue of a juvenile fish stained with Sudan Black B to show lipids is shown in Plate II, Figure 3, and is contrasted with Figure 4 of the same magnification which shows the hypertrophied core tissue from a fish which had received pituitary implants. Baker’s acid hae- matin and pyridine extraction techniques were used on the core tissue to indicate some of the cellular components. Staining results of the core tissue of normal fish showed a positive reaction to acid haematin (Plate II, Figure 5) and a neg- ative one to pyridine extraction (Plate II, Figure 6), indicating the presence of phospho- lipids. Discussion The histology of the pituitary and the identi- fication of cell types as sites of specific hormone production is still to some extent controversial. In the teleosts the transitional lobe has been found to be homologous to the anterior lobe in other vertebrates (Charipper, 1937). The pituitary basophils of the rat were de- scribed as the site of thyrotropic and gonado- tropic hormone elaboration and storage by Purves & Griesbach (1951). A subsequent re- port (1954) designated the specific basophils which produce TSH, FSH and LH. However, there is some disagreement as to whether the gonadotropic hormone is a single biochemical entity or two or even more. Evans & Simpson (1950), in a review of the literature concerning the gonadotropins, called attention to the fact that it is not certain that there are two distinct gonadotropic hormones identical with the two substances which have been isolated from pituitary tissue or that both hormones are necessary in both sexes for normal reproductive function. The gonad maturity at the time of pit- uitary injection for experimental purposes is an important factor. Matthews (1939), using injections of mam- malian pituitary extract on Fundulus hetero- clitus, found that the injections had no stimulat- ing effect in the males and only four of 35 fe- males responded to any degree. Hypophysectomy caused regression of the gonads especially in the males. On the basis of this work the pituitary gland of a teleost was seen to exert a controlling influence on the seasonal cycle and this influence was considered to be . . of greater importance in maturation than in proliferation of the germ cells.” The present study is in agreement, as pituitary implantation caused an acceleration in the rate of maturation of the germ cells. The effect of adult Fundulus pituitary im- plants to immature fish of the same species was reported by Matthews in 1940. Adult pituitary implanted intraperitoneally into immature fish at three-day intervals caused gonad stimulation by the end of four weeks. The males were espe- cially responsive and large numbers of mature spermatozoa were seen. The pigmentation of the fish was characteristic of the adult in breed- ing season. Included in this report was a review of the use of mammalian pituitary extracts for injection in fish and the conflicting results and data collected; in many cases the mammalian preparation had no effect. A similar review of the use of injecting fish pituitary preparations in fish showed a fairly uniform response. All species tried showed enlargement of the gonads and in some cases expulsion of eggs and sperm. All experiments involving hypophysectomy showed gonad regression. Burger (1941) used pituitaries from adult Fundulus as implants to other adult male Fundu- his which had been hypophysectomized at the time of maximal testicular development and which showed inhibition of the testes after hypo- physectomy. The implants caused recrudescence of the testes within two weeks. The implants were made intraperitoneally and each fish re- ceived five at a time for four times. Burger con- cluded that the Fundulus pituitary contained gonadotropic material which was responsible for spermatogonial proliferation and for the ma- turation phenomena. Riley & Fraps (1942) investigated the gonad- stimulating activity of anterior pituitary in the female domestic fowl. Glands from hens with regressed or quiescent ovaries produced a greater stimulation of the gonads in immature mice than glands from hens in full reproductive condition. Greater gonad stimulation of the gobies in the present report resulted when pituitary donors were in intermediate stages of gonadal develop- ment. The cyclic changes in the pituitary of the uro- dele amphibian (Taricha torosa) were described by Miller & Robbins (1955). In this form the 102 Zoologica: New York Zoological Society [42: 9 delta basophils of the pituitary increase in num- ber and granulation in relation to spring sperma- togenesis and oogenesis and late fall final gonad maturation. The beta basophils were considered to be related to increased thyroid activity. As part of a study of light and temperature effect on the sexual cycle of the bitterling Rho- dens amarus, Verhoeven & van Oordt (1955) studied the adenohypophysis of the fishes which had an experimentally induced sexual cycle. They briefly note that the beta cells of the gona- dotropic zone had become more numerous and staining response of these cells to PAS was a deep purple coloration which in rats indicates gonadotropic hormone. The method by which an increase in the number of beta cells was ascer- tained was not given. Rasquin & Stoll (1955) described the asso- ciation of degranulated centrally located baso- phils with hypertrophied adrenals in the fresh- water teleost Astyanax mexicanus after injec- tions of pitressin. The degranulation was thought to be a result of loss of ACTH, as the thyroid tissue was not stimulated. The peripheral baso- phils (suspected of gonadotropin elaboration) did not degranulate. The only possible explanation for the appear- ance of degranulated basophils after pituitary implantation in the present study is that the basophils are not elaborating secretion granules at the normal rate because of the excess pituitary substance which was added., The absence of gran- ules in cells usually crowded with them is gener- ally interpreted as a release of hormone, but since the fish were given more pituitary there would appear to be no need for the pituitary of the implanted animal to add its secretion to a system already over-supplied. The fact that the degranulation was so gradual— little or none could be detected after three days of implanta- tion-supports the idea that degranulation in this case is a result of an inhibition of elaboration rather than a stimulated release of secretory products which would presumably occur more rapidly. Sections of the core tissue of testes of Gobius auratus Risso were stained and found to be faint- ly fuchsinophilic and unblackened by iron haematoxylin (Eggert, 1931). The granules within the cells of the tissue apparently con- tained lipids. Included in the report is a survey of the literature on the core tissue and a discus- sion of the possibility that this is the interstitial tissue. Eggert claimed that the fine vacuolation and granulation indicated that the cells were not interrenal. He did not find core tissue in the testes of G. panizzae Verga, G. buchichi Stein- dachner, G. quagga Heck or G. jozo Linnaeus. Coujard (1941) described the core tissue in Gobius niger and G. minutus as being made up of cells with large nuclei and nucleoli, abundant in protein and containing lipids. He did not attribute a definite function to the tissue but considered that it undoubtedly had a function correlated with glandular development and the nervous mechanism of spawning. The procedure for demonstrating phospho- lipids has been described in detail by Baker (1946) . He used the two-stage technique of acid haematin staining and pyridine extraction on sev- eral tissues, including the testes of the mouse. The secretion droplets of the interstitial cells reacted positively to acid haematin and nega- tively to pyridine extraction, which identified the secretory granules as phospholipid. j Melampy & Cavozos (1954) made a compara- i tive study of lipids in the vertebrate testis, using ; Sudan dyes and Baker’s acid haematin-pyridine extraction technique. Using the Sudan dyes, a positive reaction was obtained from the inter- stitial cell cytoplasm, basement membrane, cyto- plasm of Sertoli cells, spermatogonia, spermato- cytes and spermatids of the bull, ram, boar, guinea pig, rooster and horned lizard. The tele- ost, the bluegill (Lepomis macrochirus), showed a negative reaction with the Sudan dyes. The authors suggest that the failure to get a positive reaction from bluegill testis might be due to the cyclic breeding season of this form and that at the stage of the test the reactive quantity of cholesterol in the testis was absent. Baker’s test was not applied to the bluegill testis. The implantation experiments reported here show that the core tissue of the testis is respon- sive to pituitary stimulation. The marked hyper- trophy of the cells, the high phospholipid content and the close proximity of the core tissue to the spermatogenic cells seem sufficient evidence for designating the core cells as interstitial tissue. Summary 1. Pituitary implants from adult fish to juvenile fish caused degranulation of the peripheral baso- phils of the transitional lobe of the hypophysis and gonad stimulation in receiver fish maintained for one week after implantation. 2. Basophils of fish maintained for three days after implantation showed little or no degranu- lation. 3. Acidophils of the adenohypophysis were not affected by the experimental procedure. 4. Degranulation which occurred gradually may be the result of inhibition of hormone elaboration rather than stimulation of release of hormone. 5. Basophilic degranulation, when it occurred, 1957] StoU: Changes in Bathygobius soporator after Pituitary Implantation 103 was the same in both sexes but acceleration of maturation of the ovary was greater than that in the testis. 6. The core tissue of the testes hypertrophied after pituitary implantation and had a high phos- pholipid content. On this evidence, the core tissue is designated as interstitial tissue. 7. The degree of gonad maturity of the pitui- tary donors was an important factor in the re- sults. Greater stimulation resulted when donors were in intermediate maturation stages than when donors were in advanced stages. Bibliography Atz, E. H. 1953. Experimental differentiation of basophil cell types in the transitional lobe of the pituitary of the teleost ft^h, A sty anax mexi- canus. Bull. Bingham Oceanogr. Coll., 14:94-116. Baker, J. R. 1946. The histochemical recognition of lipine. Quart. J. Microscop. Sci., 87:441-470. Burger, J. W. 1941. Some experiments on the effects of hy- pophysectomy and pituitary implantations on the male Fundiilus heteroclitus. Biol. Bull., 80:31-36. 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 of Incarassius (sic) auratiis (the goldfish) and Nectiirus maculosus (the mudpuppy). Cold Spring Harbor Symposia on Quantitative Biology, vol. 5, Cold Spring Harbor, New York, pp. 151- 164. COUJARD, R. 1941. Sur I’existence d’une glande testiculaire et d’une glande genitale annexe chez les gobies. C. R. Soc. Biol., 135:570-574. Eggert, B. 1931. Die Geschlechtsorgane der Gobiiformes und Bleniiformes. Zeit. f. wissen. Zool., 139:249-558. Evans, H. M., & M. E. Simpson 1950. Physiology of the gonadotrophins. The Hormones, Vol. II. Ed. Pincus and Thi- man, pp. 351-404. Matthews, S. A. 1939. The relationship between the pituitary gland and the gonads in Fundulus. Biol. Bull., 76:241-250. 1940. The effects of implanting adult hypophysis into sexually immature Fundulus. Biol. Bull., 19:101 -lU. Melampy, R. M., & L. F. Cavazos 1954. Comparative study of lipids in vertebrate testes. Proc. Soc. Exp. Biol. & Med., 87: 297-303. Miller, M. R., & M. E. Robbins 1955. Cyclic changes in the pituitary gland of the urodele amphibian, Taricha torosa (Triturus torosus). Anat. Rec., 122:105- 114. Purves, H. D., & W. E. Griesbach 1951. Pituitary basophils as the source of the thyrotrophic and gonadotrophic hormone. Proc. U. Otago Med. Sch., 29:3-4. 1954. The site of follicle stimulating and lutein- ising hormone production in the rat pitui- tary. Endocrinology, 55:785-793. Rasquin, P., & L. M. Stoll 1955. Effects of pitressin and water injections on the secretions of brain and hypophysis in a teleost. Anat. Rec., 122:452-453. Riley, G. M., & R. M. Fraps 1942. Relationship of gonad-stimulating activity of female domestic fowl anterior pitui- taries to reproductive condition. Endocrin- ology, 30:537-541. Scruggs, W. M. 1951. The epithelial components and their sea- sonal changes in the pituitary gland of the carp (Cyprinus carpio, L.) and goldfish (Carassius auratus, L.). J. Morph., 88: 441-469. Sokol, H. W. 1953. Selective staining of the pituitary gland of teleosts demonstrating at least six tinctorial cell types in the adenohypophysis. Anat. Rec., 117:582 and personal communica- tion. Tavolga, W. N. 1954. Reproductive behavior in the gobiid fish, Bathygobius soporator. Bull. Am. Mus. Nat. Hist., 104:427-460. 1955. Effects of gonadectomy and hypophysec- tomy on prespawning behavior in males of tire gobiid fish, Bathygobius soporator. Physiol. Zool., 28:218-233. Verhoeven, B., & G. j. van Oordt 1955. The influence of light and temperature on the sexual cycle of the bitterling, Rhodeus amarus (preliminary communication). Ko- nink. Neder. Akad. von Wetenschappen. Proc. Series C, 58:628-634. Vivien, J. H. 1941. Contribution a I’etude de la physiologie hypophysaire dans ses relations avec I’ap- pareil genital, la thyro'ide et les corps suprarenaux chez les poissons, selaciens et teleosteens. Scylliorhinus canicula et Gobius paganellus. Bull. Biol. France et Belg., 75:257-309. 104 Zoologica: New York Zoological Society [42: 9: 1957] EXPLANATION OF THE PLATES Plate I Fig. 1. Partially degranulated basophils in a 41 mm. fish three days after pituitary im- plantation. Paraldehyde fuchsin, sagittal section, 1500 X. Fig. 2. Peripherally located degranulated baso- phils from a 40 mm. fish killed one week after pituitary implantation. Paraldehyde fuchsin, sagittal section, 1500X. Fig. 3. Normal acidophils and basophils from 43 mm. control fish, paraldehyde fuchsin, sagittal section, 1500X. Fig. 4. Ovary of 55 mm. untreated juvenile. Haematoxylin and eosin, cross section, lOOX. Fig. 5. Ovary of 73 mm. untreated adult. Haema- toxylin and eosin, cross section, 100 X. Fig. 6. Stimulated ovary from 42 mm. fish after one week of pituitary implantation. Haematoxylin and eosin, cross section, lOOX. Plate II Fig. 1. Core and spermatogenic tissue of 41 mm. untreated juvenile. Haematoxylin and eosin, cross section, lOOX. Fig. 2. Core and spermatogenic tissue of 80 mm. untreated adult. Haematoxylin and eosin, cross section, lOOX. Fig. 3. Core tissue of 41 mm. untreated juvenile stained with Sudan Black B, cross section, 500X. Fig. 4. Core tissue of 35 mm. juvenile after one week of implantation. Sudan Black B, cross section, 500 X. Fig. 5. Core tissue of 81 mm. adult stained with acid haematin, cross section, 1500X. Fig. 6. Core tissue of 74 mm. adult after pyridine extraction, cross section, 1500X. STOLL PLATE I FIG. 2 FIG. 3 • CO , s> ’«s»* FIG. 4 FIG. 1 FIG. 5 FIG. 6 CHANGES IN THE CYTOLOGICAL STRUCTURE OF THE ADENOHYPOPHYSIS AND GONADS IN JUVENILE BATHYGOBIUS SOPORATOR AFTER PITUITARY IMPLANTATION STOLL PLATE II FIG. 3 FIG. 4 • # A 'V' FIG. 5 FIG 6 CHANGES IN THE CYTOLOGICAL STRUCTURE OF THE ADENOHYPOPHYSIS AND GONADS IN JUVENILE BATHYGOBIUS SOPORATOR AFTER PITUITARY IMPLANTATION 10 The Ctenuchidae (Moths) of Trinidad, B. W. I. Part I. Euchromiinae/ Henry Fleming Department of Tropical Research, New York Zoological Society, New York 60, N. Y. (Plates I-III) [This paper is one of a series emanating from the tropical Field Station of the New York Zoological Society, at Simla, Arima Valley, Trinidad, British West Indies. This station was founded in 1950 by the Zoological Society’s Department of Tropical Research, under the direction of Dr. William Beebe. It comprises 200 acres in the middle of the Northern Range, which includes large stretches of undisturbed government forest reserves. The laboratory of the station is intended for research in tropical ecology and in animal behavior. The altitude of the research area is 500 to 1,800 feet, with an annual rainfall of more than 100 inches. [For further ecological details of meteorology and biotic zones see “Introduction to the Ecology of the Arima Valley, Trinidad, B.W.I.,” William Beebe. (Zoologica, 1952, Vol. 37, No. 13, pp. 157-184).]. Contents Page Introduction 105 Euchromiinae 106 Pseudosphex 106 Pleurosoma 108 Sphecops 108 Homoeocera 108 Isanthrene 108 Phoenicoprocta 109 Loxophlebia 113 Mesothen 114 Pheia 115 Chroslosoma 116 Leucotmemis 116 Nyridela ...116 Cosmosoma 117 Dixophlebia 118 Pseudomya 118 Rhynchopyga 119 Saurita 119 ^Contribution No. 978, Department of Tropical Re- search, New York Zoological Society. Psoloptera 122 Dycladia 122 Syntomeida 122 Histiaea 123 Macrocneme 123 Calonotos 127 Introduction This paper concerns the species of moths belonging to the subfamily Euchromiinae that have been reported in the literature or collected by the Department of Tropical Re- search of the New York Zoological Society at its biological station at Simla, Arima Valley, Trini- dad, B.W.I. The subfamily Ctenuchinae will be treated in a subsequent paper which will also include keys to all the genera of Ctenuchidae found in Trinidad, as well as the bibliography. I have followed Travassos (1935: 437-451) in the selection of the family name for this group of moths. Ctenuchidae has priority over other names in use as follows; Ctenuchidae, Kirby, 1837. Syntomidae, Snellen, 1867. Euchromiidae, Neumoegen & Dyar, 1893. Amatidae, Jansen, 1917. While the type genus of Syntomidae, Syntomis, has been found to be a junior synonym of A mata, the emendation of the XIVth International Con- gress of Zoology at Copenhagen, 1953, states that a family group taxon based on a synon- omized genus is not to be changed (Copenhagen Decisions on Zoological Nomenclature: 36, par. 54 (1) (a), 1953). The Syntominae are limited to the Old World and do not concern us in this paper. 105 106 Zoologica: New York Zoological Society [42: 10 No attempt has been made to make the refer- ences under the species complete. References to the original description, pertinent or new synonomy, colored figures, helpful descriptions to the species or a specific reference to Trinidad have been cited, however. Three publications cited whenever possible are: Hampson, G. F.: Catalogue of the Lepidop- tera Phalaenae, I, Syntomidae, 1898; Sup- plement I, Amatidae, 1914. Seitz, A.: Macrolepidoptera of the World, VI, Euchromiidae, 1915 and 1917. Kaye, W. J. & N. Lamont: A Catalogue of the Trinidad Lepidoptera Heterocera. Mem. Dept. Agric. Trinidad and Tobago, No. 3, 1927. This paper includes keys to the species of moths found in Trinidad and photographs of the species collected at Simla, Arima Valley, in an attempt to make it useful to biologists working on ctenuchids in Trinidad. Five new species of Euchromiinae are described. The species of Ctenuchidae of Trinidad are continental rather than Caribbean. Trinidad and the adjacent island of Tobago are geologically part of South America rather than of the West Indies, and the faunistic and floristic character of the two islands is decidedly related to Guiana and Venezuela. Our own collecting almost without exception has been confined to the Arima Valley in the Northern Range of Trinidad. However, Kaye & Lamont (1924) listed the species known to occur on the island as a whole and Lamont & Callan (1950) added two species to the Euchromiinae. The species reported by these authors are included in this paper. There are extremely few records from the southern part of the island and additional species may be ex- pected when this region has been investigated. My thanks go to Miss Rosemary Kenedy, who made notes and took photographs of many of the holotypes of the ctenuchid species in the British Museum (Natural History) which aided in the determination of some of the species in question. She also collected the majority of the department’s ctenuchid collection. Thanks go also to Dr. William Beebe and Miss Jocelyn Crane for their part in assembling the collec- tion and for advice and criticism. Euchromiinae The absence of vein Sc in the hindwing sepa- rates the Euchromiinae from other families of Trinidad moths. In this subfamily vein M2 of the hindwing is rudimentary or absent but often represented by a vein-like line of scales, whereas in the subsequent subfamily, Ctenuchinae, vein M2 of the hmdwing is present. Veins Cui and Cu2 are stalked or united in the Euchromiinae, but in the Ctenuchinae, with the exception of the Horama and related genera, they are widely separated. PsEUDOSPHEX Hubner This and the following two genera are wasp- like in appearance, with constricted abdomens. Pseudosphex kenedyae, new species (PI. I, fig. 1; PI. II, fig. 1) Pseudosphex melanogen Beebe, not Dyar, 1955; 32, fig. 6. Length of forewing 1 1 mm. Closely related to Pseudosphex laticincta Hampson. P. kenedyae differs from P. laticincta principally in lacking the dorsal band on the first abdominal segment and on the lateral bullae. Ma/e.— Antennae bipectinate. Processes on first segment of antennae small but gradually increasmg in length to the seventh segment where it reaches a length of .142 mm. The pro- cesses are but slightly dilated, pubescent anteri- orly and with a bristle at the distal end. Near the distal third of the antennae the pectinate processes decrease in length and resolve into three dentate segments followed by 14 serrate segments. The serrate segments are blackish- brown and the remainder of the antennae red- dish-brown. Palpi concolorous yellow with a fan-shaped tuft of hairs on the basal palpal segment yellow interspersed with black hairs. Front yellow but in rubbed specimens may have a whitish cast. A large black spot on vertex of head surrounded by yelTow. Side of head behind eyes yeUow with occasional black hairs or scales. Patagia yellow with a broad black bar behind middle. Tegulae yellow edged with black; broad- ly edged with black at the costal margin of the fore wing where it projects slightly on the sub- costal and cubital veins. Mesothorax (scutum) black with two diagonal yellow stripes and a mid-dorsal line of yellow scales broadening cau- dally. Hind edge of scutum with yellow scales. Metathorax yellow. Legs yellow. The inner face of the forecoxae shining white. The distal edge with a few scat- tered yellow scales. Inner face of forefemur shining white but somewhat variable in width in different lights, with the cephalad edge yellow. Frequently on the outer edge of the distal tip of the femur of the meso- and metathoracic legs a very small patch of shining white scales. In different lights the yellow color of the thoracic structures may take on an orange-yel- low cast. 1957] Fleming: Ctenuchidae (Moths) of Trinidad, B.W.I. 107 Forewing hyaline. Costal area, area beyond discal cell, part of discal cell above discal fold and area below vein 2dA fuscus brown. Re- mainder of wing with scattered setae. Hindwing hyaline with scattered setae except the costal area above the cell and vein Rg which is fuscus, similar in color to the respective area of the forewing. Dorsum of first abdominal segment and bul- lae yellow with the anterior margin finely edged with black. Second abdominal segment (pedun- cle) translucent yellow. The subsequent caudal segment with the anterior half tan brown and the posterior half translucent yellow. The mid- dorsum of this segment may have a slight mid- dorsal line of darker scales. Fourth, fifth and sixth segments of the abdomen with the two- thirds of the cephalad portion of the segment blackish-brown and the caudal third translucent yellow. Seventh abdominal segment dark and yellow area approximately equal. Last segment with only the cephalad quarter of the segment blackish-brown. Terminal tuft of the abdomen yellow. Ventrum of the abdomen with the anterior part of the valve (peduncle) relatively unsealed, usually gray but sometimes dark brown., Often with scattered occasional yellow scales. Approxi- mately at the point where the caudal lateral edge of the dorsal sclerite of the peduncle seg- ment touches the ventral vein, the valve be- comes translucent yellow. The valve is fringed on its caudal edge with white scales approxi- mately a fifth of the length of the yellow part of the valve. Remainder of the ventrum yellow peppered with black scales. Female similar to the male except wing ex- panse greater (12 mm.), and the basal and medial part of the antennae prismatic. The last 12 distal antennal segments are slightly serrate, dark grayish-brown in our single female and the three preceding segments light brown and dentate rather than prismatic. The remaining segments of the antennae are warm reddish- brown. The legs are concolorous yellow with no white areas. Closely related to P. laticincta from which it differs by having a concolorous first abdominal segment and bullae whereas in P. laticincta the first abdominal segment has a dorsal stripe and the bullae are banded. Rosemary Kenedy compared part of our series with Hampson’s type of P. laticincta in the British Museum (Natural History) and noted that in the male of P. kenedyae the an- tennal shaft is darker than the type of P. lati- cincta so that the basal darkening is not so apparent. P. kenedyae has a longer and more restricted antennal extremity preceded by a con- trasting lighter area and the pectinations slightly longer and not so dilated. The light brown ab- dominal band on the second segment of P. kenedyae is a black band in P. laticincta and the same color as the subsequent black abdomi- nal bands. The dark abdominal bands of P. laticincta are narrower and extend to slight points behind mid-dorsally, more so than in P. kenedyae. In P. kenedyae only the fifth and sixth abdominal segments consistently have pro- nounced black points. It is also important to note from Miss Ken- edy’s memoranda that the palpi of P. laticincta are similar to P. kenedyae. The palpi are yellow with a fan-shaped process from the basal palpal segment with yellow scales interspersed with black scales. In other words, the palpi are not banded in the usual meaning of the term in P. laticincta, as the original description would lead one to beheve. P. kenedyae difi'ers from P. melanogen Dyar in having more yellow on the thorax and much wider yellow bands on the abdomen. The same considerations are true for P. deceptans Zerny which was synonymized under P. melanogen by Hampson. Neither P. melanogen nor P. decep- tans are described as having white forecoxae. For purposes of Trinidad identification, the broad fuscus costal margin of the forewing wUl separate this species from the species in the fol- lowing two genera. I take pleasure in naming this species for Miss Rosemary Kenedy, Research Assistant of the Department of Tropical Research, whose initia- tive in the use of Heliotropium as an attractant for euchromids brought to light this new species. Material— AW of the types were taken at Simla, Arima Valley, Trinidad, on Heliotropium indi- cum. Holotype, male. Catalog No. 5639, 20-11; allotype, female, (5640) 29-IV; 15 para- types, males, (5641) 14-1, (5642) 15-1, (5643) 20-1, (5644) 28-1, (5645) lO-II, (5646) 20-11, (5647) 22-11, (5648) 23-11, (5649) (5650) (5651) 4-IV, (5652) 29-IV, (5653) 1-V, (5654) 27-XII,(5655) 31-XII. Disposition of type material.— The. Depart- ment of Tropical Research, New York Zoolog- ical Society, will retain three paratypes. Catalog Nos. 5650, 5651 and 5655. Paratypes with Cata- log Nos. 5648 and 5653 are in the British Mu- seum (Natural History) collection and paratypes with Catalog Nos. 5644 and 5652 are in the United States National Museum collections. The holotype, allotype and remaining paratypes are in the American Museum of Natural History, New York. 108 Zoologica: New York Zoological Society [42: 10 Pleurosoma Orfila This and the following genus may be distin- guished from the preceding genus, Pseudosphex, by vein Rs of the forewing arising basad of vein R.3. The genus Astridia Kiriakoff, 1948: 267, is an absolute synonym of Pleurosoma Orfila, 1935: 178. Both genera were erected with Sphecosoma angustatum Moschler as the type. Pleurosoma trinitatis (Rothschild) (PI. Ill, fig. 1) Sphecosoma /rmhatw Rothschild, 1911: 24. Sphecosoma trinitatis, Rothschild, 1913: 471, pi. XIV, fig. 5. Sphecosoma trinitatis, Hampson, 1914: 90. Sphecosomatrinitatis,Y>XdL\id.i,\9l5'. 40,fig. lOh. Sphecosoma trinitatis, Kaye&Lamont, 1927: 1. Pleurosoma trinitatis, Orfila, 1935: 178. Described by Rothschild from four male specimens collected at Caparo, Trinidad, in De- cember, 1905, by S. M. Klages. The figures in Rothschild and Seitz are mis- leading, for they indicate that the color of the abdominal segment following the peduncle is black, similar to the subsequent segments, rather than rufous brown. The length of the forewing in our specimens is 12 mm. The length of the forewing of the holotype in the British Museum is 12 mm., rather than the 14 mm. given by Rothschild. Female similar to male except antennae pris- matic rather than bipectinate. Material— VitiQen males and eight females. Rfl/rge.— Trinidad. Sphecops Orfila Differs from Pleurosoma in lacking the two thoracic bladder-like processes beneath the base of the first abdominal segment. First segment of abdomen much shorter in Sphecops than Pleuro- soma. The males of Sphecops have a ventral valve covering the second and third abdominal segments which is absent in Pleurosoma. Sphecops aurantiipes (Rothschild) (PI. Ill, fig. 2) Sphecosoma aurantiipes Rothschild, 1911: 25. Sphecosoma aurantiipes, Rothschild, 1913: 471, pi. XIV, fig. 7. Sphecosoma aurantiipes, Hampson, 1914: 91. Sphecosoma aurantiipes, T>Tavidt, 1915: 40, fig. lOh. Sphecops aurantiipes, Or^Xa, 1935: 181. Four of our Trinidad specimens have been compared with Rothschild’s type from San Es- teban, Venezuela, in the British Museum by Miss Rosemary Kenedy. Rothschild gives the length of the forewing as 14 mm. in his original descrip- tion, whereas our measurement of the type is 12.5 mm. The Trinidad specimens have a fore- wing length of 10 to 11 mm. In this respect they resemble material that Rothschild had before him from Paraguay and Bolivia which he re- ported as having a wing length of 10 mm. The yellow bands on the abdomen of the Trinidad specimens are narrower than in the type from Venezuela. Material.— Seven males. Range.— Venezuela, Paraguay and Bolivia. A new record for Trinidad. Homoeocera Felder A genus related to Gymnelia, consisting of large, heavy-bodied species. We have not taken any species of this genus in the Arima valley. Homoeocera magnolimbata Dognin Homoeocera magnolimbata Dognin, 1911 : 11:3. Homoeocera magnolimbata, Hampson, 1914: 1:95, pi. V, fig. 1. Homoeocera magnolimbata, Draudt in Seitz, 1915: VI:46, pi. 27b. Kaye & Lamont (1927) report this species from Trinidad. One specimen appears to have been taken in Trinidad by F. W. Urich at St. Ann’s in October, 1899. This specimen is not in the Urich Insect Collection at the Victoria Mu- seum in Port-of-Spain. It is a large insect, ap- proximately two inches in expanse, with hyaline wings and black abdomen with conspicuous white spots and some metallic blue. Kenedy reports another specimen from Trini- dad in the British Museum (Natural History) in the Joicey Collection, collected in 1909. It was part of a series from French Guiana, Amazons, East Bolivia and Peru. Isanthrene Hubner The thorax of the species in this genus is smoothly scaled in contrast to the species of Homoeocera which have hairy thoraxes. Isanthrene tryhanei Rothschild Isanthrene try/zanei Rothschild, 1911: XVIII: 26. Isanthrene tryhanei, Rothschild, 1913: XX: 471, pi. XIV, fig. 24. Isanthrene tryhanei, Hampson, 1914: I: 98. Isanthrene tryhanei, Draudt in Seitz, 1915: VI: 47, pi. 10b. 1957] Fleming: Ctenuchidae (Moths) of Trinidad, B.W.I. 109 Rothschild described this species from St. Aim’s, Trinidad, from one female. It does not appear to have been collected since the original specimen was taken. It is another large species approximately two inches in expanse, with hya- line wings. The wings have a yellowish cast particularly along the costal and inner margin, in contrast to Homoeocera magnolimbata. Phoenicoprocta Druce (not Hampson) Stephens (1850) used the name Hyela for a generic name in Noctuidae. Walker (1854) used Hyela for one of his groups in the genus Glau- copis with the intention that the species he placed under Glaucopis would henceforth take the various group names as generic names. Hampson (1898) realized that Walker had em- ployed a name, Hyela, that was preoccupied by Stephens, and erected a new name Phoenico- procta. Druce (1898) was apparently cognizant of Hampson’s monograph of the Ctenuchidae and described a presumed new species, Phoeni- coprocta metachrysea. This species employing Hampson’s new name was published in May, 1898, in advance of Hampson’s monograph, which was not published until September, 1898. Thus Druce must be credited with the generic name Phoenicoprocta. The type of the genus is Phoenicoprocta metachrysea by monotypy which Dyar (1915) synonymized under coprocta vacillans Walker (1856) on the basis of breeding carried out by H. W. B. Moore in British Guiana. The males of this genus are easily distin- guished in Trinidad by the blue or red spotted abdomen and carmine anal tufts. The Aethria species with a carmine tuft have a concolorous blackish abdominal dorsum. The females have a Calonotus-\\ke abdomen and may be distin- guished from Calonotus by vein R2 in Phoeni- coprocta being forked with veins R3+1+5 rather than from the cell as in Calonotus. Phoenicoprocta vacillans (Walker) PI. Ill, figs. 3, 4, 5 Eunomia vacillans Walker, 1856: VII: 1617 (male). Phoenicoprocta vacillans, Hampson, 1898: I: 197, pi. VII, fig. 11 (male). Phoenicoprocta vacillans, Draudt in Seitz, 1915 : VI: 47, pi. 12a (male). 1917: Supp.: 199. Phoenicoprocta metachrysea Druce, 1898: (7) I: 404 (male). Phoenicoprocta metachrysea, Hampson, 1898: I: 196, pi. VII, fig. 12 (male). Phoenicoprocta metachrysea, Draudt in Seitz, 1915: VI: 47, pi. 12a; (male). 1917: Supp.: 199. Phoenicoprocta chrysorrhoea Hampson, 1898: I: 196 (male). Phoenicoprocta chrysorrhea, Draudt in Seitz, 1915; VI: 57, pi. 12a (male). 1917: Supp.: 199. Phoenicoprocta trinitatis Strand, 1915: 21 (male). Phoenicoprocta trinitatis, Draudt in Seitz, 1915 : VI: 58, pi. 27f (male). Phoenicoprocta nigropeltata Strand, 1915: 22 (male). Leucotmemis albigutta Schaus, 1905; XXIX, No. 1420: 185 (female). Leucotmemis albigutta, Hampson, 1914: Supp. 1 : 1 5 1 , pi. VIII, fig. 2 (female) . Leucotmemis albigutta, Draudt in Seitz, 1915: VI: 70, pi. 27k; 1917; Supp.: 201 (female). Leucotmemis thoracica Schaus, 1905: XXIX, No. 1420: 186 (female). Leucotmemis thoracica, Hampson, 1914: Supp.: I: 150, pi. VIII, fig. 1 (female). Leucotmemis thoracica, Draudt in Seitz, 1915: VI: 70, pi. 27k; 1917: Supp.: 201 (female). Antichloris trinitatis Rothschild, New Syntomi- dae. 1912: XIX: 154 (female). Autochloris trinitatis, Hampson, 1914: Supp. 1: 104, pi. V, fig. 17 (female). Antichloris trinitatis, Draudt in Seitz, 1915: VI; 136 (female). Autochloris trinitatis, Draudt in Seitz, 1917: VI: Supp.: 197, pi. 27c (female). Paramya chrysonota Hampson, 1898, I: 165, pi. VII, fig. 3 (female). Paramya chrysonota, Draudt in Seitz, 1915 : VI : 44, pi. 10k (female). Calonotus hoffmannsi Rothschild, 1911 : XVIII; 30 (female). Calonotus hoffmannsi, Rothschild, Some unfig- ured Syntomidae, 1913: XX: 470, pi. XIII, fig. 35 (female). Leucotmemis hoffmannsi, Hampson, 1914: Supp. I: 150 (female). Paramya? hoffmannsi, Draudt in Seitz, 1915: VI: 44 (female). Calonotus hoffmannsi, Draudt in Seitz, 1915: VI: 108, pi. 17k; 1917: Supp.: 201 (female). Antichloris trinitatis, Kaye & Lamont, 1927 ; No. 3; 9. Phoenicoprocta trinitatis, Kaye & Lamont, ibid: 1927; 1. Phoenicoprocta rubiventer? Kaye & Lamont, 1927: 1. Mydropastea chrysonota, Kaye & Lamont, 1927: 1. 110 Zoologica: New York Zoological Society [42: 10 The extreme variability of this species is indi- cated by the above synonymy, which undoubt- edly is incomplete. Species designation has been based on the color of the collar, shoulder-covers, thorax, legs, abdomen and the ventral valve at the base of the abdomen in the males; and in the females, aside from the fact that they have been placed in the wrong genera, largely on the amount of hyaline areas in the wings. In all of our specimens of Phoenicoprocta from Trinidad, British Guiana and Venezuela, and specimens loaned to us by the American Museum of Natural History from Mexico, Brit- ish Guiana, Brazil and Peru, there is a small accessory cell in the forewing formed by a sec- torial cross-vein between vein R2 and vein R3+4+5. This sectorial cross-vein emerges from R3+4+5 typically at the same point at which vein Rs forks but may vary to the extent of being one millimeter more basad. Hampson (1898, p. 196) makes no mention of this cross-vein, which appears to be a good generic character. In addi- tion, he states that vein 3 (vein Cui) of the hindwing is absent whereas it is invariably pres- ent though shortly stalked near the margin of the wing. Furthermore, the genus is not divisible into two sections on Hampson’s characters. In the species that he places in the first section, the discocellulars in the hindwing are not oblique throughout while in the forewing, vein 3 (vein Cui) is variable in a series from the same local- ity. The commonest form at Simla, Trinidad, is nigropeltata Strand which was described from Trinidad. The character distinguishing this form from the typical vacillans is a black ventral valve with a white posterior edge, rather than a white valve with a black base and lateral edges. We have fifteen specimens that agree with this diag- nosis of nigropeltata, but in four additional specimens the white edge is replaced by pink. Hampson (1898, p. 197), in his diagnosis of vacillans, stated “forecoxae white and crimson” (forecoxae white, red inside, as Kenedy noted on holotype). Eight of our 19 specimens agree in this respect, although more pink than red, and the remaining 1 1 specimens have the coxae white in front and blackish-brown inside. Four of these latter specimens have the pink-edged valve. Ten additional specimens appear to be trini- tatis Strand. This species is described as having short palpi. In our specimens the palpi appear to be the usual length. The origin of vein Cui in the forewing is variable. The amount, posi- tion and presence of blue, crimson and black on the collar are extremely variable, encompass- ing trinitatis, sanguinea and new forms. The colors of the patagia and tegulae vary consider- ably in different lights because of their iridescent cast alone. The subdorsal macular stripes on the abdomen also vary considerably; sometimes they are large and distinct and in others, particularly near the base of the abdomen, they may be quite small. The ventrum of the abdomen in these specimens varies between black and blackish- brown. The ventral valve is light to dark pink and in one specimen with a considerable mix- ture of black scales. The posterior edge of the valve is finely white. The forecoxae in these ten specimens are white. The inside of the fore- coxae and the remaining coxae usually red but sometimes blackish-brown. In an additional specimen, an eleventh speci- men, the shoulder covers are dark brown, fore- coxae white and remaining coxae blackish- brown, the last four abdominal segments with a macular red subdorsal line and the ventral valve black with a white edge. This specimen is an unnamed, more melanotic, form of the preced- ing ten specimens. Two specimens are very similar to sanguinea Walker. Our specimens differ in having a black transverse line at the base of the anal tuft and a narrower discocellular bar in the forewing. In the forewings of sanguinea Walker there is a small hyaline spot above vein R5 which is definitely absent in all of our specimens. In Draudt’s figure in Seitz (fig. 15a; 1915) the shoulder covers and the abdomen are reddish- brown rather than the correct color, crimson. Hampson (p. 198, 1898) gives the type locality as Honduras. Walker (1854, p. 172) in his origi- nal description did not state the origin of his specimen. The holotype in the British Museum (Natural History) has a hand-printed label “Honduras” in neither Walker’s nor Hampson’s writing. The sexes in Phoenicoprocta vacillans are dimorphic. Besides, the amount of scaling in the wings of the females in Trinidad is extremely variable. The female specimens will run in Hampson’s generic key (1898) to Mydropastea (Phaeo) or Paramya (Methysia) since Hampson’s dicho- tomous couplet (1898: 23 (B, a^), 24 (b^)) is based on the presence of abdominal tufts which are a male character. The abdominal aspect of the females differs radically from that of the males, which have a typically maculated series of red or iridescent blue spots whereas the females have blackish and iridescent blue longitudinal stripes. The abdomens of both males and females are somewhat bulbous caudad of the third or fourth abdominal segments. 1957] Fleming: Ctenuchidae (Moths) of Trinidad, B.W.I. Ill The abdomens of all the female specimens are similar in showing a mid-dorsal stripe of irides- cent blue or blue-green scales. This line is con- tinued on the metathorax, which has a single mid-dorsal spot of the same color. The mid- dorsal stripe of the abdomen is bordered by a blackish-brown stripe on each side commencing on the first abdominal segment and followed laterally by another stripe of iridescent blue or blue-green which is replaced on the first ab- dominal segments by iridescent spots on a black- ish-brown ground. Laterally the last subdorsal iridescent stripe is bordered by a fine blackish- brown line. In old or rubbed specimens the iri- descence of the abdomen may have a coppery cast. The ventrum of the abdomen is dark brown to brownish-black and iridescent in some lights. There are iridescent blue points beneath the wings on the thorax and a line of iridescent blue on the caudal edge of the metathorax continuous from each side. The amount of carmine on the under side of the coxae is variable but is present in all female specimens on the distal end of the prothoracic coxae and frequently on the whole underside of the forecoxae but usually only at the distal apex of the meso- and meta- thoracic coxae. The anterior surface of the prothoracic coxae is marked with iridescent blue on a brown background, most distinct on the outer edge of the coxae. Female Form /.—Two specimens. Both the fore and hindwings of this form are immaculate blackish-brown with no hyaline spots whatso- ever. Some blue scales along the radius vein of the forewing. Kenedy compared this form with the holotype of Antichloris trinitatis Roth- schild in the British Museum (Natural History) and believes them to be identical. Female Form //.—This form differs from Form I in having a short elliptical hyaline spot below the middle of the cell and hyaline spots below the cell on each side of vein Ms in the forewing. The hindwing has a hyaline spot below the cell, a small spot at the forking of vein Cui+2 and vein M,s and a large spot above vein Ms to vein Mi. This last mentioned hyaline spot decreases in size approximately one-third above the line of scales representing vein Ms. One of the three specimens has crimson patagia. The ab- dominal and wing patterns of this form are simi- lar to the figure in Draudt in Seitz (pi. 101; 1915) of Mydropastea chrysonota Hampson. Female Form ///.—Differs from Form II in having two or three small hyaline spots above the larger hyaline spots on each side of vein Ms of the forewing. The spot anterior to the large spot in cell Ms, if present, is minute. The hyaline area beneath the discal cell is larger and may extend slightly below the anal vein. A slight hyaline streak within the discal cell along the cubital vein. In the hindwing the hyaline spot below the cell is larger than in Form II and may extend slightly below the anal vein. Female Form /F.— Three specimens were taken of this form. The hyaline areas in the forewing are progressively more extensive by the addition of a hyaline spot in the forewing between veins Cui and Cus, a large hyaline streak below the anal vein and the posterior por- tion of the discal cell. In the hindwing the hyaline area extends below the anal vein. None of the above four female forms is to be interpreted as being stable, as they grade into one another. Paramya chrysonota Hampson is a synonym of Phoenicoprocta vacillans Walker. One speci- men of chrysonota was collected at Itaituba, Amazons, Brazil, and described as a male. Kenedy inspected the holotype of chrysonota in the British Museum (Natural History) and dis- covered that, in point of fact, it is a female rather than a male. It agrees with our female series in Form II from Simla, Trinidad. Some discussion of the past history of Par- amya chrysonota would not be out of place. Section I of Paramya Hampson (p. 164; 1898) establishes with Paramya intersecta Hampson (p. 164; 1898) as the genus type is synonymized by Hampson (p. 384; 1914) under Methysia Butler with the genus type species Methysia (Glaucopis) notabilis Walker (1854) on the basis that Hampson’s figure (fig. 205, 1898) of Methysia notabilis is correct. Thus Methysia contained the following species: notabilis Walk- er, intersecta Hampson, senetus Schaus, picta Druce and melanota Hampson. This left brice- noi Rothschild, flavia Schaus, picta Druce and chrysonota Hampson with no proper generic name as these species were in Hampson’s section II of the genus Paramya. However, Hampson was not the legitimate author of Paramya nor was he privileged to select a type species, as Druce in May of the same year (1898) and also, for that matter, Schaus in June had each described new species employing the generic name Paramya. Hamp- son in his Cat. Lep. Phal. I, published in Sep- tember, included both Druce’s and Schaus’s new species in the second section of his presumed new genus. Neither Druce nor Schaus wrote generic descriptions and what apparently hap- pened was that both were familiar with the new genus that Hampson was to erect in his mono- graph and described their new species in that genus. Unfortunately their descriptions were published a few months previous to Hampson’s monograph. The type species of Paramya is Paramya picta Druce. 112 Zoologica: New York Zoological Society [42: 10 However, Paramya was preoccupied by Con- rad (1860) in Mollusca and, consequently was unavailable for Druce’s species. Travassos (1946) erected a new name, Metamya, for Paramya Druce. The genotype is Metamya picta Druce, and Paramya flavia Schaus and Ichoria bricenoi Rothschild are congeneric. Paramya chrysonota Hampson (1898), with which we are concerned, is not congeneric with either section of Hampson’s original conception of the genus and is a female Phoenicoprocta. Kaye & Lamont (1927) placed Paramya chrysonota Hampson (1898) in the genus Mydropastea Hampson (1898). I presume they did this because Paramya as conceived by Hamp- son has veins Rg and M, of the hindwing long stalked. Apparently the specimens that Kaye & Lamont possessed from Trinidad resembled our specimens from Trinidad in which veins Rg and M, are approximate. Thus, they placed the species in Mydropastea, albeit incorrectly. In Hampson’s Supplement (p. 209; 1914) he synonymized his genus Mydropastea (1898) under Phaio Neumoegen (1894). Draudt in Seitz picked up the synonymy but changed the spelling of the genus to Phaeo following Wagner (1912). Calonotus hoffmannsi Rothschild (1911) is a synonym of Phoenicoprocta vacillans Walker. Draudt in Seitz (p. 108, 1915; p. 201, 1917) was correct in his suggested synonymy, namely, Calonotus hoffmannsi (1911) is a synonym of Paramya chrysonota (1898). Hampson in his Supplement (p. 150; 1914) placed hoffmannsi in Leucotmemis. Calonotus hoffmannsi was described from Itaituba, Brazil, the same type locality as Paramya chrysonota. The male genitalia of the specimens from Trinidad, two from Caripito, Venezuela, and a specimen from British Guiana are similar. How- ever, the genitalia of two specimens, one from Candelopa, Peru, and another from Pitaguaya, Bolivia, while very similar in other respects, have a distinctly different pair of protuberances on either side of the base of the uncus. They are rounded and bulbous in the latter specimens and pointed in our vacillans. The general facies of the Peruvian and Bolivian specimens appears to be the same as our material. Judging from the preceding discussion it is quite likely that a large number of the species described in Phoenicoprocta will be synony- mized. It is necessary, however, to study geni- talia of the holotypes or in lieu of that, of topo- typical material. In the subsequent paragraphs I shall give an indication of the probable dis- position of some of the species of Phoenico- procta. Phoenicoprocta mexicana (Walker) will have to be removed from the genus. The sexes are not dimorphic. The wing venation differs in the fore- wing by lacking the sectorial vein in the radials. It differs in the hindwing by veins Cui and Cu2 being completely united and veins Rg and M, being approximate, not connate or very shortly stalked. Forbes (1930) correctly removed Zygaena parthenii Fabricius from the genus Mallodeta where Hampson (1898) placed it. Nonetheless, it does not appear to be a Phoenicoprocta. The venation differs in the forewing by lacking the sectorial vein of the radials. It also differs in the forewing in that vein R5 forks from the radial midway between vein R2 and the forking of veins R3 and R4, whereas in vacillans it sepa- rates very near the forking of veins R3 and R4. I have not seen a female nor has the female been described, so I do not know if the sexes are dimorphic. The hindwing of partenii is similar in that veins Cu2 and Cus are forked near the margin of the wing and veins M, and Rg are slightly stalked. Phoenicoprocta paucipuncta Dyar is not a Phoenicoprocta but probably, as Forbes (1939) suggests, a color form of Cosmosoma gemmata Butler. The following species may well be only color forms of vacillans. Males: rubiventer Hampson from Panama; sanguinea (Walker) , the species type for which Hampson gives the type locality as Honduras though Walker did not know the type locality in his original description, the hand- printed label on the holotype stating Honduras being in neither Walker’s nor Hampson’s writ- ing; astrifera (Butler) from Braga (Amazons) Brazil; haemorrhoidalis (Fabricius) from Brazil; ffavipicta Hampson from British Guiana; varia- bilis Kaye from from Panama; intermedia Fors- ter from Venezuela and nigriventer Gaede from Venezuela; females: biformata Gibbs and the aberration atrapennis Strand from British Hon- duras; insperata (Walker) from Para, Brazil. Phoenicoprocta astrifera (Butler), as Draudt in Seitz figures it (pi. 12c; 1915), has wide terminal wing margins. The holotype has wing margins as in vacillans. The following were de- scribed as forms of vacillans and should be re- appraised: aurantipatagiata Draudt, aurifiua Draudt, nigricoxa Zerny and punicea Strand. While I have seen no specimens from the West Indies, the form existing there appears to be a distinct species. Hampson (1914: 125) synony- mized thomae (Lucas), cubana Druce and selecta (Herrich-Schaffer) under capistrata. Fabricius’s type locality for capistrata is “America”, which meant the mainland, and 1957] Fleming: Ctenuchidae (Moths) of Trinidad, B.W.l. 113 Herrich-Schaffer’s type locality for selecta is Brazil. Both are females and are more likely the female forms of vacillans while exima (Her- rich-Schaffer), described from Cuba, is the ac- tual female form from the West Indies. If the above suggestions prove to be correct, thomae (Lucas) has priority and capistrata (Fabricius) would preoccupy vacillans (Walker). Phoenicoprocta jamaicensis Schaus is based on a female type. Phoenicoprocta lydia (Druce) , with the synonyms thera (Druce) and demona (Druce), is a distinct species from Mexico. It has the same fore- and hindwing venation as the vacillans we have from Trinidad and is sexually dimorphic in the same fashion. The forms of Phoenicoprocta from South Brazil, Paraguay, Uruguay, Peru and Bolivia ap- pear to be distinct from the forms found from the Amazons north. Jorgensen’s species schrei- teri and sieboldi appear to be the slight variants of latimarginata Gaede (1926). The wider mar- gins of the wings alone would seem to distinguish these southern forms from vacillans. Two geni- talia from Bolivian and Peruvian specimens that were mentioned before are distinct from vacil- lans though the facies of the insects are similar. Phoenicoprocta teda (Walker) is a distinct species but quite possibly not a Phoenicoprocta. Loxophlebia Butler In our species of Loxophlebia the discocellu- lars are not oblique throughout in the hindwing as Hampson (1898: 206) states in his generic diagnosis. Furthermore, vein Cu2 of the hind- wing is present but very shortly stalked with vein Cui; the forking occurs at the inner edge of the black wing border. The following genera, Mesothen and Loxophlebia, may be separated from other Trinidad genera by vein M2 of the forewing originating well above vein Ms, more than a third but less than half the distance up between veins Mi and M3. The same vein, M2, in Pheia is approximately a fifth in our species. The genus Mesothen is very close to Loxophlebia and I am unable to separate these genera except superficially on the basis of the specimens on hand, so the following key will serve to dis- tinguish the Trinidad species in both genera. Surinam and British Guiana forms of Loxo- phlebia bisigna in the key would run to Meso- then aurantegula, so I have included an addi- tional couplet in the event that these forms are found in Trinidad in the future. 1 . Abdomen with some red or orange color 2 (Loxophlebia) Abdomen with no red or orange color 4 (Mesothen) 2. Abdomen with broad lateral orange bands diaphana Abdomen with no lateral bands of any color 3 3., Abdomen blackish-brown with subdorsal white spots on basal segment and sub- dorsal red spots on fourth abdominal seg- ment bisigna (Trinidad form?) Abdomen black with the last four segments orange postflavia 4. Disc of thorax orange -red pyrrha Disc of thorax black or blackish-brown . . 5 5. Tegulae with orange patches 6 Tegulae concolorous brownish-black. . . .7 6. Abdomen concolorous blackish-brown aurantegula Abdomen with subdorsal white spots on basal abdominal segments bisigna (extra-limital forms) 7. Male with abdominal ventral valve. Female with white subventral abdominal band endoleuca Male without abdominal ventral valve. Female without white subventral abdomi- nal band desperata Loxophlebia diaphana (Sepp) Glaucopis diaphana Sepp, 1848: II: 185, pi. 81. Glaucopis discifera. Walker, 1854: I: 178. Laeinocharis bura Butler, not Herrich-Schaffer, 1877: I: 33, pi. 11, fig. 12. Chrysostola albifronsy[os,ch\&r, 1872: XXXIII: 344. Loxophlebia diaphana, Hampson, 1898: I: 209. Loxophlebia diaphana, Draudt in Seitz, 1915: VI: 62, pi. 12g. Loxophlebia diaphana, Kaye & Lamont, 1927: No. 3 : 2., We have not collected this species at Simla, but Norman Lamont collected a specimen at Palmiste in January, 1921. Surinam and Amazons. Loxophlebia bisigna (Kaye) (PI. Ill, fig. 6) Pheia bisigna 1911: XLIV: 146. Pheia bisigna, Hampson, 1914: Supp. I: 128, pi. VI, fig. 17. Loxophlebia klagesi Rothschild, 1811: XVIII : 29. Loxophlebia klagesi, Rothschild, 1913: XX: 471, pi. 14, fig. 21. Loxophlebia clagesi, Hampson, 1914: Supp.: 135. 114 Zoologica: New York Zoological Society [42: 10 Pheia bisigna, Draudt in Seitz, 1915: VI: 61; 1917: Supp.: 199, pi. 27f. Loxophlebia klagesi, Draudt in Seitz, 1915: VI: 63, pi. 12g. Loxophlebia bisigna, Kaye & Lament, 1927 : No. 3: 2. I have followed Draudt’s (1915, 1917) syn- onymy of this species. The abdominal charac- teristics of the original type material of klagesi are variable. Rothschild’s paratype of klagesi from Caporo {sic. Caparo), Trinidad, is similar to our three specimens in having paired reddish subdorsal spots on the fourth abdominal seg- ment. Two paratypes from Maripa, Caura River, Venezuela, have paired reddish spots on the third and fourth abdominal segments. The four Suri- nam types along with Kaye’s holotype of bisigna from the Potaro River, British Guiana, lack these reddish spots. Kaye’s bisigna was described two months earlier than Rothschild’s klagesi. The label on the holotype of klagesi in the Brit- ish Museum (Natural History) has on the local- ity label “Aroewarwa Creek, Maroewyn valley, Surinam, May 1905. (S. M. Klages)” whereas in Rothschild’s original description the material from this locality is given as February. /?fl77gc.— Trinidad, Venezuela and Guianas. Loxophlebia postflavia Druce (PI. Ill, fig. 7) Loxophlebia postflavia Druce, 1898: I: 407. Loxophlebia postflavia, Hampson, 1898: I: 210, pi. VIII, fig. 4. Loxophlebia postflavia, Druce in Seitz, 1915: VI: 64, pi. 12h. Two specimens were collected. This is a new record for Trinidad. Described from French Guiana. Specimens in British Museum from British Guiana and Surinam. Mesothen Hampson In this genus, as in the former genus, Cui and Cu2 of the hindwing are forked within the dark terminal band. Thus Cu2 is present, not absent, as stated in Hampson’s original description (p. 212; 1898). The species of this genus have been keyed in the genus Loxophlebia. Mesothen aurantegula (Jones) Loxophlebia aurantegula D. Jones, 1914: 4, pi. 1, fig. 6. Mesothen aurantegula, Hampson, 1914: Suppl. I: 146, pi. VII, fig. 25. Mesothen aurantegula, Draudt in Seitz, 1917: VI: 200, pi. 27h. Mesothen aurantegula, Kay & Lamont, 1927: No. 3: 2. An unusual record based on one specimen col- lected by Norman Lamont at Palmiste in Janu- ary. The species was described from one female taken at Santos, southern Brazil. Rrtngg.— Southern Brazil and Trinidad. Mesothen endoleuca Druce (PI. Ill, fig. 8) Mesothen endoleuca Druce, 1905: XV: 460. Mesothen endoleuca, Hampson, 1914: Supp. I: 146, pi. VII, fig. 27. Mesothen endoleuca, Draudt in Seitz, 1915: VI: 66, pi. 27i. Female with a ventrolateral band of white scales on abdomen, broadest at the base of the abdomen. Black border at the apex and margins of the forewing much wider than in male. Fore- coxae white in both sexes. Eleven specimens were collected, of which five were females. Venezuela. A new record for Trini- dad. Mesothen desperata (Walker) Pseudomya desperata Walker, 1856: VII: 1602. Mesothen desperata, Hampson, 1898; 215, pi. VIII, fig. 3. Mesothen desperata, Hampson, 1914; 146, fig. 20. Mesothen desperata, Draudt in Seitz, 1915; 66, pi. 271. Mesothen desperata, Lamont & Callan, 1950: 197. One specimen reported by Lamont & Callan (1950: 197) was collected by Lamont in April at Palmiste. It is an aberrant species for the genus as the male lacks a ventral valve according to Hampson (1914: 146). Since our females of endoleuca have a ventrolateral band broadest at the base of the abdomen, not mentioned in the description of desperata, I have used this char- acter in the key to separate the female specimens of these two species. Range.— Panama, Guianas, Brazil, Paraguay and Argentina. Mesothen pyrrha (Schaus) (PI. Ill, fig. 9) Dycladia pyrrha Schaus, 1889: V: 89. Dycladia pyrrha, Druce, 1896: II: 348, pi. 71, fig. 27. Mesothen pyrrha, Hampson, 1898: I: 214. 1957] Fleming: Ctenuchidae (Moths) of Trinidad, B.W.I. 115 Mesothen pyrrha, Draudt in Seitz, 1915: VI: 66, pi. 12i. Mesothen pyrrha, Kaye & Lament, 1927 : No. 3 : 2. Only one of our male specimens has the valve white. In the remaining specimens it is brown with the margins at the base of the valve broadly and irregularly white. Only the proximal end of the coxae is white, while the remainder of the coxae is brown. Material.— Fom males and one female. Range.— MEXICO to Peru and the Guianas. Pheia Walker Very close to Cosmosoma but the lower side of the discal cell is very short and the discocellu- lar veins oblique. The only Trinidad species is considerably smaller than any of the Trinidad Cosmosoma. The wingspread of the smallest Trinidad species of Cosmosoma, klagesi and remotum, is 32 mm. whereas the following spe- cies is 22 mm. Pheia beebel, new species (PI. I, fig. 2; PI. II, fig. 2) Length of forewing of male 11-12 mm., of female 13 mm. Closely related to Pheia gaudens (Walker), from which beebei differs in its smaller size and narrower apical black patch on the forewings. Pheia gaudens was described from Para, Brazil, and has been reported from Venezuela and Peru. Antennae bipectinate with a tuft on the apex of each pectination. Pectinations along shaft of antennae on the respective sides of the an- tennae of similar length except near base and apex. Pectinations on outside of antennal shaft slightly longer than on the inside. Shaft of an- tennae of male blackish-brown and of female brown. Scape of antennae with bright yellow tuft of scales most conspicuous on the inside anterior edge. Palpi, front of head and behind eyes bright yellow. Basal segment of palpi clothed with long hair of irregular length, thus somewhat ragged in appearance. The remaining two segments of palpi finely scaled and upturned to vertex of head. Vertex of head black with some scattered iridescent blue scales most conspicuous in the center. Collar bright yellow. Tegulae bright yellow, finely edged with black scales on the outside and broadly edged with black on the inside edge. A patch of iridescent blue scales within the black border of the inside edge. Dorsum of the mesothorax yellow and of metathorax iridescent blue. Caudal edge of meta- thorax with a narrow line of bright yellow scales. Pleura and ventrum of thorax bright yellow. Legs yellow with small brown patches on the mesothoracic and metathoracic legs. Forecoxae immaculate bright yellow. Some brown on distal segments of all tarsi. Veins and margins of wings brown-black. Margins of wings narrow but broadening at apex of forewings (.91 mm. along vein Rs). In the forewing the Sc vein and area between discal cell and Sc vein to slightly beyond the end of the cell yellow. A small patch of yellow at base of wing between the discal cell and a line of black scales marking the anal wing fold. Another yel- low patch more densely scaled between the wing fold to slightly beyond vein 2dA and terminating distally at a small black patch which lies on the anal fold. The distal end of the small black patch is slightly more than 2 mm. from the base of the wing. Some scattered black scales on the part of vein 2dA which extends into the yellow basal patch, but the basal part of the vein yellow. Scales representing vein M2 extend within the discal cell for approximately half the length of the cell. Hindwing with creamy yellow patch at base of wing extending along anterior part of discal cell for two-thirds the length of cell. The black margin of the wing diminishing gradually and disappearing completely near the base of the wing. This part of the wing is normally hidden beneath the forewing. The creamy patch extends within the base of the cell and terminates at a small black spot. Area below discal cell hyaline. A small streak of yellow scales in brown-black anal margin at base of wing. Basal segment of abdomen including bullae bright yellow but each of the subsequent seg- ments except the terminal segment with the anterior margin bright yellow and posterior mar- gin brown-black. These latter brown-black bands expanding into triangular marks on the dorsum and obliterating the yellow on the mid-dorsum. The brown-black bands have scattered patches of iridescent blue or blue-green scales in vari- ous lights. The terminal segment of the abdomen is brown-black. Ventrum of abdomen yellow. I take pleasure in naming this species after William Beebe, Director Emeritus of the De- partment of Tropical Research, New York Zo- ological Society. Material.— AU types were taken at Simla, Arima Valley, Trinidad. Holotype, male. Cata- log No. 57207, 16 XII; 2 paratypes (57208) 15-1 and (57209) 21-1. Disposition of type material.— The Depart- ment of Tropical Research, New York Zoologi- cal Society, will retain two paratypes, Catalog 116 Zoologica: New York Zoological Society [42: 10 Nos. 57208 and 57209. The holotype, Catalog No. 57207, is in the American Museum of Natu- ral History. Chrostosoma Hubner Distinctive in this group of genera by having vein Ri of the forewing stalked with the other radial veins and not free from the cell. Chrostosoma viridipunctatum Rothschild (PI. Ill, fig. 10) Chrostosoma viridipunctatum, Rothschild, 1911: 30. Chrostosoma viridipunctatum, Rothschild, 1913; 471, pi. XIV, fig. 34. Chrostosoma viridipunctata, Hampson, 1914; 148. Chrostosoma viridipunctatum, Draudt in Seitz, 1915: 67, pi. 12k. Without an examination of the genitalia of the holotype in the British Museum (Natural History), it is impossible to be certain of the above identification. A genitalic examination of all the echemus-like species of this genus is necessary to determine the valid species and their respective ranges. Hampson’s (1898: 215) division of the genus into two sections on the basis of the presence or absence of vein Cui in the hindwing is incor- rect, as this vein is only forked nearer the margin of the wing in the first section than in the second section of the genus. Kenedy found that the holotypes of decisum and viridipunctum had both preserved vein Cui. The series of viridipunctatum in the British Museum (Natural History) contains three fe- male specimens from Trinidad, one from Caparo and two from Tabaquite. One of the latter speci- mens has metallic coloring similar to Roth- schild’s type, whereas in our series of males and females any metallic coloring if present is slight. In this same British Museum series are specimens from Argentina, Paraguay, southern Brazil, French Guiana and Surinam containing both males and females. The holotype is a female. The type locality is Yungas de la Paz, Bolivia, 1,000 meters. In our series of viridipunctatum the iridescent blue, while variable, is never pronounced. Some iridescent blue on the front and vertex of the head, collar and shoulder covers and faint traces on the subdorsum of the abdomen. Materials— specimens (5 males and 4 females). A new record for Trinidad. Guianas to Argentina and Bolivia, Leucotmemis Butler Leucotmemis differs from Chrostosoma in having vein Ri of the forewing rising from the cell rather than stalked on R2-Rs. Vein Cui of the forewing from close to angle of the cell. Leucotmemis nexa (Herrick-Schaflfer) Loemocharis nexa Herrick-Schaffer, 1854: f. 254. Leucotmemis nexa, Hampson, 1898: 224. Leucotmemis nexa, Draudt in Seitz, 1915: 70, pi. 13a. Leucotmemis nexa, Kaye & Lamont, 1927: 2. This species, reported from Trinidad by Kaye & Lamont from St. Ann’s, Palmiste and Rock Penal Road, has not been taken at Simla. It is probably improperly placed in Leucotmemis as the facies is entirely different from the re- mainder of the genus with the exception of L. insperata (Walker) . L. nexa resembles a female Phoenicoprocta vacillans or a Calonotus. It may be separated from the female vacillans by having white lateral spots on the first abdominal seg- ment. The sex of nexa has never been reported. The facies of L. insperata can be encompassed within the range of variation of the female P. vacillans. Walker described insperata from one female collected at Para, Brazil. There does not appear to be any distinctive venational dif- ference between Phoenicoprocta and Leucot- memis, though most likely the accessory cell be- tween vein R2 and vein R3+4+.'5 of the forewing in Phoenicoprocta is sufficient to separate the genera. In the only female of Leucotmemis I have seen, L. lemoulti (Rothschild), the anten- nal pectinations are equal on both sides of the antennal shaft, whereas in Phoenicoprocta and Calonotus the antennal pectinations in the fe- male are much longer on the outside of the antennal shaft than on the inside. In addition, Calonotus has vein R2 of the forewing rising from the cell as well as vein Ri. Calonotus hoff- mannsi Rothschild, which was removed to Leu- cotmemis by Hampson (1914), is a synonym of Phoenicoprocta vacillans. Rothschild similar- ly described lemoulti in Calonotus. Range.— Mexico to Brazil. Nyridela Lucas A genus containing only two species, which may be distinguished from related forms by their large size. Vein Rr of the forewing rises from the cell and Cu2 from near the angle of cell. Nyridela chalciope (Hubner) Isanthrene chalciope Hubner, 1827: 20, figs. 469, 470. 1957] Fleming: Ctenuchidae (Moths) of Trinidad, B.W.I. 117 Glaucopis acroxantha Perty, 1834: 156, pi. 31, fig. 4. Nyridela chalciope, Hampson, 1898: 218. Nyridela chalciope, Draudt in Seitz, 1915: 69, pi. 9k. Nyridela chalciope, Kaye & Lament, 1927, 2. A large hyaline-winged species with black margins and a black band running from the middle of the costal margin of the forewing to the anal angle. Antennae with yellow tips and the abdomen black with iridescent bluish-green spots. We have not collected this species at Simla but Kaye & Lamont report it from St. Ann’s Valley in the western part of the Northern Range. This species is the South American counterpart of the closely related species from Central America, Nyridela xanthocera (Walker). Range— Colombia and Brazil. CosMOSOMA Hubner A large, somewhat confused, genus that needs revision. Many of the species that Hampson (1898) originally included he subsequently (1914) placed in the genus Gymnelia which has a lobed hindwing. Eight species are recorded from Trinidad, of which we record six species. 1 . Discal bar broad 2 Discal bar narrow 4 2. Tegulae and patagia bright red rubriscapulae Tegulae and patagia brown or blue. ... 3 3. Abdomen with scarlet subdorsal streaks on first segment melathoracia Abdomen without scarlet subdorsal streaks on first segment subflamma 4. Antennae with white tips 5 Antennae without white tips 7 5. Costal margin and terminal band of fore- wing immaculate black anoxanthia Patches of orange-red along costal margin and terminal band at tornus of fore- wing 6 6. Abdomen with a dorsal red stripe . achemon Abdomen without a dorsal red stripe remotum 7. Abdomen concolorous orange-yellow klagesi Abdomen yellow, second and third seg- ments black with lateral blue spots . pytna Cosmosoma rubriscapulae Kaye (PI. Ill, fig. 11.) Cosmosoma rubriscapulae Kaye, 1901: 116. Cosmosoma rubriscapulae, Hampson, 1914: 156. Cosmosoma rubriscapulae, Draudt in Seitz, 1915: 71, fig. 13b. Cosmosoma rubriscapulae, Kaye & Lamont, 1924: 3. The large size combined with the scarlet pa- tagia and tegulae renders this moth distinctive in Trinidad from other Cosmosoma species. We have collected three males and three females of this species at Simla. In our specimens the discal bar of the forewings of the females is broader and heavier than in the males. Range.— Reported only from Trinidad. Cosmosoma subflamma subflamma (Walker) (PI. Ill, fig. 12) Glaucopis subflamma Walker, 1854: 159. Laemocharis panopes Herrich-Schaffer, 1854: 243. Cosmosoma subflamma, Hampson, 1898: 227. Cosmosoma siibflammum, Draudt in Seitz, 1915: 71, fig. 13b. Cosmosoma subflamma, Kaye & Lamont, 1924: 3. The bright red legs distinguish this species from other Cosmosoma species in Trinidad. Schaus described a species, lucia, from St. Lucia, B.W.I., which Hampson (1898) records as a subspecies of subflamma in which the legs are largely black. Material.— Wq have taken only one male. An- other specimen was taken by F. W. Urich in 1918 in the Arima Valley and Lamont captured a specimen on May 28, 1921, at Palmiste. Range.— St. Lucia to south Brazil on the east- ern side of South America. Cosmosoma melathoracia Kaye (PI. Ill, fig. 13) Cosmosoma melathoracia Kaye, 1901: 115, pi. V, fig. 10. Cosmosoma melathoracia,'Hav!ypson, 1914: 157. Cosmosoma melathoracia, Draudt in Seitz, 1915: 73, fig. 13e. Cosmosoma melathoracia, Kaye & Lamont, 1924: 3. Male with dorsal line of iridescent blue scales and a brown valve at base of abdomen edged laterally with white. Material.— Six males. Range.— Described and reported only from Trinidad. 118 Zoologica: New York Zoological Society [42: 10 Cosmosoma anoxanthia Druce (PI. Ill, fig. 14) Cosmosoma anoxanthia Druce, 1905: 460. Cosmosoma bolivarensis Klages, 1906: 536. Cosmosoma anoxanthia, Hampson, 1914: 165, pi. VIII, fig. 25. Cosmosoma anoxanthium, Draudt in Seitz, 1914: 80, fig. 27m. Cosmosoma achemon, f. bolivarensis, Draudt in Seitz, 1914: 79. Cosmosoma anoxanthia, Kaye & Lament, 1924: 3. Female similar to male except that on the forewing the oblique quadrate patch of blackish- brown scales from the lower extremity of the discal cell to tornus (between veins Cui and Cuo) is absent in the female. Material —Six males and three females. Trinidad, Venezuela and Bolivia. Cosmosoma achemon (Fabricius) (PI. Ill, fig. 15) Zygaena achemon Fabricius, 1781: 162. Euchromia tyrrhene Hubner, 1827: 23, figs. 483-484. Cosmosoma voltumna Druce, 1897: 303. Cosmosoma achemon, Hampson, 1898: 247. Cosmosoma achemon, ab. tyrrhene, Hampson, 1914: 527. Cosmosoma perfenestratum Dyar, 1899; 175. Cosmosoma achemon, Draudt in Seitz, 1914: 79, fig. 14c. Cosmosoma perfenestratum, Draudt in Seitz, 1914: 80, 1917: 201. The very bright red dorsal abdominal stripe will separate this species from all other Trinidad Cosmosoma. A new record for Trinidad. Material— One male from St. Augustine, Trinidad. Range.— Jamaica, Haiti, Venezuela and Brazil. Cosmosoma remota (Walker) Glaucopis remota Walker, 1854: 170. Cosmosoma remotum, Hampson, 1898: 248, pi. IX, fig. 22. Cosmosoma remotum, Draudt in Seitz, 1914: 80, fig. 14c. Cosmosoma remota, Kaye & Lamont, 1924: 3. Reported by Kaye & Lamont from Tobago. This species is very close to achemon, from which it may be distinguished by the absence of the red abdominal dorsal stripe. We have not taken this species in the Arima Valley but Kaye has specimens from Trinidad in his collection. Range.— Trinidad, Tobago and Venezuela. Cosmosoma klagesi Rothschild (PI. Ill, fig. 16) Cosmosoma klagesi Rothschild, 1910: 509. Cosmosoma klagesi, Rothschild, 1913: 470, pi. XIII, fig. 11. Cosmosoma clagesi, Hampson, 1914: 170. Cosmosoma klagesi, Draudt in Seitz, 1914: 82, fig. 14g. Cosmosoma klagesi, Kaye & Lamont, 1924: 4. Mater/a/.— Fifty-one males and twenty fe- males. Our commonest Cosmosoma in the Ari- ma Valley. Range.— Trinidad and Brazil. Cosmosoma pytna Druce Cosmosoma pytna Druce, 1906: 78. Cosmosoma pytna, Hampson, 1914: 157, pi. VIII, fig. 12. Cosmosoma pytna, Draudt in Seitz, 1914: 83, fig. 27m. Cosmosoma pytna, Kaye & Lamont, 1924: 4. We have not collected this species in the Arima Valley nor, insofar as the literature reveals, has it been collected since the holotype was taken in Trinidad. Dixophlebia Butler The fringes of hair along the lower radial, median and cubital veins of the forewing will distinguish this genus from other genera of Ctenuchidae. Dixophlebia holophaea Hampson (PI. Ill, fig. 17) Dixophlebia holophaea Hampson, 1909: 346. Dixophlebia holophaea, Hampson, 1914: 174, pi. IX, fig. 9. Dixophlebia holophaea, Draudt in Seitz, 1914: 85, fig. 14k. A grayish-black moth with hyaline in the dis- cal cell areas in both wings. Material— Two males. Raaga.— Surinam and British Guiana. A new record for Trinidad. PsEUDOMYA Hubner A genus probably derived from Saurita, from which Pseudomya may be distinguished by hav- ing the hind tarsal joints fringed with scales in the males. 1957] Fleming: Ctenuchidae (Moths) of Trinidad, B.W.l. 119 Pseudomya melanthus (Stoll) (PI. Ill, figs. 18, 19) Sphinx melanthus Stoll, 1782: pi. 367, C. Pseudomya melanthus, Hampson not Stoll, 1898: 264. (/ngM/fa Walker). Pseudomya melanthus, Hampson, 1914: 175. Pseudomya melanthus, Draudt in Seitz, not Stoll, 1914: 87, fig. 14m. (trigutta Walker). Pseudomya melanthus, Draudt in Seitz, 1917: 202. Pseudomya melanthus, Kaye & Lamont, 1927: 4. The facies of this species is very similar to Pseudomya sanguiceps Hampson from Panama (Hampson, 1898: 264, pi. X, fig. 4) figured in Seitz on line 14m. The medial black band of the forewings is darker and wider than shown in the Seitz figure and extends from the costal to the inner margin in the male. Our single female, while having a very distinct quadrate patch of dark scales at the discal veins, has only a scat- tering of scales below the cell. Length of fore- wing of the male 10 mm. and of the female 11 mm. Material— One. male and one female. Trinidad and Surinam. Rhynchopyga Felder Differs from Saurita and Pseudomya in hav- ing Cui and Ms on a long stalk in the hindwing. Rhynchopyga flavicollis (Druce) (PI. Ill, fig. 20) Amycles flavicollis Dmce, 1884: 46, pi. 7, fig. 11. Rhynchopyga flavicollis, Hampson, 1898: 270. Rhynchopyga flavicollis, Draudt in Seitz, 1914: 90, fig. 15e. The long, very narrow brown wings of this slender, small species is distinctive. Our single male specimen lacks the orange streak on the patagia present on the holotype. Our specimen also has less orange on the tegulae. Material.— One male. A new record for Trini- dad. Guatemala, British Honduras, Costa Rica, Panama and Colombia. Saurita Herrick-Schaffer Probably not a natural genus but attempts to divide it have thus far been unsuccessful. Veins Ms and Cui of the hindwing arise from the same point or are very shortly stalked. 1 . Abdomen of male with ventral valve cover- ing basal segments. Abdomen of male and female large, with iridescent blue sublateral spots and crimson anal tufts .... cassandra Abdomen of male with no ventral valve. Abdomen of male and female other- wise 2 2. Thorax with crimson patches only 3 Thorax orange red or crimson 6 3. Abdomen with whitish patches 4 Abdomen with brown or black patches . . 5 4. Discal cell of forewing brown-scaled lacteata Discall cell of forewing hyaline . . orimensis 5. Forewing uniformly hyaline or smoky hyaline perspicua Forewing with post-discal milky hyaline band clusia 6. Vertex of head crimson red salta Vertex of head black 7 7. Wings fully scaled with only slight thin- ning of scales between veins concisa Discal cell and area below discal cell of forewing hyaline 8 8. Underside of body yellow te menus Underside of body black afflicta Saurita cassandra (Linnaeus) (PL III, fig. 21) Sphinx cassandra Linnaeus, 1758: 494. Saurita cassandra, Hampson, 1898: 274, fig. 111. Saurita cassandra, Draudt in Seitz, 1915: 93, fig. 15f. Saurita cassandra, Kaye & Lamont, 1927: 4. A robust, brown-winged Saurita with the dis- coidal patch darker and extending somewhat along the cubital veins. Males and females simi- lar except for the longer pectinations on the male antennae. Material.— One male. Range.— Venezuela to Argentina. Saurita clusia (Druce) (PI. Ill, figs. 22, 23) Laemocharis clusia Druce, 1897: 303. Hypocharis clusia, Hampson, 1898: 271, fig. 125. Hypocharis clusia, Draudt in Seitz, 1915: 91, fig. 15e. Hypocharis clusia, Kaye & Lamont, 1927: 4. Saurita clusia, Forbes, 1939: 121. Forbes (1939: 121) placed clusia in Saurita, pointing out that the character Hampson (1898: 271) employed to erect the genus Hypocharis is equally true of Saurita, namely, vein Mi of 120 Zoologica: New York Zoological Society [42: 10 the forewing is more or less from below the angle of the cell in all species of Saurita as well as Hampson’s Hypocharis. This species will key to Saurita nox (Druce) in Hampson’s key (1898; 275). Hampson’s footnote indicates that he examined the holo- type of nox which was in the Staudinger collec- tion. In his description of the species he men- tions “a broad diffused whitish band from just beyond middle to termen.” He gives the wing expanse as 22 mm. Druce, on the other hand, neither in his orig- inal description (1896: 30) nor in the Biologia (1897: 341, pi. 71, fig. 11) where he copies his original description, makes any mention of a whitish postmedian band but states simply, “primaries and secondaries smoky hyaline, with the veins all black.” His figure (/.c.) illustrates a smoky hyaline-winged moth with blackish veins. Druce records the wing expanse as 114 inches. Draudt in Seitz (1915: 93, fig. 15g) shows a figure resembling the figure of Druce. Forbes conjectures that Draudt’s figure is of S. fumosa. In his description, however, Draudt writes of a “postdiscal, faded, whitish spot.” He gives the wing expanse as 22 mm. He compares the species to Saurita lacteata Butler which has a different-shaped inner margin of the hindwing. (See Forbes, l.c. for grouping of species of Saurita) . Hampson (1914: 186) synonymizes Chrosto- sonia maratha Druce as the female of Saurita nox. The holotype is in poor condition with the wings torn and rubbed and the abdomen miss- ing. However, the whitish postdiscal band crosses the forewing in the photograph slightly distad of the discoidal veins. The facies of the forewing looks very similar to that of clusia. Specimens from Trinidad are included within the series of Saurita clusia in the British Museum (Natural History) collection. These Trinidad specimens and also those in Kaye’s private collection under clusia are similar to the speci- mens we collected. However, if I were to follow my interpretation of Forbes (1939: 121) I should have identified them as nox. Forbes states that clusia is smaller than nox, the white post- medial band narrower sex for sex, and the blue spotting is distinctive, and for nox, “There are no blue spots; in the male the white area invades the outer third or half the cell, in the female the extreme apex of the cell may be pale.” First of all, our series of specimens varies from having a distinctively blue spotted abdomen to a com- pletely unspotted concolorously colored ab- domen with the completely unspotted section representing approximately half the collection. Our specimens are much larger than the speci- mens in the American Museum of Natural His- tory from Panama identified as clusia by Forbes. All of our specimens have the postmedial band invading the discal cell to the extent of half or more. Lastly, our specimens differ in genitalia from those in the American Museum under clusia. Hampson (1898: 271) mentions the blue spotting in regard to clusia. His figure of clusia is similar to our photograph of the holotype of clusia, but his wing expanse is much larger (30 mm.) (and Druce 114") than the six specimens named clusia by Forbes (23-25 mm.) in the American Museum of Natural History. In other than the size difference, Hampson and Forbes agree. Aside from the discrepancy between Druce’s original description of nox and the subsequent descriptions by other authors which make the identity of nox questionable, it is quite possible that we are concerned with more than two species. However, I consider that it would be unwise to describe the Trinidad form without having material throughout the range of the above two species. Saurita lacteata (Butler) Dycladia lacteata Butler, 1877: 34, pi. 17, fig. 3. Saurita lacteata, Hampson, 1898: 276, fig. 128. Saurita lacteata, Draudt in Seitz, 1915: 93, fig. 15g. Saurita lacteata, Kaye & Lamont, 1927: 4. This species is listed by Kaye & Lamont as occurring in Trinidad but the specimens in Kaye’s collection under this name are similar to the following new species. The specimens under lacteata in the British Museum (Natural History) from Trinidad are likewise in agree- ment with the following species. Butler describes lacteata from the Rio Jutahi, Amazons. Saurita orimensis, new species (PI. I, fig. 3; PI. II, fig. 3) Length of forewing of male 10.5-11.5 mm. Related to Saurita lacteata Butler from which arimensis differs in having the discal cell hyaline and in not having a yeUowish-white patch below the cell and another yellowish-white patch be- yond the cell. 'Antennae bipectinate in the male with each pectination dilated and bristled on the distal end. Palpi brown, reaching vertex of head with fan-shaped brown tuft on first segment. Whole head uniform brown. 1957] Fleming: Ctenuchidae (Moths) of Trinidad, B.W.I. 121 Patagia and disc of thorax uniform brown with a small white patch on the mid-dorsum of the metathorax. Tegulae brown with a crimson patch on anterior margin near patagia (shoul- ders) and a white bar at discal edge of tegulae on the anterior margin of forewing. The inner half of tegulae along the disc of the thorax crimson. Some white on the outer edge of the basal half of the forecoxae on the episternum (2nd). The remainder of the legs and the lateral and ventral parts of the thorax brown. Fore- and hindwings hyaline. Veins brown. The forewings narrowly bordered with brown but the apex of the wings broadly brown. The brown patch at the apex extending within the wing to the point where veins R4 and R5 fork. A relatively broad brown discal bar (1.25 mm.) . Hindwings with narrow brown margins broaden- ing at apex. At the 1st anal fold the brown color extends within the wing for approximately one- third of the length of the wing from the margin, and a brown patch extending into the wing at the anal angle for approximately the same distance. Abdomen with the dorsum of the first and second segments and to a variable extent the third segment creamy white. The bullae creamy white. A creamy white band extends from bullae along the spiracular region of the abdomen for four segments. The band is broadest at the base and narrows toward the distal end of the abdomen. The remainder of the abdomen brown. The name arimensis has been taken from the type locality, Arima Valley. Material— All types were taken at Simla, Ari- ma Valley, Trinidad. Holotype, male, Catalog No. 57200, 2-V; 6 paratypes, (57201) 2-VI, (57202) 16-III, (57203) 21-IV, (57204) 13-11, (57205) 7-1, (57206) 3-VI. Disposition of type material— The Depart- ment of Tropical Research, New York Zoologi- cal Society, will retain two paratypes. Catalog Nos. 57205 and 57206. Paratype, Catalog No. 57204, is in the collection of the British Museum (Natural History) and paratype. Catalog No. 57203, is in the United States National Museum collection. The holotype, Catalog No. 57200, and the paratypes. Catalog Nos. 57201 and 57202, are in the American Museum of Natural History. Saurita perspicua Schaus Saurita perspicua Schdios, 1905: 187. Saurita perspicua, Hampson, 1914: 185, pi. IX, fig. 31. Saurita perspicua, Draudt in Seitz, 1915: 93, fig. 28e. Saurita perspicua, Kaye & Lamont, 1927: 4. This species has not been collected since the original type. Subsequent mention in the litera- ture has been based on the original description. The type is a female in the U.S. National Mu- seum. Schaus (1905) states that the wings are smoky hyaline and Hampson (1914) simply says that they are hyaline. We have not taken this species at Simla. Trinidad. Saurita salta (Schaus) Thrinacia salta Schaus, 1894: 226. Saurita salta, Hampson, 1898: 277, pi. X, fig. 12. Saurita salta, Draudt in Seitz, 1915 : 94, fig. 15h. Saurita salta, Kaye & Lamont, 1927: 5. Described from Venezuela. Kaye & Lamont report it from Trinidad, with no data. Saurita temenus (Stoll) Sphinx temenus Stoll, 1781: pi. 367, D. Saurita temenus, Hampson, 1898: 279. Saurita temenus, Draudt in Seitz, 1915: 94, fig. 16b. Saurita temenus, Kaye & Lamont, 1927: 5. See following species, Saurita afflicta, for dis- cussion. Rnn^e.— Surinam, Amazons. Saurita afflicta (Walker) (PI. Ill, figs. 24, 25) Glaucopis (Pseudomya) afflicta Walker, 1854: 144. Glaucopis afflicta, Butler, 1877 : 29, pi. 7, fig. 12. Saurita temenus, Hampson, 1898: 279 (in part). Saurita venezuelensis Klages, 1906: 538. Saurita venezuelensis, Draudt in Seitz, 1917: 94, fig. 15h. Saurita temenus, Draudt in Seitz, 1917: 94, fig. 16b (in part). Saurita afflicta, Forbes, 1939: 123. The Saurita temenus record of Kaye & La- mont (1927: 5) may be in error and their specimens rightfully placed in this species. The specimens we have taken at Simla are S. afflicta. This species was listed by Hampson (1898: 279) as a synonym of S. temenus and subsequent authors until Forbes (1939: 123) followed Hampson. Forbes was the first to point out that the two names represented different species most easily separated by the yeUow ventrum of S. temenus and the blackish ventrum of S. afflicta. 122 Zoologica: New York Zoological Society [42: 10 However, since S. afflicta was described from the Amazons and S. temenus from Surinam, it is possible that both species inhabit Trinidad. The females may be distinguished from the males by having the spaces between the veins beyond the cell hyaline. Material —11 specimens: 12 males and five females. Honduras to the Amazons. Saurita concisa (Walker) Euchromia concisa Walker, 1854: 243. Thrinacia afflicta L)T\ice, 1884: 56 (not Walker). Saurita concisa, Hampson, 1898: 279. Saurita thoracica Klages, 1906: 538. Saurita concisa, Hampson, 1914: 189. Saurita thoracica, Draudt in Seitz, 1915: 94. Saurita concisa, Kaye & Lamont, 1927: 5. Kaye & Lamont placed Saurita venezuelensis Klages as a synonym, but I have followed Forbes and synonymized venezuelensis under Saurita afflicta Walker. Kaye & Lamont record this species from Palmiste, Trinidad, but we have not collected it as yet at Simla. Panama to the Amazons. PsoLOPTERA Butler A small genus closely related to Saurita, from which it differs by having vein Ri forked with R2 in the forewing. Psoloptera leucosticta (Hubner) Glaucopis leucosticta Hubner, 1827: t. 162. Psoloptera leucosticta, Hampson, 1898: 285. Psoloptera leucosticta, Draudt in Seitz, 1915: 96, fig. 15 m. Psoloptera leucosticta, Kaye & Lamont, 1927: 5. Kaye & Lamont report this species from Trinidad, one specimen from Guaico at the southern foot of the Northern Range and two specimens from Palmiste. We have not taken it at Simla, but we collected the species at Cari- pito, Venezuela, across the Gulf of Paria from Trinidad. The wings and body are purplish- black with two white points at the base of the abdomen. Range.— Venezuela, Trinidad, Guianas and Amazons. Dycladia Felder A small genus of moths that are very beetle- like. The following species in particular, like the species of the genus Correbidia in the next sub- family, resemble a Lycidae beetle. 1. Orange lateral line of the abdomen not reaching beyond middle of the abdomen and no black spot at the base of the fore- wing correbioides Orange lateral line of the abdomen absent only on the last segment and base of fore- wing with large black spot. . . basimacula Dycladia correbioides Felder Dycladia correbioides Felder, 1874: pi. 102, fig. 20. Dycladia correbioides, Hampson, 1898: 293, fig. 139. Dycladia correbioides, Draudt in Seitz, 1915: 99, fig. 16d. Dycladia correbioides, Kaye & Lamont, 1927 : 5. Kaye & Lamont record this species from Pal- miste but see following species, Dycladia basi- macula Schaus. Dycladia basimacula Schaus (PI. m, fig. 26) Dycladia basimacula Schaus, 1920: 9. It is most likely that the D. correbioides of Kaye & Lamont is this species. D. correbioides is the name used for the form found in Colombia and Panama, and D. emerita for the form from Costa Rica to Mexico. Schaus described D. basi- macula from Venezuelan and Trinidad material. Material— Vova males. Range.— Trinidad and Venezuela. Syntomeida Harris The three posterior veins from the discal cell of the hindwing are forked. Syntomeida melanthus (Cramer) Sphinx melanthus Cramer, 1779: pi. 248, C. Sphinx nycteus Stoll, 1780: pi. 325, F. Euchromia apricans Walker, 1854: 224. Syntomeida albifasciata Butler, 1876: 366. Syntomeida melanthus, Hampson, 1898: 306, fig. 138. The lustrous blue-black wings with a variable number of quadrate yeUow spots and orange- red and black bands on the abdomen separate this insect from other Trinidad ctenuchids. The names albifasciata, nexilis and nycteus apply to variants in either the number of yellow wing spots or abdominal coloration. Kaye & Lamont report this species from San Fernando. 1957] Fleming: Ctenuchidae (Moths) of Trinidad, B.W.l. 123 Mexico to Uruguay and Peru. Histiaea Walker The largest of the Trinidad ctenuchids, with densely scaled, brown wings variously marked with red and yellow or grayish spots. 1. Abdomen with two basal segments com- pletely yellow dorsally. Forewing with ex- tensive longitudinal red rays running from base meldolae Dorsum of abdomen brown with two sub- dorsal spots on basal segment, and triangu- lar, subdorsal yellow spots on the anterior parts of the following two segments. Fore- wing with at most red or reddish-tinged spots or small streaks cepheus Histiaea meldolae Butler (PI. Ill, fig. 27) Histiaea meldolae Butler, 1876: 362. Histiaea meldolae, Druce, 1884: 42, pi. 6, fig. 14. Histiaea meldolae, Hampson, 1898: 311. Histiaea meldolae, Draudt in Seitz, 1915: 101, fig. 16g. Histiaea meldolae, Kaye & Lamont, 1927: 5. Material —ThrQC males and three females. Range— Ths holotype is a Trinidad specimen. British Guiana, Venezuela, and Panama. Histiaea cepheus (Cramer) (PI. Ill, fig. 28) Histiaea cepheus Cramer, 1780: pi. 109, E. Histiaea cepheus, Hampson, 1898: 313. Histiaea monticola Klages, 1906: 538. Histiaea monticola, Hampson, 1914: 203. Histiaea cepheus, Draudt in Seitz, 1915: 101, fig. 16g. Histiaea monticola, Draudt in Seitz, 1915: 102. Histiaea cepheus, Kaye & Lamont, 1927: 6. Ma/ma/.— Forty-six males and 17 females. TJange.— Venezuela to Surinam. Macrocneme Hubner This genus of iridescent wasp-like moths is in a very confused condition. The iridescence of the wings, while diagnostic in many instances, is extremely difficult to describe. Forbes (1939) has written the only paper that is of any real assistance. The descriptions in Hampson (1898 and 1914) are valuable but his taxonomic treatment is inadequate. Genitalic studies of the types and series of specimens must be made and associated with other characters before identification in this genus will have any relia- bility. The following key is based on specimens we have collected and some unidentified Trinidad material loaned by the British Museum (Natu- ral History). 1. First segment of abdomen with four white spots; two subdorsal and two lateral. No ventral valve in male. Large species with a wing length of 16 mm. or more 2 First segment of abdomen with only two iridescent blue or green subdorsal spots on conspicuous abdominal bullae. Ven- tral valve present in male. Smallest species with a wing length of 14 mm. or less 6 2. Males 3 Females 5 3. At least the basal 2/3 of the ventrum of the abdomen with a broad uninterrupted white band thyra Otherwise; white of underside of the ab- domen broken up into spots 4 4. Forecoxae iridescent blue species? Forecoxae white spinivalva 5. Iridescence of the forewing uniform green to end of discal ceU except slight black along base of costal margin and small black point on the base of the anal vein. (True of males also) thyra Iridescence of forewing interrupted by a black band from inner margin to at least the anterior part of discal cell. (True of males also) spinivolva 6. Forewing with streaks of iridescent blue or green from base of wing to end of discal cell vittata Forewing with only blue or green spots at base of wing 7 7. Abdomen with sub ventral white spots on two medial segments plumbea Abdomen without subventral white spots albitarsia Macrocneme plumbea (Hampson), new combination (PI. Ill, fig. 29) Poliopastea plumbea Hampson, 1898: 337, pi. XII, fig. 26. Poliopastea plumbea, Draudt in Seitz, 1915: 110, fig. 18b. Our specimens agree with Hampson’s original description of plumbea but the description is inadequate for proper determination. Hampson states that the abdominal valve is blue edged 124 Zoologica: New York Zoological Society [42: 10 with white, with a white patch behind it. Our specimens have these characters, but in addition the two segments caudad of the white patch have small subventral white spots. Draudt in Seitz may have had similar specimens, as he states: “—the next rings indistinct, white sublateral spots.” Kenedy, in notes, writes that the type has small, sublateral spots on the next segment and sublateral metallic blue spots on next two segments. One of our specimens shows only faint traces of the second pair of spots. AH of our specimens have subventral blue spots on the sub- terminal abdommal segments but they may be very faint. The hind legs are missing on the holo- type, but in the series of specimens of plumbea (not type material) in the British Museum (Natural History) the hind tarsi are white. Hampson’s male holotype came from the lower Amazons, Paratins. Draudt records the species from the Amazon and French Guiana, and specimens in the British Museum (Natural His- tory) collection are from Trinidad, Venezuela, British Guiana and Sao Paulo, Brazil. My identi- fication of the Trinidad material is provisional on a genitalic examination of the holotype in the British Museum. Hampson erected a new genus, Poliopastea, for plumbea and made plumbea the type species of the genus. Hampson’s conception of the dif- ferences between Macrocneme and Poliopastea appears to be the porrect palpi in Poliopastea, veins Cui, Ms and Ms of the forewing close to the angle of the cell and vein Rs of the fore- wing from the cell. The specimens that I have seen from Trinidad have individuals with both porrect and upturned palpi. Vein Cui of the forewing is variable in Macrocneme so that the character, veins Cui, Ms and Ms from close to angle of cell, has no significance and some spe- cies of Macrocneme have vein Rs of the fore- wing from the cell. Consequently, I consider Poliopastea a synonym of Macrocneme. I have not examined any of the other species that have been placed in Poliopastea but it seems likely that obscura (Wallengren) and viridis (Druce) can be included in Macrocneme. Hampson (1914: 207) placed ochendeni Rothschild in Macrocneme. He erected a new genus Pseudo- phaio for rosenbergi Rothschild and provision- ally transferred verdivittata to Calonotus. Polio- pastea pava (Dognin) was aberrant in the genus and required a separate section. It was originally described by Dognin in the genus Thysanopry- ana, which is a synonym of Baritius in the family Arctiidae. Kaye & Lamont listed plumbea as a synonym of eacus along with vittata and nigritarsia. The figure of eacus in Stoll (1781, fig. 335 C) has evanescent, iridescent blue in the ceU and some- what beyond, and vittata (see above under vittata) has iridescent blue above and below the cell as well as within the cell. Hampson’s nigri- tarsia is a larger moth than eacus and the iri- descent blue is more distinct and extensive. Lamont & Callan (1950: 197) report albitarsia from Trinidad (Palmiste) . This species may be separated from plumbea by the absence of the white spots on the ventrum of the abdomen. Both eacus and nigritarsia have black hind tarsi, whereas albitarsia, plumbea and vittata have the terminal segments of the hind tarsi white. Material.— Four males from Simla and two males from the British Museum (Natural His- | tory) from Caparo and San Fernando, Trinidad. j Macrocneme thyra thyra Moschler Macrocneme thyra Moschler, 1883: 334, 1. 18, ij fig. 24. 1 Macrocneme thyra, Hampson, 1898, 321. \ Macrocneme thyra, Draudt in Seitz, 1915, 103, fig. 17a. Macrocneme albiv enter Dognin, 1923 : 2. Macrocneme thyra, Kaye & Lamont, 1927 : 6. Macrocneme thyra, Forbes, 1939: 129, 1. 1, fig. 3 (genitalia). This is the nomenclatural type that was de- scribed from Surinam. One male specimen from the collection in the British Museum (Natural History) I have tentatively assigned to this form. | It was collected at Tabaquite, Nariva District, ; Trinidad. This single specimen has a very short streak of iridescent blue at the base of the wing ! in the lower part of the cell and just below the cell. Iridescent blue in the distal half of the cell, just above the cell and belov/ the anal vein, the | latter extending almost to the base of the wing. I The area between the cell and the anal vein black | but with a trace of iridescence below the irides- ; cence in the cell end. This specimen has more extensive black in the basal part of the wing than Hampson indicates for thyra. The ventrum of the abdomen is white. While I have not re- moved the genitalia, it is well extruded and I have been unable to see any significant difference between the genitalia of this specimen and the genitalia of thyra that Forbes (1939) figures, and the following race. Range.— Guianas, Brazil, Peru to Panama? Macrocneme thyra intacta Draudt (PI. Ill, fig. 31) Macrocneme thyra, subspecies 1, Hampson, 1898: 321. Macrocneme thyra intacta, Draudt in Seitz, 1915: 103. 1957] Fleming: Ctenuchidae {Moths) of Trinidad, B.W.l. 125 I am unable to find any essential details of the genitalia of the Trinidad specimens different from the genitalic figures of Forbes (1939). Both processes of the male valves are curved as in the figure and the upper process of the left valve has a distinct tooth on the inner edge. This last feature is not shown in the figure but stated in the text. The juxta is squarely cut off and short. The uncus is the same shape as in the figure, but with small, narrow, lateral-winged margins which I do not see in the figure. This is subspecies 1 of Hampson. Draudt ap- pears to have given the subspecies a name on the basis of Hampson’s description without having specimens before him. Hampson (1898) char- acterized it in part by stating “Forewing with blue-green at base entire.” This is true of our Trinidad specimens except as noted in the key; namely, the Trinidad specimens have black on the base of the costal margin and a small black streak or spot on or near the base of the anal vein. This last-mentioned black streak or spot is never very conspicuous and is occasionally absent. Material— lAA specimens from Simla (137 males and 7 females) and 9 specimens (6 males and 3 females) from Port of Spain, Guaico, Ariapite Valley and St. Ann’s borrowed from the British Museum (Natural History). Stated by Hampson as Trinidad and Colombia. Macrocneme spinivalva, new species (PI. I, fig. 4; PI. II, fig. 4) Length of forewing of male 17 mm., of female 18 mm. Antennae dark brown, bipectinate. Each pectination tufted on distal end and length of pectinations rapidly decreasing at proximal and distal ends of antennal shaft. Pectinations shorter in male than in female. Palpi upturned to vertex of head. First seg- ment of palpi clothed with ragged, dark brown scales with an immaculate white medial tuft of broad scales on anterior face, for two-thirds to three-quarters the length of the second segment of the palpi from the proximal end. The line of white scales absent in female. Ground color of second and third palpal segments dark brown in both sexes. Front of head dark brown with two white spots on upper edge beneath the antennal scapes in both sexes. Vertex of head brown. Patagia (coUar) with a pair of dorsal white spots and a pair of subdorsal white spots. In the male, but not the female, each dorsal spot joined to its respective subdorsal spot by a fee line of white scales on the anterior margin of the patagia. Immediately below the subdorsal spots and hardly separated from them, anterior to the bases of the tegulae and approximately in line with the middle of the eyes, a small white spot. A rectilinear episternal white spot in line with the lower part of the eyes and base of palpi. The episternal spot is present in the female though usually smaller than in the male, but the spot below the subdorsal spot is absent in the female. Tegulae dark brown with iridescent green reflections most pronounced on the an- terior edge. Dorsum of thorax dark brown with iridescent reflections in various lights, most dis- tinct on the metathorax. Forecoxae of the male white but the re- mainder of the forelegs brown with faint irides- cent reflections except for the caudal edge of the femur of the forelegs which is gray. Each of the forecoxae of the female with two small white spots on the proximal end, one laterad and one dorsad. The remainder of the female legs brown with varying iridescent reflections. Mesothoracic legs in both sexes brown with some blue reflec- tions, with a small white spot on the distal end of the femur. Metathoracic legs of the same brown color with reflections but with the distal one-half to two-thirds of the tibia in both sexes fringed. The metatarsus with long fringe. The fringe on the third, fourth and fifth tarsal seg- ments white. Forewing with two white spots on base of wing near tegulae. Distinct blue iridescence in the forewing is confined to the following five areas : (1) A streak the length of the discal cell above the discal cell, (2) to the distal half of the discal cell, (3) to the same respective part of the wing below the discal cell except that the iridescence foUows vein Cu2 in the direction of the tornus a little beyond the end of the discal cell, (4) a basal patch below the base of the discal cell, and (5) a median streak below the anal vein. The remainder of the wings including the veins through the iridescent areas blue to brownish- black depending on the age of the specimen. The above description of the forewing pattern may be summarized by saying that the iridescence is confined to the plane of the end of the discal cel! interrupted by a dark transverse band perpen- dicular to a dark anal streak. Underside of fore- wing with the basal half of the wing to cell end iridescent blue. Upperside of hindwing bluish or brownish- black depending on the age of the specimen. Underside of hindwing iridescent blue except 126 Zoologica: New York Zoological Society [42: 10 for apical and anal area. In old specimens the iridescence may be restricted to the cell and costal area. Abdomen with two pairs of white spots, one pair subdorsal and the other on the hoods or bullae. First segment of abdomen blackish- brown and the remainder of the abdominal seg- ments iridescent blue-green with bluish-black subdorsal longitudinal bands. The background color may become quite brassy in old specimens. Ventrum of male without a basal ventral valve but a pair of subventral white spots on the first segment in both sexes. Subsequent segments of the abdomen in both sexes with a midventral white spot except the last abdominal segment which, like the first segment, has a pair of white subventral spots. Male genitalia with base of uncus broadly in- flated and the edges thin in cross-section with the distal process normally directed ventrally and slightly spatulate. Valves large with two processes, the dorsal process curved with a dis- tinct thorn-like structure on inner edge of the distal half and a ventral process which is slender and hirsute. Juxta very slender, so slender as to be remarkable for the genus. Female genitalia with signa of bursa copula- trix spherical, with long spines, sea-urchinlike, with small anterior portion unspined. In the female genitalia of thyra, the other species of this section of the genus in Trinidad, the signa is ovoid or almost subquadrangular with short spines. This is most likely the species mentioned by Forbes ( 1939: 126, pi. II, fig. 8) which he refers to as Macrocneme species. It will run to this species in his key to the species of Macrocneme and his figures are extremely similar. The tegu- men and uncus are less inflated laterally than in spinivalva and do not show any part of the distal end of the uncus. The thorn or tooth on the dorsal process of the valve appears more pronounced and acuminate than in Forbes’s figure of the valve, but this may be a matter of perspective. This species seems to be most closely related to thyridia with which we had it confused in our collection. The tooth or thorn-like structure on the upper process of the valve and extremely slender juxta will separate spinivalva from thyridia. The spined condition of the valve can often be seen without removing the genitalia in the male. Thus far, we have not found thyridia in Trinidad. The specific name spinivalva, meaning thorn- valve, refers to the thorn-like structure on the upper process of the male genitalic valve. Material.— male. Catalog No. 5716, Simla, Arima Valley, 27-III; allotype, female. Catalog No. 5717, Simla, 26-III; paratypes, male and female in coitu, (5718) Simla, 22-XII; paratypes, males, (5719) Simla, 16-11— (5720) #288, Trinidad, B.M.-(5721) June, 1902, Aria- pite Valley, B. M.— (5722) June, 1902, Ariapite Valley, Trinidad, B. M.; paratypes, females, (5345) Simla, 17-III-(5724) Simla, 6-III- (5725) Simla, 27-III-(5726) Simla, 13-III- (5727) Trinidad, B. M.-(5728) Trinidad, B. M. —(5729) Trinidad, Dr. Jackson, 1921-1922, B. M.-(5730) Trinidad, A. HaU, Feb. 1930, B. M.— (5731) Ariapite Valley, July, 1902— (5732) Trinidad, B. M.— (5733) Trinidad, B. M.— (5734) Trinidad, F. Shade, Nov. 1920, B. M.-(5735) Trinidad, B. M.-(5736) Trini- dad, B. M. Disposition of type material.— The Depart- ment of Tropical Research, New York Zoo- logical Society, retains the two paratypes in coitu (5718). Those paratypes listed above with abbreviation “B.M.” are specimens borrowed from the British Museum (Natural History) and have been returned to England. Paratype 5723, female, is in the United States National Museum collection. The holotype, allotype and the remaining paratypes are in the American Museum of Natural History, New York. Macrocneme vittata Walker (PI. Ill, fig. 30) Macrocneme v/rtota Walker, 1854: 249. Macrocneme vittata, Hampson, 1898: 249, PI. XII, fig. 23. Macrocneme nigritarsis, aberration 1, Hampson, 1898: 326. Macrocneme caurensis Klages, 1906: 540 (after Hampson: 1914: 207). Macrocneme v/ttatfl,Draudt in Seitz: 1915: 105, fig. 17d. Macrocneme nigritarsis, aberration trinitatensis. Strand, 1917: 84. Macrocneme vittata, Forbes, 1939: 127, 133. This species has not been reported by Kaye & Lamont. They did, however, list it as a synonym of eacus (Stoll). Macrocneme eacus does not seem to have been found in Trinidad. It is a small Macrocneme with a suffused bright blue fore- wing in contrast to vittata which has definite if variable iridescent blue marks. The latter seems closely related to alesa Druce, which differs in having more green than vittata. The costal green of vittata extends but little beyond the discal cell from the base, whereas in alesa it runs to near the termen. The blue streak on the anal vein ex- tends beyond the middle of the wing in alesa, but in vittata from Trinidad this streak in the 1957] Fleming: Ctenuchidae (Moths) of Trinidad, B.W.l. ni specimen with it most enlarged is still only half the length of the discal cell. This species will key to alesa in Hampson (1898: 322) since in the key vittata is separated on the basis of having green only in and above the cell, which is con- trary to what Hampson states in his description (1898: 325). Forbes (1939: 127) similarly has vittata restricted to blue in costal half or less. The amount of blue in our specimens is variable in the anal region of the forewing; from only a spot near the base of the wing to streaks above and below the anal vein reaching to approxi- mately the end of the discal cell, above the anal vein (origin of vein Cu2) and half the length of the discal cell below the anal vein. The holotype of alesa is a female from Bolivia and the holo- type of vittata from Para, Brazil. Hampson (1898: 326) on the basis of one female specimen identified as nigritarsia, de- scribed an aberration from Trinidad distin- guished by having white sublateral spots on the four medial segments of the abdomen. Hampson did not mention in his description that this speci- men has some white on the tarsi, which has been seen by Kenedy. Later Strand (1917: 84) named this aberration of Hampson trinitatensis. Hamp- son’s female specimen can hardly be anything else than a normal vittata female, and thus, since the Trinidad locality record for nigritarsis appears to be dependent on this single specimen, nigritarsis cannot be considered as having been taken in Trinidad. Material.— ffimQ specimens (5 males and 4 females). Thirty specimens (14 males and 16 females) in the British Museum (Natural His- tory). One from Caparo, four from Ariapite Valley and the remainder labelled only Trinidad. Range.— Amazon^ and Venezuela (caurensis) . Calonotos Hubner Key to the Trinidad Species of Calonotos 1. Wings without hyaline spots, .helymus Wings with hyaline spots 2 2. Abdomen cupreous with black stripes tiburtus Abdomen iridescent green with black stripes 3 3. Basal segment of abdomen black with a pair of subdorsal white spots and lateral hoods white craneae Basal segments of abdomen green without a pair of white spots, but with dorsal black stripes and lateral white hoods tripunctatus Calonotos helymus (Cramer) Sphinx helymus Cramer, 1775 : 1, pi. 2, figs. D, E. Glaucopis aterrima Sepp, 1848: p. 17, pi. 97. Calonotos helymus, Hampson, 1898: 335. Calonotus helymus, Draudt in Seitz, 1915: 109, fig. 18a. Calonotus helymus, Kaye & Lamont, 1927: 7. This species was collected at Palmiste by Lamont. We have not collected it in the Arima Valley as yet. Hampson in his key (1898: 333) states, “Wings without hyaline spots,” which separates this species from all other Calonotos collected to date from Trinidad. Draudt’s figure in Seitz (1915: 18a) has a single hyaline spot in the discal cell area of the forewing, although his text states that the wings are unspotted. Range.— French and British Guiana. Calonotos tiburtus (Cramer) (PI. Ill, fig. 32) Sphinx tiburtus Cramer, 1780: pi. 237C. Calonotos tiburtus, Hampson, 1898: 333, fig. 154. Calonotus tiburtus, Draudt in Seitz, 1915: 108, figs. 17i 3 & $. Calonotus tiburtus, Kaye & Lamont, 1927: 6. The bright, metallic coppery stripes on the abdomen serve to distinguish this species from other Calonotos of Trinidad. We have found the species quite common flying in the daytime in forested parts of the Nariva Swamp south of Brigand Hill. Male genitalia with a pair of lobe- like processes at base of uncus. Uncus curved ventrally and tapering abruptly to a point at distal end. Dorsal edge of harpe sharply curved ventrally at distal end to form a large ventral process and from inner edge of dorsal edge of harpe a smaller terminal process. From near the base of the harpe on the ventral edge a long slender process. Scoup (juxta?) curved dorsally at end and terminating in a thorn-like spine pre- ceded by a pair of double-pointed thorns. Aedeagus terminating in two long but unequal spines. Strand (1915: 25), on the basis of a male and two females collected in Trinidad, named a local form trinidadensis. The description simply states “Without white spots on the palpal base and on the thorax. Wing expanse 41, length of forewing 20 mm. female.” Materials.— Two male specimens were col- lected on May 5 and June 17 at Simla. Range.— Costa Rica to Surinam. 128 Zoologica: New York Zoological Society [42: 10 Calonotos cmnoae, new species (PI. I, fig. 5; PI. II, fig. 5) Length of forewing of males and females average 19-20 mm. Females average but slightly larger than males. Antennae of male bipectinate with the distal end of each pectination dilated and bristled. The pectinations on each side of the antennal shaft subequal. In the female the antennae are similar, but the pectinations are shorter than in the male and the pectinations on the inside of the antennal shaft of the female are very short in comparison with the pectinations on the outer side of the antennal shaft. Antennae in both sexes with the pectinations decreasing in size towards the apex of the shaft. The dorsal side of the distal end of the antennal shaft white (approximately the distal fifteen segments) but with the terminal one or two segments brown. Palpi normal in shape and position for the genus. The fan-shaped tuft on the base of the first segment of the palpi concolorous brown, as well as the three palpal segments. Front of head blackish-brown with two white spots on upper lateral edges below antennae. Vertex of head, collar, tegulae, thorax and legs concolorous dark brown. A small amount of white on the distal end of the hindcoxae in both sexes but smaller and less conspicuous in the female. In some lights there is a slight bluish- green reflection from the dark brown of the various structures mentioned above. Forewings blackish-brown. Typically with three white spots, one elongate spot below the middle of the discal cell, a subspherical spot be- yond the cell bordering on the discoidal veins proximally and between vein Mi and vein Ms and a third spot between veins Cui and Cus. However, this last-mentioned spot is always smaller in the males than in the females and may be absent, although usually a few white scales persist. In addition, in one female this same spot is so enlarged as to extend half way between veins Cus and Ms. A short iridescent blue sub-basal streak on upper edge of basal third of cell but not extending to base of wing. In two males and one female there is a short iri- descent blue steak within the basal part of the discal cell. Hindwing concolorous black-brown with one white hyaline spot beyond the cell. Underside of forewing concolorous dark brown except for the inner margin below the anal vein which is light brown. The underside of the hindwing concolorous dark brown with iri- descent blue-green patches above and within the discal cell. The whitish-hyaline spots on the upper side of the wings are present on the under- side of the wings. Abdomen shining iridescent, silvery blue with two longitudinal subdorsal black-in-brown stripes. The basal segment black-brown with two gray or whitish subdorsal spots. These spots are seldom conspicuous and may be all but absent. Two lateral white spots on bullae which are larger than the subdorsal spots. A narrow black- ish-brown lateral stripe with a very narrow stripe below of iridescent silvery blue ground color. This silvery blue stripe diminishes and finally disappears on the terminal segments of the ab- domen. The ventral surface of the abdomen blackish-brown with a midventral white stripe. This white stripe is always present but is stronger in some specimens than others. Male genitalia massive and symmetrical. A pair of appendages arises cephalad and dorsad of the uncus. These processes extend caudally beyond the uncus and have a lobe-like structure subventrally at their bases. The base of the uncus is in the plane of the tegumen, but the uncus narrows and bends ventrally at right angles for a distance longer than its horizontal length. The harpe with two processes. One process long and slender with long hair on its ventral edge produced from the ventral edge of the harpe. The ventral edge of the harpe nar- rows abruptly after this process to produce just before the terminus of the harpe a small slender process directed ventrally. Scoup (juxta ?) long, broad and massive, terminating in two long processes bent at right angles in a dorsal direc- tion. Most closely related to triplagus Hampson and chalcipleurus Hampson. In both of these species, as in craneae, the iridescent blue mid- dorsal line commences at the anterior edge of the second abdominal segment rather than at the anterior edge of the first abdominal segment which is characteristic of tripunctatus. The wing expanse of the holotype of triplagus is 38 mm. as against 42-45 mm. of craneae. The collar of triplagus has paired white spots while craneae is concolorous black-brown. The dorsum of the thorax is spotted metallic green in triplagus but is immaculate black-brown in craneae. The coxae and extremities of the femora are spotted white in triplagus and unspotted in craneae. The hyaline white on the wings of triplagus and craneae appear to be similar. The type locality of triplagus is Manaos, Brazil. The legs and thorax are also spotted white in chalcipleurus and the dorsum of the thorax with golden green spots, contrary to craneae in which these areas are black-brown. The pectinations of the an- tennae appear to be longer in chalcipleurus than in craneae. The ventrum of the abdomen with a series of white spots in chalcipleurus, whereas in craneae it has a ventral white stripe. Hampson 1957] Fleming: Ctenuchidae (Moths) of Trinidad, B.W.I. 129 gives the wing expanse of chalcipleurus as 46 mm., which is considerably larger than craneae. I take pleasure in naming this species after Jocelyn Crane, Assistant Director of the De- partment of Tropical Research, New York Zoo- logical Society. All of the type material was collected at Simla, Arima Valley, Trinidad. Holotype, male. Cata- log No. 5656, I-VI; allotype, female (5657) 25-IV; paratypes, 29 males, (5658) 3-1, (5659) 7-1 (5660) ll-I, (5661) ll-I, (5662) 12-III, (5663) 27-m, (5664) 2-IV, (5665) 2-IV, (5666) 3-IV, (5667) 6-IV, (5668) 8-IV, (5669) 16-IV, (5670) 18-IV, (5671) 25-IV, (5672) 25-IV, (5673) 27-IV, (5674) 29-IV, (5675) 1-V, (5676) 2-V, (5677) 4-V (5678) 7-V, (5679) 8-V, (5680) 10-V, (5681) 12-V, (5682) 12-V, (5683) 13-V, (5684) 13-V, (5685) 28-V, (5686) 1-VI; 23 females, (5687) 7-1 (5688) 5-II, (5689) 13-11, (5690) 20-11, (5691) 27-11, (5692) 28-11, (5693) 2-III, (5694) 6-III, (5695) 7-III, (5696) 14-III, (5697) 20-III, (5698) 27-III, (5699) 27-III, (56100) 29-m, (56101) 4-IV, (56102) 19-IV, (56103) 27-IV, (56104) 3-V, (56105) 4-V, (56106) 6-V, (56107) 16-V, (56108) 28-V, (56109) 1-VI. Disposition of type material.— The Depart- ment of Tropical Research, New York Zoologi- cal Society, will retain four paratypes. Catalog Nos. 5662, 5669, 5699 and 56108. Paratypes with Catalog Nos. 5683 and 56106 are in the British Museum (Natural History) collections and paratypes with Catalog Nos. 5682 and 56109 are in the United States National Museum collections. The holotype, allotype and remain- ing paratypes are in the American Museum of Natural History, New York. Calonotos tripunctatus Druce (PI. Ill, fig. 33) Calonotos tripunctatus TiTMCg, 1898: 401. Calonotos tripunctatus, Hampson, 1898: 335 (in part), pi. XII, fig. 7. Calonotus tripunctatus, Draudt in Seitz, 1915: 109, fig. 18a. Calonotus tripunctatus, Kaye & Lamont, 1927: 7. Superficially similar to craneae but may be easily separated by the dorsal green band in tripunctatus running to the thorax and no white points on either side of this line on the basal segment of the abdomen. Kaye & Lamont synonymized chalcipleurus Hampson under tripunctatus, and while I have not seen chalcipleurus, it appears to be a valid species. First of all the basal abdominal segment of chalcipleurus is black-brown as in craneae and tiburtus and does not have the dorsal silvery blue that is present on tripunctatus. A pair of subdorsal white spots above the bullae on the basal abdominal segment of chalcipleurus which is absent in tripunctatus. The thorax and legs of tripunctatus are black-brown, whereas in chal- cipleurus the legs and thorax are spotted white with the dorsum of the thorax with metallic green spots. The type locality of chalcipleurus is Aroa, Venezuela. Male genitalia with large flat lobes on base of uncus. Uncus curved ventrally with distal end inflated but with acuminate terminus. Harpe with ventral process long and slender and dorsal process with a broad ventral spine near bifurca- tion of dorsal and ventral harpal processes and a long finger-like process directed caudally at approximately the mid-point. Terminus of the dorsal process blunt. Scoup (juxta ?) with a single small spine at caudal end and directed laterally. Caudal end of aedeagus with long process at one side. One female collected on May 14 with a Cata- log Number 5491 is provisionally placed in this species. It has only one white spot on the fore- wing but otherwise appears identical. Material.— Tvitniy-one specimens (7 males and 14 females). Holotype from Trinidad. Reported from St. Vincent and Venezuela. 130 Zoologica: New York Zoological Society [42: 10: 1957] EXPLANATION OF THE PLATES Plate I Fig. 1. Pseudosphex kenedyae, dorsolateral view of male genitalia. Fig. 2. Pheia beebei, lateral view of male genitalia. Fig. 3. Saurita arimensis, dorsal view of male genitalia. Fig. 4. Macrocneme spinivalva, dorsal view of male genitalia. Fig. 5. Calonotos craneae, lateral view of male genitalia. Plate II Fig. 1. Pseudosphex kenedyae, new species. Holo- type. Fig. 2. Pheia beebei, new species. Holotype. Fig. 3. Saurita arimensis, new species. Holotype. Fig. 4. Macrocneme spinivalva, new species. Hol- otype. Fig. 5. Calonotos craneae, new species. Holo- type. Plate III The following figures are of species collected at Simla, Arima Valley, Trinidad, except for Cosmo- soma achemon collected at St. Augustine, Trinidad. Fig. 1. Pleurosoma trinitatis 2. Fig. 2. Sphecops aurantiipes $. Fig. 3. Phoenicoprocta vacillans (nigropeltata) $. Fig. 4. Phoenicoprocta vacillans 2- Fig. 5. Pheonicoprocta vacillans 2. Fig. 6. Fig. 7. Fig. 8. Fig. 9. Fig. 10. Fig. 11. Fig. 12. Fig. 13. Fig. 14. Fig. 15. Fig. 16. Fig. 17. Fig. 18. Fig. 19. Fig. 20. Fig. 21. Fig. 22. Fig. 23. Fig. 24. Fig. 25. Fig. 26. Fig. 27. Fig. 28. Fig. 29. Fig. 30. Fig. 31. Fig. 32. Fig. 33. Loxophlebia bisigna Loxophlebia postflavia $. Mesothen endoleuca $. Mesothen pyrrha $. Chrostosoma viridipunctatum $. Cosmosoma rubriscapulae 2. Cosmosoma subflamma subflamma fi. Cosmosoma melathoracia $. Cosmosoma anoxanthia Cosmosoma achemon $. Cosmosoma klagesi Dixophlebia holophaea Pseudomya melanthus $. Pseudomya melanthus 2- Rhynchopyga flavicollis $. Saurita cassandra Saurita clusia $. Saurita clusia 2* Saurita afflict a Saurita afflicta 2- Dycladia basimacula $. Histiaea meldolae $. Histiaea cepheus $. Macrocneme plumbea $. Macrocneme vittata $. Macrocneme thyra intacta $■ Calonotos tiburtus 2. Calonotos tripunctatus $. FLEMING PLATE 1 FIG. 1 FIG. 3 FIG. 4 FIG. 5 THE CTENUCHIDAE (MOTHS) OF TRINIDAD, B.W.I. PART I. EUCHROMIINAE FLEMING PLATE II FIG. 2 FIG. 3 FIG. 4 FIG. 5 THE CTENUCHIDAE (MOTHS) OF TRINIDAD, B.W.L PART I. EUCHROMIINAE FLEMING PLATE III PART I. EUCHROMIINAE NEW YORK ZOOLOGICAL SOCIETY GENERAL OFFICE 30 East Fortieth Street, New York 16, N. Y. PUBLICATION OFFICE The Zoological Park, New York 60, N. Y. OFFICERS PRESIDENT VICE-PRESIDENTS SECRETARY TREASURER Fairfield Osborn Alfred Ely Harold J. O’Connell David H. McAlpin Laurance S. Rockefeller SCIENTIFIC STAFF: Zoological Park and Aquarium John Tee-Van General Director ZOOLOGICAL PARK Leonard J. Goss Assistant Director and Veterinarian John L. George Associate Curator of Mammals William G, Conway. .Associate Curator of Birds Grace Davall Assistant Curator, Mammals and Birds James A. Oliver Curator of Reptiles Charles P. Gandal. . .Associate Veterinarian Lee S. Crandall General Curator Emeritus William Beebe Honorary Curator, Birds AQUARIUM Christopher W. Coates . Director James W. Atz Associate Curator Carleton Ray Assistant to the Director Ross F. Nigrelli Pathologist Myron Gordon Geneticist C. M. Breder, Jr Research Associate in Ichthyology Harry A. Charipper. . .Research Associate in Histology Homer W. Smith Research Associate in Physiology GENERAL William Bridges Editor & Curator, Publications Sam Dunton Photographer Henry M. Lester . . . Photographic Consultant DEPARTMENT OF TROPICAL RESEARCH William Beebe Director Emeritus Jocelyn Crane Assistant Director David W. Snow Resident Naturalist Henry Fleming Entomologist John Tee- Van Associate William K. Gregory Associate AFFILIATES C. R. Carpenter Co-ordinator, Animal Behavior Research Programs L. Floyd Clarke Director, Jackson Hole Research Station SCIENTIFIC ADVISORY COUNCIL A. Raymond Dochez Alfred E. Emerson W. A. Hagan Caryl P. Haskins K. S. Lashley John S. Nicholas EDITORIAL COMMITTEE Fairfield Osborn, Chairman James W. Atz William Beebe William Bridges Christopher W. Coates William G. Conway Lee S. Crandall Leonard J. Goss James A. Oliver John Tee-Van 200L0GICA SCIENTIFIC CONTRIBUTIONS OF THE NEW YORK ZOOLOGICAL SOCIETY VOLUME 42 • PART 4 • DECEMBER 31, 1957 • NUMBERS 11 TO 14 PUBLISHED BY THE SOCIETY The ZOOLOGICAL PARK, New York Contents PAGE Part 4. December 31, 1957 11. The Ability of the Saprolegniaceae to Parasitize Platyfish. By Helen S. Vishniac & R. F. Nigrelli. Plate 1 131 12. Imaginal Behavior in Butterflies of the Family HeUconiidae: Changing Social Patterns and Irrelevant Actions. By Jocelyn Crane. Plate I.. . . 135 13. Habits, Palatabihty and Mimicry in Thirteen Ctenuchid Moth Species from Trinidad, B.W.I. By William Beebe & Rosemary Kenedy. Plates I & II. 147 14. Serological Relationships among Members of the Order Carnivora. By Ludwig K. Pauly & Harold R. Wolfe 159 Index to Volume 42 167 11 The Ability of the Saprolegniaceae to Parasitize Platyfish' Helen S. Vishniac & R. F. Nigrelli Department of Microbiology, Yale University, and New York Aquarium, N. Y. Zoological Society (Plate I) A NUMBER of species of fungi belonging to the Saprolegniaceae have been iso- ^ lated from naturally occurring infec- tions in fish and other aquatic animals (Table 1 ) . The list is so diverse, from a taxonomic point of view, as to suggest that any saprolegniaceous fungus might be capable of parasitism under ap- propriate conditions. The experiments reported here were undertaken to determine whether genera and species not previously reported to be parasitic, as well as saprophytic isolates of species known as parasites, could infect platyfish under conditions which allowed infection by two fungi isolated from naturally occurring infections. The platyfish is known to be susceptible to Sapro- legnia (Gordon, 1936). However, there is no evidence of host specificity for these fungi even though there is variation in the susceptibility of different fish in captivity (Tiffney, 1939a) as well as in nature. Tiffney (1939a) reported that Saprolegnia parasitica can attack at least 16 species of fishes and 2 species of amphibians. Saprolegnia parasitica also attacks aquatic rep- tiles both in nature and in captivity. Other species have been reported in crustaceans (Atkins, 1954; Hohnk & Vallin, 1953; Prowse, 1954; Smith, 1940), lamellibranchs (Atkins, 1954), 1 It is a pleasure to acknowledge our debt to Dr. Myron Gordon and the Genetics Laboratory of the New York Zoological Society for the fish used in these experiments. We are indebted to Dr. J. R. Raper, Dr. A. W. Ziegler and Dr. L. Shanor for the isolates desig- nated by their initials. The experimental infections were made at Haskins Laboratories, New York City, in the laboratory of Dr. S. H. Hutner, for whose interest we are grateful. This study was supported in part by re- search grant E 1179 from the National Institute of Allergy and Infectious Diseases of the National Insti- tutes of Health, U. S. Public Health Sendee. larval mosquitoes (Rioux & Achard, 1956) and rotifers (Valkanov, 1931). Mycosis caused by Saprolegniaceae is a disease of temperate and tropical freshwater and possibly brackish water fishes. Those affecting temperate water fishes, e.g. carp, have been recognized in Central Europe since the Middle Ages. The num- ber of fish species susceptible to infections by Saprolegniaceae has never been properly re- corded. From more than 25 years’ experience with the diseases of fish, it can be stated without exaggeration that most freshwater fishes are sus- ceptible to infection by some species of Sapro- legniaceae, especially under tank and hatchery conditions and following injury of the skin or gills. In fishes, primary infections occur in in- dividuals that have lesions caused by injury. Lesions caused by protozoan and metazoan parasites are the loci of secondary infections. At the New York Aquarium, more than 45 deaths were reported in a single year as being caused by “Saprolegnia” (Nigrelli, 1943). A re-check of the Aquarium’s records for 1939, 1940 and 1941 showed that 24 species of temperate and tropical freshwater fishes, 4 species of urodeles and 4 species of turtles died from such infections. Saprolegnia-VikQ infections were also found in the feet of alligators which had developed ulcers from other causes. No efforts were made at the time to determine which species of Sapro- legniaceae were involved in these deaths. The results of our experiments indicate that although the family Saprolegniaceae is primarily saprophytic in nature, the potentiality of para- sitic existence is a familial characteristic. More- over, imder our conditions, at least, the infecting fungus can be a primary invader. 131 132 Zoologica: New York Zoological Society [42: 11 Table 1. Saprolegniaceae Occurring as Parasites of Aquatic Animals Fungus Reference Achlya flagellata see Tiffney, 1939b A. polyandra see Tiffney, 1939b A. prolifera see Tiffney, 1939a, b A. racemosa see Tiffney, 1939b Achlya sp. Tiffney, 1939b Aphanomyces laevis Smith, 1940 Ap. astaci^ see Prowse, 1954 Ap. daphniae^ Prowse, 1954 Ap. hydatinae^ Valkanov, 1931 Ap. ovidestruens'- see Prowse, 1954 Dictyuchus monosporus Tiffney, 1939b Leptolegnia baltica Hohnk & Vallin, 1953 L. caudata see Coker, 1923 L. marina Atkins, 1954 Saprolegnia diclina Rioux & Achard, 1956 S. ferax see Tiffney, 1939a, b S. mixta see Tiffney, 1939b S. monoica see Tiffney, 1939a, b S. parasitica see Tiffney, 1939a, b S. torulosa see Tiffney, 1939b 1 If these are indeed valid species. Materials and Methods The 19 isolates of saprolegniaceous fungi used are listed in Table 2. Brevilegnia unisperma and Dictyuchus monosporus were originally included in this group, but were omitted when they failed to produce zoospores under the conditions of the experiments, since infection in nature is normally accomplished by zoospores (Tiffney, 1939a). Two of the fungi listed were isolated from naturally infected aquarium fish: Achlya sp. from an electric catfish (Malapterurus elec- tricus) and Saprolegnia parasitica from the mos- quito fish (Gambusia affinis). The Achlya sp. was similar to the sterile Achlya described by Tiffney (1939b) from the shell of Chelydra serpentina, except in the ap- pearance of the “chlamydospores.” In our isolate, these structures were spherical to ovoid, never clavate. They were formed first terminally, later on racemose branches or less frequently in basi- petal chains. These “chlamydospores” were shown in our isolate to be oogonia. When mats (grown in pure culture) of Achlya sp. were placed in a petri dish in proximity to mats of male (E247) and female (355) thalli of A. bi- sexualis, abundant antheridial initials formed on E247 within three to four hours. After 24 hours, well differentiated oospheres could be seen in the “chlamydospores” of Achlya sp. A few eccentric oospores maturing in the oogonia of Achlya sp. attested to the completion of the mating reaction with E247. There was no reaction between our Achlya sp. and the female isolate 355. This fun- gus cannot, however, be considered a female isolate of A. bisexualis or of any other described heterothallic species of Achlya. Not only were the majority of oospores aborted in the mating just described, but the characteristic of produc- ing oogonial initials in the absence of hormonal stimuli provided by the presence of antheridial initials is present in no other heterothallic Achlya (see Raper, 1955). Since specific distinctions in j this family of fungi are made largely on the basis j of characteristics of sexual reproduction, we con- sider it inadvisable to formally describe this species in the absence of a male isolate giving a normal complete mating reaction. Table 2. The Ability of Various Saprolegniaceae to Infect Platyfish ^ _ Uninfected Fish Infected Fish J7UU^US JTiCSiCUL Alive \ Died Alive 1 Died None 2 2 0 0 None 4 0 0 0 Achlya sp. 1 0 1 2 A. ambisexualis E87 S (J.R.R.) 0 0 0 4 A. ambisexualis 302 $ (J.R.R.) 0 0 0 4 A. bisexualis BIM $ (J.R.R.) 0 0 0 4 A. bisexualis 355 $ (J.R.R.) 0 0 0 4 A. glomerate (A.W.Z.) 2 2 0 0 A. Klebsiana (L.S.) 0 0 0 4 A. Sparrowii^ 0 0 0 4 A phanomyces laevis (A.W.Z.) 0 0 0 4 Calyptralegnia achlyoides (L.S.) 0 0 2 2 Isoachlya monilifera 3 0 0 1 Protoachlya paradoxa (L.S.) 0 1 0 3 Saprolegnia delica (L.S.) 0 0 0 4 S. ferax 0 0 0 4 S. megasperma^ 0 1 0 3 S. mixta 0 0 0 4 S. parasitica 1 0 0 3 Thraustotheca clavata 0 0 0 4 T. primoachlydr (A.W.Z.) 0 0 2 2 1 Johnson (1956) considers this species to be synony- mous with A. racemosa Hildebrand. Since A. racemose has centric oospores while A. Sparrowii has subcentric oospores (a distinction which Johnson considers, in the same study, to be of subgeneric rank), we prefer to retain A. Sparrowii. 2 Two broods of young were born in the course of this experiment. Normal young (9 and 5) remained healthy; premature fry (i.e. with visible yolk sac) be- came parasitized and died (5 and 3). It was the mother of one of these broods that died without becoming parasitized by Saprolegnia megasperma. 1957] Vishniac & Nigrelli: Ability of Saprolegniaceae to Parasitize Platyfish 133 The Mexican platyfish, Xiphophorus macu- latus, was used as host in these experiments. The conditions chosen for these experiments were determined, by means of preliminary ex- periments, to provide the greatest chance of in- fection by the two parasitic isolates (Achlya sp. and Saprolegnia parasitica) together with the lowest mortality from causes other than induced fungal infection. Pyrex kitchen trays (ca. 8"X 12"X2") were washed thoroughly, steamed, filled with 1.5 liters of tap water, and placed on an illuminated bench at room temperature (ap- proximately 20° during the course of these ex- periments). An excess of fish food was then placed in them and inoculated with a pure cul- ture of a fungus. When the growth of the fungus on the fish food was producing an abundance of zoospores, four platyfish, roughly evenly dis- tributed as to size and sex, were placed in each tray, together with a few strands of Nitella. Just before exposure to the fungi, these fish were in- jured by scraping the scales from an area ap- proximately 2X2 mm. on one side of the caudal peduncle. Uninjured fish were not attacked; fish injured by mere rubbing or nicking of the caudal peduncle were not consistently attacked. The fish were then observed until death, at which time they were removed from the tray, examined to verify the identity of the fungus and preserved in formalin (Plate I) , or until the disappearance of the fungus from the tray. Paraffin sections of the diseased peduncle were prepared and stained with haematoxylin-eosin and with Masson’s trichrome stain; whole mounts of the skin and scales showing the my- celia were treated with a modification of Mal- lory’s method in which the following stains were used: Harris’ haematoxylin, Phloxin B and Stir- ling’s anilin crystal violet. The tissues were then then treated with Gram’s iodine and differen- tiated with several changes of anilin oil until no more color was removed. The stained material was then passed through several changes of xylene and mounted in Permount. With this method, the mycelia stained blue and the spo- rangia red and the extent of the infection was followed with ease. Results The results of these experiments are given in Table 2. It is evident that under the conditions which permit infection by the two parasitic isolates, nearly every saprolegniaceous fungus used can attack, and usually kill, platyfish. The two apparent exceptions, Achlya glomerata and Isoachlya monilifera, did not in fact present quite the same conditions as the parasitic isolates. Achlya glomerata grew sparsely, although what growth there was produced zoospores. Isoachlya monilifera grew moderately well before the ad- dition of the fish, which promptly ate up the mycelium, thus greatly reducing their chance of infection. Consumption of mats of living my- celium of Saprolegnia parasitica by fish which suffered no harm thereby has been noted by Tiffney (1939a). The fungus appearing on in- fected fish was in every case the species with which the tray had been inoculated. The first signs of infection always appeared at the injured area on the caudal peduncle in the form of a tuft of hyphae. Later, in fatal infec- tions, hyphae often emerged from the gills and mouth, and in tufts over the body, or the entire fish became covered with the fuzzy growth of the fungus. The pathological lesions were more or less similar, varying only in degree. In relatively light infections, the mycelia penetrated the epi- thelium of the scales and skin with some necrosis of the involved areas. In heavy infections, the epithelium was often sloughed and the scales and fin rays were softened or completely destroyed. The growth penetrated the deeper tissues, the hyphae often passing into the muscle bundles and resulting in hyalinization or complete destruc- tion. Macrophages filled with melanin, cellular debris and blood cells, together with lympho- cytes, were invariably found massed in the areas of the mycelial growth. Inflammatory reaction of varying intensity occurred in regions immediately adjacent to the infection. Surprising as it may seem, little or no bacterial infection was present in any of the sections studied, indicating that the tissue destruction was due almost exclusively to the fungi. Discussion and Conclusions The criteria listed by Henle in 1840 for es- tablishing a causal relationship between an organism and a disease in its putative host begin with the isolation of the organism from cases of the disease. We have shown that given a set of conditions which allow infection by two sapro- legniaceous fungi isolated from diseased fish, all other saprolegniaceous fungi for which the same conditions (including abundant sporulation) could be provided would infect platyfish. Some of the species used, although not isolated from diseased animals in this instance, have pre- viously been reported to be parasitic, viz. Aphanomyces laevis, Saprolegnia ferax and S. mixta. The remaining 12 species have not been so reported. Since from these results and pre- vious reports at least 27 species in 10 genera of the Saprolegniaceae have been found capable of attacking animal hosts, we may conclude that 134 Zoologica: New York Zoological Society [42: 11:1957] potential parasitism is a familial characteristic. But from the ecological point of view the 12 species not known to be involved in natural in- fections are not parasites. There is, to be sure, a very marked difference in frequency of oc- currence of reported parasites. Saprolegnia parasitica is universally reported as most com- mon; other species are reported in frequencies ranging down to only a single established case. It would hardly be surprising if continued search should disclose that the forms we have desig- nated as potential parasites are actually parasitic in nature. The reasons for the rarity or absence of the parasitic habit in some species of this family are probably to be found in their specific ecology. A successful parasite in nature must be abundantly sporulating under the conditions of temperature, etc., that exist when and where susceptible animals appear. While the report of Coker (1923) on seasonal occurrence of sapro- legniaceous fungi and the extensive studies of Hohnk (1934, 1956), of Hohnk & Bock (1955) and of Bock (1956) are outstanding as contribu- tions to our knowledge of the specific ecology of the Saprolegniaceae, more precise quantitative techniques would be desirable before attempts are made to correlate the production of zoospores and infections by various species of the Sapro- legniaceae in nature. Summary Eighteen of 19 isolates of saprolegniaceous fungi, including 16 species belonging to 7 genera, infected platyfish having a standardized wound on the caudal peduncle. The infection usually resulted in the death of the fish. Histological ex- amination indicated that tissue destruction was due almost exclusively to the infecting fungus. Literature Cited Atkins, D. 1954. Further notes on a marine member of the Saprolegniaceae, Leptolegnia marina N. Sp., infecting certain invertebrates. J. Mar. Biol. Ass. U. K., 33: 613-625. Bock, K. J. 1956. Zur Okologie und Systematik saprophy- tischer Wasserpilze aus dem Silbersee bei Bremerhaven. Veroff. Inst. Meeresforsch. in Bremerhaven, 4: 25-44. Coker, W. C. 1923. The Saprolegniaceae. U. of North Caro- lina Press, Chapel Hill. 201 pp. Gordon, M. 1936. Fishes, Beware the Fungus. Nature Maga- zine, 27: 145-146; and (reprinted) The Aquarium, 5: 27-29. Hohnk, W. 1934. Saprolegniales und Monoblepharidales aus der Umgebung Bremens, mit beson- derer Beriicksichtigung der Okologie der Saprolegniaceae. Abh. Naturwiss. Verein Bremen, 29: 207-237. 1956. Mykologische Abwasserstudie I. Veroff. Inst. Meeresforsch. in Bremerhaven, 4: 67-110. Hohnk, W., & K. J. Bock 1955. Ein Beitrag zur Okologie der saprophy- tischen Wasserpilze. Veroff. Inst. Meeres- forsch. in Bremerhaven, 3 : 9-26. Hohnk, W., & St. Vallin 1953. Epidemisches Absterben von Eurytemora, verursacht durch Leptolegnia baltica nov. spec. Veroff. Inst. Meeresforsch. in Bre- merhaven, 2: 215-223. Johnson, T. W., Jr. 1956. The Genus Achlya. U. of Michigan Press, Ann Arbor. 180 pp. Nigrelli, R. 1943. Causes of Diseases and Death of Fishes in Captivity. Zoologica, 28: 203-216. Prowse, G. a. 1954. Aphanomyces daphniae Sp. Nov., Para- sitic in Daphnia hyalina. Trans. Brit. Myc. Soc., 37: 22-28. Rarer, J. R. 1955. Some problems of specificity in the sex- uality of plants. In Biological Specificity and Growth. E. G. Butler, ed. (12th Sym- posium Soc. for the Study of Development and Growth). Princeton U. Press, Prince- ton, N. J. Pp. 119-140. Rioux, J.-A., & F. Achard 1956. Entomophytose mortelle a Saprolegnia diclina Humphrey 1892 dans un elevage d'Aedes berlandi Seguy 1921. Vie et MUieu, 7: 326-335. Smith, R. 1940. Studies on two strains of Aphanomyces laevis found occurring as wound parasites on crayfish. Mycologia, 32: 205-213. Tiffney, W. N. 1939a. The host range of Saprolegnia parasitica. Mycologia, 31: 310-321. 1939b. The identity of certain species of Sapro- legniaceae parasitic to fish. J. Elisha Mitchell Sci. Soc., 55: 134-151. Valkanov, a. 1931. fiber Morphologie und Systematik der rotatorien-befallenden Pilze. Arch. Pro- tistenk., 74: 5-17. EXPLANATION OF THE PLATE Plate I Fig. 1. Platyfish infected with Isoachyla moni- lifera (upper) and Saprolegnia mega- sperma (lower). Preserved in formalin. 2X. VISHNIAC & NIGRELLI PLATE I THE ABILITY OF THE SAPROLEGNIACEAE TO PARASITIZE PLATYFISH I . ■“ u. . i\'^i -• rr''^ "'ji %;'S ' I- .t-; r " t' ■ 1'. '" ’ *., ■' ' I? JjjiS- ')lj. \H" < V 12 Imaginal Behavior in Butterflies of the Family Heliconiidae: Changing Social Patterns and Irrelevant Actions^ Jocelyn Crane Department of Tropical Research, New York Zoological Society, New York 60, N. Y. (Plate I) [This paper is one of a series emanating from the tropical Field Station of the New York Zoological Society at Simla, Arima Valley, Trinidad, British West Indies. The Station was founded in 1950 by the Zoological Society’s Department of Tropical Re- search, under the direction of Dr. William Beebe. It comprises 200 acres in the middle of the Northern Range, which includes large stretches of undisturbed government forest reserves. The laboratory of the Station is intended for research in tropical ecology and in animal behavior. The altitude of the research area is 500 to 1,800 feet, with an annual rainfall of more than 100 inches. [For further ecological details of meteorology and biotic zones see “Introduction to the Ecology of the Arima Valley, Trinidad, B.W.I.,” William Beebe. (Zoologica, 1952, Vol. 37, No. 13, pp. 157-184.)] Contents I. Introduction 135 II. Historical Review 136 III. Materials and Methods 137 IV. Survey of Social Behavior in Six Species of Trinidad Heliconiidae 137 V. Changes in Social Patterns with Age. .. . 138 VI. Irrelevant Actions 139 VII. Discussion 141 VIII. Summary 143 IX. References 143 I. Introduction This contribution is one of a series on the biology and ecology of butterflies of the neotropical family Heliconiidae. It con- cerns post-imaginal changes in social behavior patterns, and various kinds of irrelevant be- havior, including the displacement of actions ^Contribution No. 979, Department of Tropical Re- search, New York Zoological Society. characteristic of one pattern by actions from another behavioral field. The observations presented arose as unex- pected incidental results of a comparative etho- logical study of six species common on the island of Trinidad, British West Indies. The publication of these data appears to be a prerequisite to the adequate exposition of comparative specific characteristics and to an illumination of the group’s phylogeny. A first paper in the series, on Heliconius erato hydara Hewitson, has been published in which the effect of age on behavior and the occurrence of displacement behavior were mentioned (Crane, 1955). Five additional species, along with H. erato, are considered below. These consist of Dry as Julia Julia (Fab- ricius), Heliconius melpomene euryades Rif- farth, H. sara rhea Cramer, H. ricini insulana Stichel and H. isabella isabella Cramer. Details of their behavior patterns, accounts of releasing mechanisms and discussion of phylogenetic as- pects are being reserved for more appropriate inclusion in forthcoming contributions. The risks of inaccuracy and incompleteness in the description of butterfly behavior became ap- parent very early in the work, particularly in the study of courtships. The patterns seemed at once surprisingly variable within a species, un- expectedly similar among species and altogether too unpredictable to be characteristic of arthro- pods, with their high degree of dependence on stereotypic forms of behavior. Since the butterflies were being studied prin- cipally in outdoor insectaries, it was thought at first that the discrepancies might be the result of unnatural conditions— although even if that were so the interest of the variability of the be- havior would scarcely have been lessened. 135 136 Zoologica: New York Zoological Society [42: 12 Enough corroborative observations have now been made in the field, however, to show that the apparent eccentricities are equally charac- teristic of free-flying butterflies. Further cor- roboration of the natural prevalence of the types of behavior under consideration comes from the fact that, now that the causes are better under- stood, the patterns shown in the insectaries are highly predictable. Some of the behavioral discrepancies have proved to be due simply to age differences in the individuals observed. The full courtship pattern characteristic of the species, for example, is elicited only in and by individuals between cer- tain ages, although successful reproduction can take place throughout a much longer portion of imaginal life. Other irregularities appear when, as Tin- bergen (1952, p. 26) detoes the conditions for displacement activities, “a strongly activated drive is denied discharge through its own con- summatory act(s).” However, in the present paper the term “irrelevant behavior” (suggested by Rand, 1943) will be employed as a more general term than “displacement behavior.” The latter, it seems, may be usefully restricted to the definition given by Bastock, Morris & Moynihan (1953, p. 25): “A displacement activity is an activity belonging to the executive motor pat- tern of an instinct other than the instincts activated.” This distinction between terms seems desir- able since some of the butterfly actions under consideration do not at all appear to belong to the motor pattern of another instinct, and hence will be simply referred to as “irrelevant actions.” Others, which fulfil the conditions of the more restricted definition of displacement activities, will be so designated here and treated as a sub- division of irrelevant actions. My thanks go to the National Geographic Society for a grant-in-aid, to Dr. William Beebe, Mr. Henry Fleming and Dr. D. W. Snow for helpful suggestions, and to Miss Barbara P. Young for rearing numerous larvae. II. Historical Review The special aspects of social behavior under consideration are little-known fields in the study of invertebrates. The gradual development of behavior pat- terns in physiologically adult vertebrates has been extensively studied, and it is well known that changes occur in response to physiological alterations due both to increasing age and to seasonal causes. Corresponding information has been gath- ered on few invertebrates, although the Hymen- optera include outstanding exceptions. The sequence of changes in colony functions have been studied in worker honeybees (Rosch, 1925), Polistes (Steiner, 1932) and in various ants {e.g. Buckingham, 1910). Verlaine (1932) reported differences between young and old mason wasps in nest-repairing and provision be- havior, that of old bees late in the season being incomplete; this observation invalidated a con- clusion drawn by Fabre (1879; ed. 1920) from experiments which he performed, using aged individuals. Pardi (1947) found that age was one of the factors determining the status of in- dividual Polistes females in a dominance hier- archy. Nielsen & Nielsen (1952) reported that the migratory period of a pierid butterfly (Astia) was confined to a single day of the five- day life-span. As an example of intrageneric variation, the parasitic wasp genus (Opius) may be cited: males of certain species cannot mate for five days or more after emergence, although in other species they do so early in the imaginal period (Hagen, 1953). Finally, entomologists would probably agree that female insects that have recently molted into the final instar are in general more attractive to males than are older individuals. It will be noted, however, that use of the inexact term “recently” is necessary. Rockstein (1956) discussed the unreality of a sharp boundary between the pupal and imag- inal stages in insects, citing recent reasearch on biochemical changes occurring after emergence in the worker honeybee, house fly. Drosophila, moths and the Japanese beetle. In Drosophila there are concomitant increases in glycogen con- tent and wing-beat frequency during the first week of imaginal life. Although this frequency change is not directly related to social behavior, the phenomenon illustrates the kind of corre- lation which may be brought to light in investi- gations linking invertebrate physiology and be- havior. Studies of irrelevant actions, including dis- placement behavior, in vertebrates are increas- ing in number, following the pioneer work of Lorenz, Tinbergen, Makkink, Koorlandt and Armstrong. General accounts and references are given by Armstrong (1950), Lorenz (1950), Tinbergen (1951, 1952), and Bastock, Morris & Moynihan (1953). Recent studies on particu- lar species of birds and fish include those of Hinde (1953), van lersel (1953), Moynihan (1953) and Morris (1954). The probable occurrence of equivalent be- havior in invertebrates has apparently not been suggested until recently. Armstrong (1950, pp. 379 ff.) summarized the situation as it appeared at the time of his writing as follows: “Probably displacement activities are commoner in some other groups besides birds than is at present 1957] Crane: Imaginal Behavior in Butterflies of the Family Heliconiidae 137 known, but they have reached their highest de- velopment in birds ... So far as is at present known they have a comparatively insignificant role among insects. A thwarted solitary wasp, Bembex rostrata, when forced to remove pebbles repeatedly from the mouth of its burrow merely buzzed loudly and ran around in a wide arc (Nielsen, 1945). When experimenting with but- terflies attracted to pieces of coloured paper Dr. D. Use noticed movements which might have been displacement activities (personal communication). Possibly one of the factors re- sponsible for the apparently slower speciation of insects than birds (Mayr, 1942) is the greater displacement-proneness of the latter.” A few recent observations and comments, however, suggest that displacement activities may prove, after all, to be widespread among higher arthropods. References to date appear to be confined to the following: salticid spiders (Crane, 1949), mantids {idem, 1952), the but- terfly erato {idem, 1955) , Drosophila (Bastock & Manning, 1955) and fiddler crabs, genus Uca (Gordon, 1955, and Crane, 1957). III. Material and Methods The studies were all conducted in out-of-door wire mesh insectaries in Trinidad (Crane & Flem- ing, 1953; Crane, 1955) between 1954 and 1957. During the past two seasons a new in- sectary, designed as were the earlier ones by Henry Fleming, has been in operation. Con- structed entirely of aluminum, it measures 24 X 36 feet, the dimensions of the larger of the two earlier structures, but it is higher than its prede- cessor, measuring 12 feet at the ridgepole. It also has two doors with a small vestibule be- tween, forming a baffle which has proved very useful in preventing the escape of butterflies. A small pond and bog have been added near one end; they form an efficient aid in maintaining the necessary high humidity. The aluminum netting reflects far more heat than does bronze mesh; it also diffuses the light better, making it excellent for photography. Finally, species suitable for keeping in a cage of this size tend to bat against the netting less, even when they have just been released into it, than they did against the bronze netting of the previous insectary. For this reason, as well as because of the in- sectary’s relative coolness, heavy vines giving large areas of dense shade have been found to be not only unnecessary but undesirable. Instead, the planting is kept to several well-separated major groups of shrubs, saplings and wild bananas of varying degrees of height, density and leaf size. The rest of the space is occupied by flowering weeds and by the pond. Branches with bromeliads attached stand upright at inter- vals against the netting, as a further aid in main- taining humidity and natural conditions. The wild banana group {Heliconia) beside the bog forms the coolest, most shady corner of the cage. H. erato, melpomene and other shade-lovers in- variably seek it out during the heat of the day. In contrast, other species, such as H. isabella, which is most active around noon, frequent the open center of the cage around a group of Lan- tana, Bidens and Asclepias. In this cage all of the latter, favorite food blossoms of the heli- coniids, thrive and are allowed to grow freely over most of the cage. This scattering of a natural food supply encourages a normal amount of flight by the butterflies. The success of the new design is attested by the fact that all of the six species (p. 135) of heliconiids discussed in the present paper feed, court, mate and lay eggs. Many individuals re- main alive, barring accidents, for one to three and one-half months, although differences in viability are shown. H. isabella is the most dif- ficult to maintain, and observations on this species are still somewhat deficient. Two locally rare heliconiids, Philaetraea dido (Linnaeus) and Heliconius wallacei Reakirt, have not been successfully maintained. They live a few days and feed, but do not “settle in” and spend most of their active periods batting against the roof. Presumably a higher cage is needed. Broods from aU the species of heliconiids in- cluded in this study were raised in the labora- tory. The young imagoes were kept in small cages out-of-doors until needed for observation or testing in the large insectary. General meth- ods are given in an earlier paper (Crane, 1955). Table 1 gives an idea of the number of healthy imagoes that were used in the preparation of this study. Substandard specimens, as well as those obesrved in the early seasons before tech- niques were perfected and the present problems formulated, are not included. IV. Survey of Social Behavior in Six Species OF Trinidad Heliconiidae The social behavior of all six of the Trinidad species discussed below consists of three general types— courtship, “social chasing” and roosting. These have already been described in some de- tail for H. erato (Crane, 1955). Except for minor differences they are characteristics of the other five species as well. Although these slight specific differences are of great potential interest from a phylogenetic point of view, their detailed dis- cussion belongs in subsequent papers on the ethology of the genus. 138 Zoologica: New York Zoological Society [42: 12 Table I. Numbers of Individuals upon which Present Data Are Based (From broods reared during the seasons of 1954-1957, incl.). Species Males Females Dryas julia 49 46 Heliconius melpomene 27 28 Heliconius erato 71 62 Heliconius ricini 38 27 Heliconius isabella 21 22 Heliconius sara (1957 only) 16 19 A. Courtship. Since the similarities of the behavior within the genus are far greater than the differences, a comparative chart (Table 2) of the patterns of fully developed courtship will in- dicate the trends to the extent needed for present purposes. It will be seen from the table that courtship usually begins and always ends similarly in all species, while differentiation is shown principal- ly in the first and second stages of the second, sedentary phase. In brief summary, the sequence in its most complete form is as follows: 1. Aerial Phase. Stage 1. Nudging. A flying male approaches a resting female from the rear. She then takes wing, usually without his actually touching her in any way. Stage II. Flight. The male chases and over- takes the female, rises above and in front of her and fans her with the rapid vibration of his wings, so spreading the products of his scent scales. She then descends or is forced down to a perch. Chases sometimes include mutual circling and spiralling in all the species. At these times the circling by the female is apparently always merely the result of her temporarily successful effort to duck out from under and behind the flying male and rise above and in front of him; he in turn repeats the manoeuvre and the result- ing vertical circling may continue for several minutes. Horizontal spiralling is less frequent but seems to have a similar basis. In the pre- viously published account of erato (Crane, 1955) this flight stage, which is little developed in that species and often omitted, was not separated from Stage I of the Sedentary Phase below. 2. Sedentary Phase. Stage I. Primary Fanning. The male fans the alighted female from the front or rear, de- pending on the species, facing in the same direction as the female. The latter, meanwhile, flutters her wings, also characteristically, elevates the abdomen and, in this stage or the next, ex- trudes the subterminal scent glands. Stage II. Secondary Fanning. The male fans the female from a different position from that in Stage I, but still facing in the same direction as she. The female meanwhile extrudes the sub- terminal scent glands if she has not already done so. Stage III. Alighting and Engaging. The male alights beside the female, either just in front of or just behind her, moves back or forward ap- propriately and, as she closes her wings, curves his abdomen up between her posterior wings, en- gages her genitalia with his harpes, and swings around so that the two insects now face in op- posite directions. B. Social Chasing. Discussed first in 1955, by Crane, in erato, “social chasing” was the term given to social flights that are not apparently di- rectly of a sexual nature. It was found to take similar form in the other five species. In all, it consists of the pursuit of males and old females by males of any age, and of either sex by old females. It will be discussed on pp. 141 and 142. C. Roosting. Four of the six species roost gregariously, namely H. erato, melpomene, ricini and Sara. All return to the same bush or vine, and often to the same twigs or tendrils, night after night. Usually the perch selected is dry. Although the four species tend to maintain sep- arate roosts, erato, melpomene and ricini often roost together, as do ricini and sara. A trace of gregariousness is found in Dryas, which sometimes hangs up for the night near one or two others of its kind. H. isabella, how- ever, always roosts alone. The two latter species always hang from beneath green leaves. Roost- ing will be further considered on p. 139. V. Changes in Social Patterns with Age None of the social behavior patterns sum- marized in the foregoing section is fully ex- pressed when the insect emerges from the chrys- alid, and it now appears that the maturation time for the various responses is similar in all six species. Table 3 shows these periods. “Age” indicates time after emergence from chrysalid. It will be seen from the table that the full pattern of courtship— including Stage II of the Aerial Phase and the specifically characteristic wing flutters of the Sedentary Phase— is evinced by females only on their second and third days, that is, between 24 and about 56 hours after [DS Species y Phase Stage III M; Female Male Female Dryas jidia Nudges rear ification of Stage I Alights beside 9 head, moves backward, curves abdomen forward between 9 wings, en- gages harpes, swings around to face in opposite direction from 9 Closes wings; lowers abdo- men; withdraws scent glands Heliconius melpomene Dryas As in Dryas As in Dryas Heliconius erato Stage I, but lower in- j As in Dryas As in Dryas Heliconius ricini Dryas Alights beside 9 abdomen, moves forward, then pro- ceeds as in Dryas. (Rarely alights beside 9 head as in Dryas). As in Dryas Heliconius isabella Dryas As in Dryas As in Dryas Heliconius Sara Dryas As in Dryas As in Dryas ■ 1. ■» I \ 'I' I I I 't, V,' t i ' # u 'it ) U £A4 J • •' » > Table II. Basic Motion Sequence in Fully Developed Uncomplicated Courtships of Trinidad Heliconuds Species Aerial Phase Sedentary Phase Stage I Stage n Stage 1 Stage n Stage III Male j Female Male I Female Male Female Male Female Male Female Dryos jiilia Nudges from rear Takes flight Pursues; aims to fly above Aims to back out from under and in front of $. (Scent and rise above Eventually scales probably in use) . alights on exposed perch. Front fanning, facing in same direction as $. All wings flattened and vi- brated in narrow arc, the hindwings more flattened than forewings and vibrated more rapidly; abdomen raised; scent glands extruded Brief dorso-posterior ning; then return to fanning fan- front Intensification of Stage I Alights beside $ head, moves backward, curves abdomen forward between $ wings, en- gages harpes, swings around to face in opposite direction from 9 Closes wings; lowers abdo- men; withdraws scent glands Heliconius melpomene < Entire Aerial Phase essentially as in Dryas but often omitted > As in Dryas All wings flattened equally and slightly vibrated; other- wise as in Dryas Rear fanning, then return to front fanning As in Dryas As in Dryas As in Dryas Beliconius erato < Entire Aerial Phase essentially as in Dryas but often omitted > Rear fanning Forewings closely appressed; hindwings flattened and quivered; otherwise as in Dryas Front fanning As in Stage I, but lower in- tensity As in Dryas As in Dryas Heliconius ricini ^ Entire Aerial Phase essentially as in Dryas ^ As in Dryas As in H. melpomene Dorso-rear and rear fanning As in Dryas Alights beside 9 abdomen, moves forward, then pro- ceeds as in Dryas. (Rarely alights beside 9 head as in Dryas). As in Dryas Heliconius isabella < Entire Aerial Phase essentially as in Dryas > As in Dryas 1st day: As in H. erato 2nd & 3rd days: As in Dryas As in Dryas As in Dryas As in Dryas As in Dryas Heliconius Sara < Entire Aerial Phase essentially as in Dryas > As in Dryas As in H. melpomene, but wings vibrated strongly through wider arc As in Dryas As in Dryas As in Dryas As in Dryas 1957] Crane; Imaginal Behavior in Butterflies of the Family Heliconiidae 139 Table III. Approximate Ages after Emergence AT WHICH Various Types of Activity Occur in Six Species of Heliconiids Activity Age Flutters to ground when disturbed 1 hr. First flight when undisturbed 2 hrs. First feeding, rarely 6 hrs. First feeding, usually 2nd day Earliest copulation, female 10-30 minutes Earliest copulation, male 3rd day Complete courtship pattern elicited, female 2nd & 3rd days Complete courtship pattern 3rd day— elicited, male 2% months Latest copulation, unmated females 6th-8th day Latest copulation, males (2 species) 2% months First eggs laid 4th- 12th day Maximum ages reached, males and females (2 species) 'iVx months emergence. In Isabella, it is not even possible to see all the specLQc characteristics at one time. During the female’s first day, the wing fiutters of the sedentary phase are very similar to those of erato and the aerial phase is, as usual, normal- ly non-existent; yet by the second day the special- ized flutter is already disintegrating into the more generalized Dryas-type flutter (see Table 2). In all the species after their third day the wing- flutterings of the females are diminished pro- gressively in intensity and characteristic form. In males, copulation does not occur until the third day (48 hours after emergence) and may not take place until the fourth or fifth; these later dates are apparently always characteristic of H. Sara. No social activity by the males what- ever is shown until the second day, 24 hours after emergence, when males sometimes nudge young females from the rear. After noon of the second day slight chasing may occur, but this behavior does not usually take place until the third day. Unlike females, males court and can mate practically throughout life. In older males, al- though complete and successful courtship is swiftly elicited by second-day females, relatively little attention is paid to older unmated or egg- laying individuals. Males chase each other freely throughout life, especially in the absence of young females, ex- cept during their first two days. In both sexes and all species, roosting accord- ing to the species habit begins on the second or third night. Even in the most gregarious species {erato, melpomene and sara), however, young females often hang up for the night alone, unless courting or mating has proceeded during the late afternoon. At such times the female, too, often roosts with the group. VI. Irrelevant Actions A. Males. Under certain conditions the court- ship pattern characteristic of the various species of heliconiids becomes atypical. Instead of court- ship proceeding in the usual fashion to copula- tion or, alternatively, to the point where one partner stops responding and both go their sep- arate ways, the male continues special behavior that never ends in copulation. On the very rare occasions when mating soon ensues between the ^ same partners following the first stages of this irregular behavior, the male has returned to an early stage of courtship, and then followed the typical sequence. Irrelevant actions never begin before the male has reached Stage II of the Sedentary Phase. In other words, he is in the final phase of fanning above, in front of, or behind the alighted female, his position depending on the pattern character- istic of his species (Table 2). Sometimes the male has reached Stage III, having alighted be- side the female and tried unsuccessfully to attach his harpes. One of two major types of irrelevant behavior follows, depending on whether the female flies away and evades the male or whether she stays in place. In the first type, having lost track of the fe- male, the male flies about at unusual speed for up to five minutes at a time, without pausing, feeding or making any apparent “searching” motions (as he may do, on other occasions, among the vegetation) . This type of behavior has been artificially induced by the observer’s simply picking up the female and keeping her tem- porarily out of sight. The second type of irrelevant behavior follows when the female stays alighted and often appears to the observer to be making full courting re- sponses. The first manifestation is always poor orientation in the fanning (PI. I, Figs. 7-8). Normally in all the species the male faces in exactly the same direction as the female, al- though the longitudinal axes of the two insects are usually more or less oblique to each other since the male fans characteristically from down- in-front to up-in-back. In disoriented fanning the male may face in any direction, even backward 140 Zoologica: New York Zoological Society [42: 12 with respect to the female, and frequently changes the direction without any regular se- quence.^ After disoriented fanning has continued for any period up to five minutes in length, there are a number of alternate sequels. Most frequently the partners separate, either the male or the female flying off, and neither one thereafter evinces either excitement or special behavior. Ocaasionally the female takes wing and eludes the male, whereupon the male flies about rapidly as described under the first type of irrelevant behavior. 2In the study of H. erato already published (Crane, 1955), Plate I, Figs. 5 and 6, are excellent examples of disoriented fanning. A third example from the same series is published as Plate I, Fig. 8, of the present con- tribution. At the time these photographs were made, the existence of this type of irrelevant behavior had not been recognized. It is now clear that the sequence of stiU photographs in the earlier plate could not be taken, with present-day still camera apparatus, to illustrate a complete, normal consumated courtship between the same two individuals, photographed in sequence. This is because these courtships always proceed too svsdftly to permit recharging of the speedlamp between flash shots. Only during the prolonged repetitions of irrelevant be- havior can still photographs be made of a single pair during a consecutive period of courting behavior. A mo- tion picture camera, operated at high speed, is of course suitable for the work. In a common sequel to disoriented fanning, the male settles near the female. He may be beside her, back of her or, most frequently, facing her. He then extends the proboscis, some- times uncoiling and coiling it repeatedly and with speed and force; when uncoiled it often touches or even palpates the female’s head, thorax, legs or, rarely, abdomen (Oane, 1955, PI. Ill, Figs. 13-14). Sometimes the male palpates similarly with his antennae. In these cases the proboscis is not uncoiled. Once only a male Dryas, after prolonged dis- oriented fanning, settled obliquely to the side of the female, facing her, and alternately brushed each of his own eyes with the palp of the cor- responding side. The motion was repeated sev- eral times and the performance was repeated an hour later. Occasionally a period of proboscis extension or antennal palpation will be followed by re- sumed disoriented fanning, and then a second period of the more extreme form of irrelevant behavior. Sometimes the male persists in alter- nate disoriented fanning and its sequels until he appears altogether exhausted and remains perched motionless nearby. At these times he cannot even be urged into flight by a sudden close wave of the observer’s hand. Under ordi- nary conditions such a gesture sends into the air even butterflies that are fully accustomed to moving human beings and insectary conditions. Nine situations regularly elicit irrelevant be- ■ havior in courting males. As previously stated, i] an advanced stage of courtship must have been reached before irrelevant behavior begins. 1. Courting of an unmated female, four to eight hours after her emergence; that is, on the afternoon of her first day. 2. Courting of a young mated female. It is now certain that second matings in females are very rare, and are practically confined to indi- viduals that have already reached egg-laying age and are being persistently courted by young males in the absence of young unmated females. 3. Courting of a mated female about to lay eggs. 4. Courting of a male on his first or second day. 5. Courting of a very young imago of either sex, belonging to another species. 6. Loss of an unmated female, during court- ship, through her flying out of sight. 7. Courting of females, unmated or mated, more than three days old by males more than about one month old. 8. Overcrowding of the insectary. This results in somewhat indiscriminate courting activity, apparently clearly attributable to the operation of the principle of heterogeneous summation and resulting in the courting of unsuitable in- dividuals. 9. Unresponsiveness of unmated young fe- males, or other failure in the final stages of courtship, due to unknown causes. B. Females. Equivalent irrelevant behavior of females following broken-off courtships has not been detected. However, later in life a female’s resistance to courtship appears to change in char- acter to such an extent that it may be termed irrelevant behavior. A recrudescence of courtship of egg-laying females is usual in all species. Very rarely these attempts end in copulation. The usual procedure is for the male to chase the female when she is beginning to hover around the foodplant, and to try to force her to alight for the Sedentary Phase in the usual fashion. At these times the female, flying with increasing strength as she is freed of the weight of eggs, usually manages to duck out from under the flying male and fly well above him. Then, instead of flying away as unrespon- sive younger females do, she hovers, mounting higher as he in turn tries to come on top. As 1957] Crane: Imaginal Behavior in Butterflies of the Family Heliconiidae 141 she grows older her flight above him more and more resembles the swift, vigorous flight of the fanning male, in which the fore- and hindwings are well separated (whether or not the friction surfaces in the species concerned hold the scent scales) . The male usually soon gives up and flies away, without irrelevant behavior (see p. 139). As the female grows older she chases passing butterflies with increasing frequency, regardless of their age or sex. Some chasing of other spe- cies, either related or of similar size and color, also occurs. This behavior continues even after egg-laying has ceased and, at least in H. erato, is strongest in those females that have completely finished egg-laying; in that species no eggs are laid after the age of about six weeks, while the female’s imaginal life often extends to more than three months. Only in the last two weeks of life does this energetic chasing decrease. The following special behavior has been ob- served only in females of H. sara. It occurs only on the two days preceding their first egg-laying or, alternatively, when another female has al- ready laid on the only available space. (This species in Trinidad lays eggs in a cluster on the terminal leaflets of Passiflora auriculata). On these occasions sara uncoils the proboscis and with it touches either the spot on which she would normally lay, or the eggs already laid by another female. This procedure never occurs when she is actually ready to lay her own eggs, or when no other eggs are already in place. VII. Discussion A number of points which have emerged from the observations described in the preceding pages now require comment. It is clear that, in this group of butterflies, there can be no accurate description of court- ship or other social behavior that is based on a few observations and individuals, any more than this is possible in the case of a higher vertebrate. The ages and past experience of each butterfly observed, particularly females, must be known in order properly to describe and evaluate their activities. A. Age. As in many birds, the patterns of reproductive behavior are not closely linked with successful reproduction. In the first place, the full specific courtship pattern is unnecessary for copulation. Before pairing with first-day females, males omit all except Stage III of the Second (Sedentary) Phase of courtship, and the young- est females gives no overt responses at all. These exceedingly simple courtships, in which prac- tically all appetitive behavior is almost always omitted, seem to depend predominantly on fe- male scent, as do courtships in moths. All spe- cific differentiation shows in courting females later in the first afternoon and, particularly, on the second or third days. Beyond that age, unmated females are receptive for at least three to six days, although their characteristic actions decline in both intensity and specialization. Courting of very young males by other males, or of young individuals of other species, is un- doubtedly due to the strong family odor of re- cently emerged insects. It is interesting that this odor apparently does not develop until the insect is at least ten minutes out of the chrysalid. Once a male H. sara actually copulated with a female H. melpomene when the latter, aged about one hour, was still unable to fly. Usually these two species, which are strikingly different both in color and, even to the human sense, in odor, have very little attraction for each other. It is unlikely that females, which spend the first day practically inactive, are often found by males during these early hours. Odor alone is not an adequate releaser. A male, after losing sight of a newly alighted young female that has slipped underneath a leaf, sometimes searches around with every appearance of vagueness and inefficiency, and only rarely locates her once again. B. Social Chasing and Related Topics. It now seems clear that all the activities referred to in the paper on H. erato (Crane, 1955) as “social chasing” are appetitive fragments of the courtship pattern. They correspond to instances in numerous other animals where, as in frogs, a male in response to an incomplete stimulus situa- tion embraces another male and, in the absence of appropriate response, releases his grip. It is possible that in the wild some species of heliconiids maintain territories. All that can be said now is that in this family no trace of terri- torial behavior, or of a dominance hierarchy, has been observed either in the insectaries or during field observations. There is no patrolling of definite routes, no special display that lends itself to an interpretation of threat behavior, no overt fighting, no individual that usually is the pursuer in the frequent inter-male chases. On the other hand all of the chases of one male by another can be explained satisfactorily as merely the chasing phase of normal courtship which, because of the inadequacy of the sexual situation, breaks down. The break usually occurs after more or less mutual circling, which exactly resembles the circling resulting from the evasion attempts of unresponsive females (p. 138). When two males are courting the same female. 142 Zoologica: New York Zoological Society [42: 12 the latter often escapes in the general excitement and the two males continue for a short time chasing and fanning each other. Even when, through the breakdown of the releaser sequence, they “discover” the mistake there is no evidence of agonistic behavior; the two simply separate and go their ways. Rarely one or both of the males may continue rapid flight, apparently of an irrelevant character (p. 139). It seems likely that the phylogenetic origin of gregarious roosting was the continuation of social chasing until time to hang up for the night. This could have become flxed, through the action of natural selection, into the stereo- typed patterns of the species, as it came to have protective value. Presumably the strong species odors, which laboratory tests show to be un- pleasant to predators, become intensified through crowding and could be a strong deterrent to nocturnal enemies that depend largely on the sense of smell. These enemies are probably chiefly reptiles and small mammals. The increase of chasing by older females is of particular interest. This masculine type of activity would, in a vertebrate, be subject to a hormonal interpretation, in which the decline of female reproductive hormones leads to visible effects of male hormone activity. However, in insects, morphological sexual characters, at least, are not glandularly controlled; rather “. . . the sex of every part is controlled directly by the chromosome constitution of the cells composing it. Hence the sexual characters are unaffected by the removal of the ovary or the testes or even by their transplantation” (Ford, 1945, p. 192). Schneirla (1953, pp. 677-678) , after citing refer- ences concerning mating behavior in insects that were castrated or otherwise sexually abnormal, commented, “Such results suggest that factors governing susceptibility in male and female in- sects are not directly dependent upon testes or ovaries, but may concern hormones already in the blood ... the activity of some cephalic endocrine secretion, or other physiological agencies such as neural processes.” The present instance, therefore, of pseudo-masculine be- havior in aged female butterflies, is one more instance of the desirability of cooperative studies between physiologists and students of behavior. C. Irrelevant Behavior. Irrelevant behavior proves to play a large part in the apparently normal social life of all butterflies under obser- vation. It is not confined to species in captivity, since it has also been frequently observed in the wild. In fact, if only a few pairs of butterflies of a given species are observed in the Sedentary Phase in the field, it is far more likely that the observer witnesses irrelevant behavior of some kind than a specifically characteristic courtship. The various phases of irrelevant behavior and the situations which produce it have proved to be similar in all six species. In males, it is known certainly to take place only following an ad- vanced stage of an incomplete courtship that is directed toward an unsuitable individual. It occurs in females after their most receptive period (see preceding section). Experiments to determine whether irrelevant behavior takes place in hungry individuals that are prevented from reaching visible and olfactorily detectable food have so far been inconclusive. The instances of irrelevant behavior described are in completely different categories from “so- cial chasing” since they are in no sense merely unfinished portions of the regular reproductive or roosting patterns. Like “social chasing,” how- ever, they never end in copulation. Two of them, irrelevant proboscis-uncoiling and excessive, rapid, undirected flight, can ap- propriately be termed displacement activities. Occurring when the strongly activated sex drive is denied expression through its own consum- matory act, the motions are clearly associated with other patterns of behavior, namely feeding and flight. The once-observed eye-rubbing with the palps may be equivalent to the apparently displaced cleaning motions described in mantids (Crane, 1952), Drosophila (Bastock & Manning, 1955) and ocypodid crabs (Crane, 1957). The two remaining types of irrelevant be- havior in courting male heliconiids cannot be further classified at present. They should not, it seems, be termed displacement activities in the restricted sense (p. 136) since they do not occur in other behavior patterns found within the spe- cies. These motions are disoriented fanning and palpation of the female with the antennae. It may be, of course, that this palpation with the antennae, as well as with the proboscis, produces chemotactic sensations in a situation where the usual olfactory stimuli from the female are likely to be weak or incomplete. In the European satyrid, Eumenis semele, movements of the an- tennae form an integral part of courtship (Tin- bergen et al., 1943). Disoriented fanning, in which the male may face in any direction with respect to the female, appears to be a disintegration of the normal pat- tern, resulting from a breakdown in the usual sequence of male-female responses. Subsequent- ly the pattern often collapses altogether, the remaining energy being channelled into displace- ment-feeding or displacement-flight. 1957] Crane: Imaginal Behavior in Butterflies of the Family Heliconiidae 143 In all irrelevant behavior appearing during courtship, the non-responsiveness of the female, for whatever reason, appears to be the major factor. Males that are not in a physiological con- dition to complete courtship simply break it off at an early stage. As stated in earlier pages, it is now certain that females normally mate only once. Each develops a specifically characteristic odor, given off by the abdominal glands, which becomes ap- parent to the human observer about an hour or more after the butterflies have separated. It is the same odor given off by mated females when seized. It seems likely that it is one of the de- terrents leading to irrelevant behavior when males court mated females and do not follow through to copulation even though the females are making, to the human eye, all the visible responses characteristic of the species. Another deterrent in such cases seems to be the failure of the female to fold the wings, which have been more or less flattened during the courtship flut- tering, above her back when the male alights beside her in Stage III. It is impossible for him to reach her abdomen with his harpes so long as her wings are flattened. Use of the proboscis by H. sara, when egg- laying is apparently thwarted, is no part of the normal egg-laying procedure of any of the mem- bers of the family that have been studied, and perhaps should be termed displacement be- havior. Unlike members of some other families (Use, 1955 and refs.), none of the heliconiids drum the foodplant with their feet before laying. Finally, it may be re-emphasized that almost all these types of irrelevant behavior occur in the lives of all individuals of the appropriate sexes and ages that have been studied. No reliable interspecific differences have yet been observed. VIII. Summary Six species of heliconiid butterflies from Trini- dad, B.W.I., were reared in the laboratory and their post-emergence behavior studied in insec- taries. Their social behavior patterns are briefly compared. Social responses, including courtship, vary with age. In females the full courtship pattern is elicited only on the second and third days, although they can mate successfully when both younger and older. Younger females not only give little or no overt response but draw scarcely any pre-copulatory behavior from males. Older females show progressively fewer specifically characteristic responses. Brief chasing of males or unreceptive females by males of any age and by old females is a com- mon form of activity. It apparently represents simply a fragment of the appetitive portion of the courtship pattern. Neither territorial defense nor inter-male threat behavior seems to be in- volved. Irrelevant actions of a number of kinds are frequent. Most occur in males that are thwarted, usually by the unresponsiveness of the female, after an advanced stage of courtship has been reached. Two kinds of irrelevant actions, a dis- oriented type of wing-fanning and palpation with the antennae, are not known in any normal behavior pattern of the species. Other types of irrelevant behavior appear to be true displace- ment activities, since they occur as appropriate actions in other fields of the insects’ behavior, such as in feeding and flight. IX. References Armstrong, E. A. 1950. The nature and function of displacement activities. Symposia of the Society for Ex- perimental Biology, No. 4. Physiological mechanisms in animal behavior. Academic Press, N. Y., 361-384. Bastock, M., & A. Manning 1955. The courtship of Drosophila melanogaster. Behaviour, 8: 85-111. Bastock, M., D. Morris & M. Moynihan 1953. Some comments on conflict and thwarting in animals. Behaviour, 6: 66-184. Beebe, W. 1952. Introduction to the ecology of the Arima Valley, Trinidad, B.W.I., Zoologica, 37: 158-184. Buckingham, E. 1910. Division of labor among ants. Proc. Amer. Acad. Arts & Sci., 46: 425-507. Crane, J. 1949. Comparative biology of salticid spiders at Rancho Grande, Venezuela. Part IV. An analysis of display. Zoologica, 34: 159-214. 1952. A comparative study of innate defensive behavior in Trinidad mantids (Orthoptera, Mantoidea). Zoologica, 37: 259-293. 1955. Imaginal behavior of a Trinidad butterfly, Heliconius erato hydara Hewitson, with special reference to the social use of color. Zoologica, 40: 167-196. 1957. Basic patterns of display in fiddler crabs (Ocypodidae, Genus Uca). Zoologica, 42: 69-82. Crane, J., & H. Fleming 1953. Construction and operation of butterfly insectaries in the tropics. Zoologica, 38: 161-172. 144 Zoologica: New York Zoological Society [42: 12 Fabre, J. H. 1879. Souvenirs Entomologiques. Paris, Dela- grave. (Ed. def. illustree, 1914-1924. Vol.. 2). Ford, E. B. 1945. The New Naturalist: Butterflies. Collins, London, xiv -f 368 pp. Gordon, H. R. S. 1955. Displacement activities in fiddler crabs. Nature, 176 (4447): 356-357. Hagen, K. S. 1953. A premating period in certain species of the genus Opius (Hymenoptera: Eurytom- idae). Proc. Hawaiian Ent. Soc., 15(1): 115-116. Hinde, R. a. 1953. The conflict between drives in the court- ship and copulation of the chaffinch. Be- haviour, 5: 1-31. Iersel, J. J. a. van 1953. An analysis of the parental behaviour of the male three-spined stickleback {Gas- terosteus aculeatus L.). E. J. Brill, Leiden. 159 pp. Ilse, D. 1955. Behaviour of butterflies before oviposition. Jour. Bombay Natural Hist. Soc., 53: 486-488. Lorenz, K. Z. 1950. The comparative method in studying in- nate behaviour patterns. Symposia of the Society for Experimental Biology, No. 4. Physiological mechanisms in animal be- haviour. Academic Press, N. Y., 221-268. Mayr, E. 1942. Systematics and the origin of species, Columbia University Press, N. Y., viii -f 334 pp. Morris, D. 1954. The reproductive behaviour of the zebra finch {Peophila guttata), with special ref- erence to pseudofemale behaviour and displacement activities. Behaviour, 6: 271-322. Moynihan, M. 1953. Some displacement activities of the black- headed gull. Behaviour, 5: 58-80. Nielsen, E. T. 1945. Moeurs des Bembex. Spolia Zool. Mus. Haun. Copenhagen 7: 1-174. Nielsen, A., & E. T. Nielsen 1952. Migrations of the Pieride butterfly Ascia monuste L. in Florida. Ent. Meddel., 26(5): 386-391. Pardi, I. 1947. Beobachtungen fiber das interindividuelle Verhalten bei Polistes gallicus (Unter- suchungen fiber die Polistini no. 10). Be- haviour, 1(2): 138-172. Rand, A. L. 1943. Some irrelevant behavior in birds. Auk, 60: 168-171. Rockstein, M. 1956. Metamorphosis: a physiological interpre- tation. Science, 123: 534-536. Rosch, G. a. 1925. Untersuchungen fiber die arbeitsteilung im bienenstaat. I. Teil: Die Tatigkeiten im normalen bienenstaate und ihre bezie- hungen zum alter der arbeitsbienen. Zeitschr. Vergleich. Physiol., 2: 571-631. SCHNEIRLA, T. C. 1953. Basic problems in the nature of insect be- havior. (In: Insect Physiology, edited by K. D. Roeder; John Wiley & Sons, Inc., New York, xiv -[- 1100 pp.) . pp. 656-684. Steiner, A. 1932. Die arbeitsteilung der feldwespe Polistes dubia K. Zeitschr. Vergleich. Physiol., 17: 101-152. Tinbergen, N. 1951. The study of instinct. Oxford, Clarendon Press, 228 pp. 1952. “Derived” activities; their causation, bio- logical significance, origin, and emancipa- tion during evolution. Quart. Rev. Biol., 27: 1-32. Tinbergen, N., B. J. D. Meeuse, K. K. Boereme & W. W. Variosseau 1943. Die balz des sampfalters, Eumenia (Satyrus) semele (L.). Zeitschr. Tierpsy- chol., 5: 182-226. Verlaine, L. 1932. L’instinct et I’intelligence chez les Hy- menopteres. Bull. Soc. Roy. Sci. Li&ge, 2: 248-253. 1957] Crane: Imaginal Behavior in Butterflies of the Family Heliconiidae 145 EXPLANATION OF THE PLATE Plate I Figs. 1-6, incl. Species and subspecies of butterflies discussed in this contribution. Photograph by Sam Dunton, New York Zoological Society. Fig. 1. Dryas julia julia. General color; Orange. Fig. 2. Heliconius melpomene euryades. General color: Black; forewing band scarlet. Fig. 3. Heliconius erato hydara. Color as in H. melpomene. Fig. 4. Heliconius ricini insulana. General color: Black; forewing bands pale yellow; central portion of hindwing scarlet. Fig. 5. Heliconius isabellaisabella. General color: Reddish - brown to yellowish - brown; marked with black. Fig. 6. Heliconius sara rhea. General color: Black; forewing bands pale yellow; basal half of hindwing with dark blue irides- cence. Fig. 7. Irrelevant courtship behavior in Dryas julia: Sedentary phase, Stage II, showing poor orientation in fanning. In normal fanning the flying male faces in exactly the same direction as the sedentary female (Table II and p. 138). Note that female has forewings almost closed as required in Stage III. However, the hindwings are still fluttering, as is characteristic of earlier courtship stages, but atypical here. Photo- graph by M. Woodbridge Williams, © Na- tional Geographic Society. Fig. 8. Irrelevant courting behavior in Heliconius erato, corresponding to stage shown in Fig. 7. Photograph by Rosemary Kenedy. CRANE PLATE 1 FIG. 7 FIG. 8 13 Habits, Palatability and Mimicry in Thirteen Ctenuchid Moth Species from Trinidad, B.W.Id William Beebe & Rosemary Kenedy Department of Tropical Research, New York Zoological Society, New York 60, N. Y. (Plates I & II) [This paper is one of a series emanating from the tropical Field Station of the New York Zoological Society, at Simla, Arima Valley, Trinidad, British West Indies. The Station was founded in 1950 by the Zoological Society’s Department of Tropical Re- search, under the direction of Dr. William Beebe. It comprises 200 acres in the middle of the Northern Range, which includes large stretches of undisturbed government forest reserves. The laboratory of the Station is intended for research in tropical ecology and in animal behavior. The altitude of the research area is 500 to 1,800 feet, with an annual rainfall of more than 100 inches. [For further ecological details of meteorology and biotic zones see “Introduction to the Ecology of the Arima Valley, Trinidad, B. W. I.,” William Beebe. (Zoologica, 1952, Vol. 37, No. 13, pp. 157-184.)] Contents I. Introduction 147 II. Material and Methods 148 III. Observation and Results 148 Pseudosphex kenedyae Fleming 148 Saurita clusia (Druce) 150 Histiaea cepheus (Cramer) 151 Macrocneme thyra intacta Draudt 151 Calonotos crane ae Fleming 152 Aethria carnicauda (Butler) 152 Episcepsis hypoleuca Hampson 152 Napata walkeri (Druce) 152 Agyrta dux (Walker) 153 Eucereon maia Druce 153 Eiicereon obscurum (Moschler) 154 Correbidia assimilis (Rothschild) 154 Ctenucha andrei Rothschild 155 IV. Discussion 155 V. Summary 156 VI. References 157 ^Contribution No. 980, Department of Tropical Re- search, New York Zoological Society. I. Introduction The highly specialized moths of the family known as the Ctenuchidae (Syntomidae, Amatidae, Euchromiidae) are among the most interesting and colorful of the Lepidoptera. In general, they are medium to small in size; many are brightly colored, some with metallic luster on wings and abdomen. More than 2,000 species are known, the great majority being re- stricted to the American tropics. Although they are commonly referred to as “day-flying” moths, many species fly also, or even exclusively, at night. Some species “mimic” other insects, and the larvae of some show remarkable adaptations (see especially Beebe, 1953). The taxonomy of this group in Trinidad has been discussed by Kaye & Lamont (1927) and Fleming (1957), but relatively few observations have been reported on these moths in life : flight pattern, position at rest, frequency, time of flight and seasonal occurrence. Also, although it is well known that this family includes species which are obvious wasp and beetle “mimics,” little experimental work has been done on the palatability of these moths to possible predators. For this paper field observations and palata- bility tests were limited to 13 of the more than 80 species so far recorded in Trinidad. The choice of species was based on the following criteria: most are relatively common, therefore more readily available for observation and col- lection; they include both day- and night-flying species; they include genera from the most high- 147 148 Zoologica: New York Zoological Society [42; 13 ly specialized (Pseudosphex) to the most gener- alized {Eucereon, Ctenucha); they include species which show great differences in wing scaling, i.e., completely hyaline (Pseudosphex), completely scaled and having bright patches of color {Histiaea), cryptic fuscous {Episcepsis) and metallic lustrous (Macrocneme); some spe- cies which are obvious “mimics” are included, and some which are drawn to Heliotropium in- dicum and others which apparently are not at- tracted to it. This attractant is reported in detail by Beebe (1955 and refs.) but a brief summary is perti- nent here. Heliotropium indicum Linnaeus is a common weed which, when uprooted and al- lowed to shrivel, exercises a remarkable selective attraction for several lepidopteran families including the Ctenuchidae. For some unknown reason, however, only certain species appear to be attracted to it, while other species common in the same locality ignore it completely. The authors are indebted to Mr. Henry Flem- ing for identification and range of the ctenuchid species used in this study. II. Material and Methods Nearly all observations and collections were made within a radius of approximately a quar- ter of a mile from the Simla field station, the majority within 100 yards of the laboratory. At night they were made at several incandescent electric light bulbs outdoors and at a spotlighted sheet, by day along trails, roadside or in the laboratory itself where a large screened window trapped many of the day-flyers which entered at an open end of the room. H. indicum was used both night and day as a collecting and observing site. The best results were had with this attractant when a clump of the weed was suspended from a low tree branch along a shaded trace. Abundance records used in this paper were kept during the three-year period 1953-1955 but cover only the months of December to June, the extreme limits of our stay at the field station. Rest position was usually observed on a cap- tured uninjured moth placed in a large glass jar or terrarium. Most flight observations were made on free-flying moths at night lights or at H. indicum, or inside a screened insectary mea- suring 12 X 18 feet. As shown in Table 1, the relative acceptability of moths to various vertebrates and invertebrates was tested. In experiments with ants, the ctenu- chids were either freshly dead or completely in- capacitated. No chemicals were used for killing. With all the other testers, living active specimens were used. There is an occasional record of a mantid or a spider capturing a ctenuchid in the field, but all experiments with bird, lizards, frog, tarantula and with most mantids were done with the testers in captivity. The Polychrus lizard and the bird, a Moriche Oriole, were tamed , specimens which readily accepted food held by forceps in the hand. In each test, with the exceptions just men- tioned, an active moth was placed in the ter- rarium or cage with the tester and the response of the latter was noted. Each test was controlled, i.e., a negative result was fisted only if another lepidopteran specimen, usually a pyralid moth or a riodinid butterfly of comparable size, was accepted immediately afterwards. The results of the tests were classified as follows: A— Moth eaten entirely (with occasionally the , obvious exception of wmgs or genitalia) . 1 B— Moth eaten partially or with apparent dis- J taste, i.e., mouthed (by bird or mantid), sucked for short time (spider), partially eaten only (ants, mantid, bird), or eaten with apparent \ distaste (frog). i C— Moth originally noticed (antennaed or dragged by ants, seized by mantid, spider, frog, 1 lizard or bird) but eventually released without | being injured. D— Moth ignored completely or fled from. The numeral following each of the above classifications in Table 1 indicates the number of tests performed with that result. III. Observation and Results Pseudosphex kenedyae Fleming (Plate I, Figure 1; Plate II, Figure 14) Range.— Trinidad. Field Name.— Dusky-winged Wasp-mimic. Field Characters.— Thordix. yellow with dark brown markings; abdomen yellow banded with brown, constricted at base; wings hyaline with fuscous streak along costal margin of forewing. In flight, cannot be distinguished from another Trinidad ctenuchid, Sphecosoma trinitatis Roth- schild, nor from its vespid wasp “models” (see below) . Sexes differ by presence of ventral valve in $. Frequency.— This species was not seen by us until dead Heliotropium indicum was used as an attractant. Since then it has been very common on the weed. The attraction is almost limited to 53, as out of 114 individuals seen or taken, only one was 5. Active Period.— One, of the most decidedly Palatability of Ctenuchid Moths 1957] Beebe & Kenedy: Habits, Palatability & Mimicry in Ctenuchid Moths 149 T3 a? s a cl e S S s S ft Kl 'la ft- -5 ft fei 8 "S ^ Si Eq -e -e 2 S H >1 Ph Q 53 r? im' P Q cq P P (m" c m S' n ft H S Eft O Eo E~I p O O P =« -s S3 S, S' H W ~s ‘S. o o P P (M 53 <; P p p o p <1 o p p . P o < O P OO S’ p p p CO <>1 O CO p p If CO <1 O P P < O (M ^ p P to 05 p ?? -Jl o CO p - - CO CO -H C O P P <1 o p 150 Zoologica: New York Zoological Society [42: 13 diurnal of the ctenuchids, flying, often with its model, from early morning (6:00 to 6:15 A.M.) to nearly dusk. In a terrarium it is invariably positively phototropic. Observed every month from December to June. Rest Position.— Wings 75° back, at times ver- tical. When walking, wings 45° back, 45° up. Antennae in constant motion, often alternately. When pausing between steps, the abdomen is frequently twitched up and down several times. Flight.— Quick, jerky darts, wholly wasplike; sometimes remains suspended before landing, motionless except for the wings, the hindlegs extended downward and slightly backward. It has been seen to alight directly on the underside of a leaf without any hesitation and without alighting on the upperside flrst. Palatability.— (Table 1). In every one of the 22 tests made, this species proved to be unac- ceptable to lizard, bird and three species of ants. In 15 tests the moth was ignored completely; in only seven cases was it approached or noticed. Army ants dragged a living moth by one leg for a few seconds before releasing it, uninjured. In another instance, ants approached a freshly dead specimen, touched it with their anteimae for a short time but ignored it thereafter. In five tests the tame Moriche Oriole started to approach a living moth but backed away almost imme- diately; in one case it pecked at the forceps but not at the insect held by it. In five other tests with the same bird the moth was ignored com- pletely. Mimicry.— There exists a remarkably exact resemblance between this day-flying moth and several of the common Simla wasps. In our first search for this mimicry we captured seven of the wasps and a single ctenuchid. The latter was resting head down on a dry leaf of Heliotropium indicum, while the former were either flying or resting on leaves or inside the laboratory. This particular hymenopteran was Stelopolybia palli- pes var. anceps (Saussure). The two field characters by which the “model” and “mimic” may best be distinguished are the head, that of the wasp being wider and bearing a pair of jaws rather than a proboscis, and secondly, the antennae. The solidity of the an- tennae in the case of the wasp or the pectination in the case of the moth is made more difiBcult for differentiation by the fact that in the moth the pectination comes to an abrupt ending about two-thirds from the base, the distal end being reduced to a slender, easily overlooked filament. Other insects bearing a close resemblance to this moth are Oriodes in Syrphidae and one of the Stratiomyidae, both dipterans. The force of the mimicry is enhanced by the following characters: Size: In total length the wasp is 12 mm., the moth 10 mm. Shape: The abdomen of the moth is proximaUy constricted to a most unmothlike degree. Color and Pattern : | Basic amber color the same, also the thoracic i and abdominal pattern. Antennae: Nervous j twiddling. Walk and Flight: Character, speed I and details of manoeuverability similar in moth | and wasp. When the moth is feeding, its abdo- J men is lifted into a curve and occasionally raised I and lowered, emphasizing the wasplike shape and pattern. Wings: As in the wasp, the wings are held 45° up and out, and the costal area is darker than the remainder of the forewing. An- other remarkable wing character was noted by Kaye (1913) in Pseudosphex spp. He observed that they “curve the forewing while the hind- wing is folded up, so that the wings appear to be as narrow as the completely folded wings of the Vespidae and Eumenidae.” This character applies also to the Trinindad species, and is shown to some extent in Plate II, figure 14. On the whole, the present species of ctenuchid appears to offer an almost perfect example of Mullerian mimicry: the moth is definitely dis- tasteful, and its extremely close hymenopteran counterpart possesses an efficient sting. Saurita clusia (Druce) (Plate I, Figure 2) Range. — French Guiana and the Amazon Region. Field Name.— Black-tipped Whitewing. Field Characters. — Wings yellowish-white, dusky at base with large apical spot. Five red spots, paired on shoulders and collar and one at base of dorsal abdomen. Hindwings short and rounded. $$ with more extensive black scaling at apex and base of forewing and antennae less heavily pectinated, almost threadlike. Frequency.— One of the most common species at Simla, but of the 156 individuals seen or taken, only five have been $?. Active Period. — Strictly nocturnal. Taken every month, December to June. Rest Position.— Wings 60° back, flat, tips clear of surface. When walking the wings are raised 45° and held well apart. Thorax and narrow anterior segment of abdomen exposed, revealing the five scarlet spots. Antennae 45° apart. Flight.— At first, forewings are waved slowly, then all four are whirred, before taking off. Flight is flylike, usually more or less straight upward, the insect finally alighting and walking at once. When liberated from a hand-held con- 1957] Beebe & Kenedy: Habits, Palatability & Mimicry in Ctenuchid Moths 151 tainer in an insectary, the moth seeks shelter beneath fingers or in a dark crevice in the rafters, as if to avoid the light. Twice found rest- ing beneath a leaf. Palatability— 1). This species appears to be acceptable to mantids, moderately so to spider and frog. A specimen was seized by a tarantula but immediately released; later the same specimen was captured and the abdomen sucked for several minutes, but it was eventually released alive and able to walk. Two specimens were eaten by frog with apparent distaste. Two were refused by Moriche Oriole. But as the moth is strictly nocturnal, these negative results with the bird appear to have little or no survival value. Mimicry —1^0 definite models have been ob- served in association with this species. It is flylike in many respects but is strictly nocturnal. Histiaea cepheus (Cramer) (Plate I, Figure 3) Venezuela to Surinam. Field Name.— Large Red-and-yeUow-spotted Brown. Field Characters. — Large (expanse 2 Vi inches), dark brown; forewings with three buffy yellow spots, hindwings with two large, irregu- lar, scarlet spots; basal segments of dorsum with three incomplete yellow bands. Sexes very simi- lar; abdomen of ? slightly broader. Frequency.— Common. Of 81 moths seen or taken, two-thirds were 35. Active Period. — Strictly nocturnal, usually taken at lights before 10 P. M. Twelve imagoes, reared from larvae (Kenedy, unpub.), were placed in an insectary for observation. They were inactive during the day, usually resting, singly, beneath a large leaf. From 6:00 to 6:30 P.M. they began to be active, flying and walking on the screening of the roof. Seen or taken all months, December to June. Rest Position.— Wmg?, 50° back, tips not touch- ing surface. Abdominal bands exposed. The moth has often been observed to wave its wings up and down, through a rather large arc, when standing still or walking, or to whirr its wings rapidly. Flight.— In an insectary, it took off with moth- like flight, the slow beat of the wings rendering them visible. Liberated in the open, it rose swift- ly and smoothly, circled twice in mid-air, then flew straight over the valley. Palatability.— (Table 1 ) . Thirty-four tests were completed, 26 of which demonstrated that this species could be considered unpalatable. A rather unique phenomenon has been observed in this species. When a specimen is immobilized suddenly, whether captured by a caged tester such as a mantid or held by observer, it has been seen in many cases to exude droplets of thick yellow fluid from the thorax. Sometimes the droplets are bubbly and almost completely cover the dorsal thorax. This seems to serve as an effective defense against at least some predators. In two cases mantids bit into the dorsum but dropped the moth when the yellow froth came in contact with their eyes and mouthparts; a lizard five times threw a captured but uninjured insect from its mouth with violent shaking of its head. In the single test with the Moriche Oriole it ate the interior of the thorax and abdomen, leaving the head, genitaliaand entire exoskeleton. In this case, no drops of fluid were observed on the insect’s thorax. A tarantula ignored two specimens, and captured but immediately re- leased two others undamaged. Macrocneme thyra intacta Draudt (Plate I, Figure 4; Plate II, Figure 15) Range.— Trinidad and Colombia. Field iVa/ne.— White-bellied Greenwing. Field Characters.— Wings black, basal half with metalhc blue-green sheen; dorsal base of abdomen with four white spots; hind tarsi white tipped; S white below on thorax and abdomen, $ venter with white spots. Frequency.— Abundant. S $ dominate in num- ber: of 326 seen or taken, only 10 were 99. Active Period.— ChieQy night flyers. Of 326 individuals, 14 (nine $$, five 99) were netted by day. Seen or taken every month, December to June. Rest Position.— W'mgs 20° back, flat, the fore- wings overlapping the hindwings so that the latter are invisible. Antennae 45° apart. When walking, movements are nervous and jerky, wasplike; wings held up and back 30° to 45°; antennae in motion. F/ig/it.— Wasplike, direct. The long hindlegs are held down and backward in flight. When several captured SS were released they spiralled upward fairly slowly or circled several times before flying away. Half of the total 99 collected were netted while they were flying slowly along roadside or near stream, in bright sunshine. Palatability.— (Table 1). Fifty-two tests were made with this species. In 29 the moths were ignored completely and in eight they were origi- nally noticed but uneaten. In five tests with the Moriche Oriole, the bird would not approach the insect; in one case the bird appeared to jump agitatedly around its cage during the several 152 Zoologicar N ew York Zoological Society [42: 13 minutes the moth was crawling inside. Ants ignored freshly dead specimens in all 14 tests. Mantids, however, ate 75 per cent, of the speci- mens offered to them, and frog and Polychrus lizard occasionally accepted the moths. Mimicry— A% noted by Kaye (1913), mem- bers of the genus Macrocneme seem to be ex- cellent mimics of fossorial or pompilid wasps, especially the genera Salius and Pepsis. M. thyra intacta is wasplike both in facies and in behavior. The wings and abdomen possess a metallic blue- green sheen; the long hindlegs, more heavily scaled than is usual in ctenuchids, are extended downward and backward in flight. Kaye also noted that they wave their antennae and vibrate their wings rapidly when alighted, and that they settle usually, like fossorial wasps, on ground, bank or leaf. Calonotos craneae Fleming (Plate I, Figure 5) Trinidad. Field Name.— White-spotted Green-stripe. Field Characters.— Wings black, with two or three white spots on forewing ($$ have two and sometimes a smaller third; $$ have three); a single spot on hindwing; abdomen black with three longitudinal, iridescent, pale green stripes. Frequency.— Common, 78 having been re- corded or collected, with $9 slightly outnumber- ing $$. Active Period.— This species has been taken at Simla only at night lights, chiefly before 10 P.M. However, one specimen was captured by Fleming in central Trinidad in daytime, in palm and immortelle forest. Have been taken Decem- ber to June, with March, April and May the time of heaviest flight. Rest Position.— W'mgs 50° back, 30° raised. Hindwings concealed. Black abdomen exposed, showing conspicuous streaking of iridescent green. Antennae 45° up and out. Relatively slow, direct. On one occa- sion at night light a moth was seen to “play dead” when taken. Palatability. -{Tahie 1). Specimens refused or rejected by bird, frog and tarantula. Ants ate interior of two moths, ignored two others. Eaten by mantids. Aethria carnicauda (Butler) (Plate I, Figure 6) Venezuela and Brazil. Field Va/ne.— Red-tailed Clearwing. Field Characters.— with red-tipped ab- domen; wings hyaline with margins and veins black. with palpi, shoulder spot, forecoxae and ventral base of abdomen white; 99 lack white. This species may be mistaken for Dinia mena (Hiibner) but the latter has the scarlet abdominal fringe extended well up the sides. Frequency.— A common species; 66 were ob- served or collected; 99 slightly outnumbered $$. Active Period.— Conhned to daylight, from 8:30 A.M. to 4:30 P.M. Observed from Decem- ber to June. The moths are more numerous at the end of the dry season, in March and April. Rest Position.— Wings 60° back, flat, inner edges not quite touching. In walking the wings are raised 40°, spread so that the hindwings are exposed; wings in slight but constant motion. When the moth is walking about on the attract- ant H. indicum the antennae are constantly in play, tapping the surface. Apparently more swift and direct than is usual in the group. A moth liberated five times in an insectary flew straight upward toward the light. Palatability.— (Table 1). This species was test- ed with ants and bird only. Refused in single test with Moriche Oriole; ants ate five of 16 specimens. Episcepsis hypoleuca Hampson (Plate I, Figure 7) Range.— Central America. Field Vame.— Red-collared Brownwing. Field Characters.— W'mgs unicolored snuff brown; forecoxae red; dorsal abdomen metallic blue. 33 with white at base of ventral abdomen. Frequency.— Common. Total recorded 128, of which 82 were 33, 34 99 and 12 undetermined. Active Period.— ChieQy nocturnal, both at lights and H. indicum. Approximately 20 per cent, were taken in the daytime; the sex ratio remained constant. Seen December to May, with heaviest flight in December and January. Rest Position.— W'mgs 70° back, not touching surface, overlapping at anal angle, meeting me- dially. Flight.— Slow, slightly wavering until full speed is attained, then direct, not very rapid. Fairly sluggish on H. indicum, crawling slowly in and out of dried branches, sometimes drop- ping instead of flying off when approached. Palatability.— (Table 1). Refused or rejected by bird in three tests, by house gecko in a single test; frog ate one, rejected another; eaten by mantids; ants ate four, ignored six. Napata walker i (Druce) (Plate I, Figure 8; Plate II, Figure 16) 1957] Beebe & Kenedy: Habits, Palatability & Mimicry in Ctenuchid Moths 153 Mexico, Central America. Field Name— Barred-tip Yellow. Field Characters— Wings orange-yellow; apex of forewings white barred with black; palpi, legs and ventral surface black and white. Sexes indis- tinguishable in the field. Frequency.— Taken occasionally at Simla. Of a total of 25 specimens, 22 were $$, three 99. Active PmW.— Nocturnal. Taken from De- cember to June, with the exception of May. Rest Position.— Wings 70° back, just meeting at midline over dorsum; tips not touching sur- face. Flight.— Wavering and not rapid. At end of flight may seek shelter under leaf. Palatability.— Because of the comparative rarity of this species, only one specimen was tested, and that inadvertently. It was a rare 9 which was being kept in captivity in the hope of obtaining eggs. When freshly dead, ants swarmed over it and ate it before it could be salvaged for the collection. M/m/cry.— This orange-yellow ctenuchid is quite unlike other members of its family, but there are several species of other families which in color and pattern might be considered as models. Among these are the riodinid butterfly, Mesene semiradiata, and the zygaenid moth, Malthaca radialis. A pyralid moth, Mapeta xan- thomelas, fulfills all the demands in numbers and appearance of a model, although it is a day- flying species. At rest the ctenuchid and the pyralid are remarkably similar, and not until one is close enough to see the backward angle of the antennae and the prominent palpi of the pyralid can they be distinguished (see Plate II, fig. 17) . In flight, however, the barred hindwings of the pyralid are quite visible; the ctenuchid lacks these markings. Two of the pyralids were tested for palata- bility with ants and were eaten except for the shell of the thorax. However, two specimens were ignored by the Moriche Oriole. Agyrta dux (Walker) (Plate I, Figure 9) Range.— Central America to Venezuela and Brazil; St. Lucia. Field Vume.— Six-windowed Black. Field Characters. — Relatively large moth, black shot with blue; wings with six large, ir- regular hyaline areas, two in forewing, one in hindwing, crossed by black veins; wing bases, thorax and abdomen iridescent blue; palpal base and tongue scarlet; narrow dorsal abdominal streak and venter white. Sexes similar except for slightly shorter antennal pectinations in 9. Dif- fers from closely related A. micilia by hyaline forewing area arising near base. Frequency.— OccasionaBy recorded at Simla. A total of 23 consisted of 19 three 99 and one undetermined. Active Period.— Exclusively diurnal, taken from early morning to late afternoon, with the majority seen in late afternoon. December and January have been the best months for recording this species, but they have also been taken in March, April and May. Rest Position.— Wings 20° to 40° back, meet- ing over back. When walking, wings are waved through a 45° arc; antennae in constant motion, moving in unison. Flight.— Fairly rapid, darting, low, circling before alighting. It was seen to alight on dead H. indicum three times at the same spot. At once it began to search eagerly with its tongue, within a small area. After feeding on a panicle it liter- ally ran through the dense, dry foliage, its wings rubbing against the leaves in the narrow places. Its speed was remarkable in the deeper inter- stices. The wing iridescence and the deep red collar and tongue could be plainly seen when the rest of the insect was invisible. Palatability .-{Table 1). This species is ap- parently palatable to ants. A small ponerine ant was once seen carrying off a 9 specimen, holding it at the tip of the abdomen. The insect was rescued, but died. The Moriche Oriole twice picked up a specimen hesitantly, but dropped it without injury. Mimicry.— An unidentified geometrid moth is a close model for the species except for its smaller size, spreading only 28 mm. whereas A. dux extends 40 mm. between wingtips. The rest posi- tion of the geometrid is similar to that of the ctenuchid and it also is a day-flyer. Both in facies and flight, A. dux appears more butterfly- like than is usual in the group. Eucereon maia Druce (Plate I, Figure 10) Range.— Mexico, Central America, Venezuela, British Guiana. Field Name.— Red-bodied Brown. Field Characters.— Tborax and forewings light brown, mottled with darker; hindwings pale; ab- domen red above, dotted with black, whitish below. Forewings of 9 paler. Frequency.— of this species are abundant. Total of 260 seen or taken consisted of 196 12 99 and 52 undetermined. 154 Zoologica: New York Zoological Society [42: 13 Active Per/oci.— Nocturnal, flying throughout the night, appearing at dusk and seen also at dawn with records every hour to 3 A.M. The peak of the flight seems to be 10:30 to 11:30 P.M. In an insectary the moth is negatively pho- totropic. Recorded every month from December to June, with December and January the months of heaviest flight. Rest Position— 70° back, overlapping, completely covering abdomen. Flight.— Rapid and direct, occasionally cir- cling once or twice. When disturbed, the moth often drops to the ground, beetlelike, before tak- ing flight. Or it may crawl along the ground instead of flying, sometimes hiding under a low- growing leaf. The habit of dropping is so usual that a specimen can almost always be collected merely by placing the open end of a vial be- neath it. Palatability.— (Table 1). Ants ate seven speci- mens, ignored five; spiders sucked three; one was captured and eaten by a centipede on H. indicum. In majority of tests the frog captured but spit out this species. The Moriche Oriole ate one, ran two others through its bill, then flicked them away, injured. Eucereon obscurum (Moschler) (Plate I, Figure 11) Mexico and Central America south to the Amazon. Field Name.— Blue-spangled Bronzewing. Field Characters.— Abdomen black, the termi- nal segments shot with blue; forewings bronze- black, faintly spangled and lined with bluish- white; hindwings semi-hyaline, pale brown. Sexes indistinguishable in the field. Frequency.— A common species. Of 87 speci- mens 58 were $$, 10 52 and 19 undetermined. Active Exclusively nocturnal, like others of the genus. It first appears in early evening and has been found as late as 3 : 30 A.M. Has been taken from December to June, with the heaviest flight in December and February. Rest Position.— Wings 80° back, almost par- allel; inner edges overlapping, concealing body; tips resting on surface. Flight.— A rapid fluttering, but rather slow flight. The moth circles slowly, hovers before alighting. When released in insectary, it invari- ably flew to the ground and hid among grass. Palatability.— (Table 1). Eaten by ants, man- tids and frog; Moriche Oriole ate head and thorax of one, seized four others but dropped them immediately. Correbidia assimilis (Rothschild) (Plate I, Figure 12; Plate II, Figures 18, 19) Venezuela, British Guiana, Surinam, Brazil. Field ATamc.— Yellow-banded Beetle Mimic. Field Characters.—Small moth. Wings black, crossed in center by a wide, pale yellow band. Sexes indistinguishable in the field except by slightly shorter antennal pectinations in 2. Frequency.— The most abundant ctenuchid at Simla. Total recorded 337, of which 205 were 55, 30 22 and 102 undetermined. Active Period.— More than 95 per cent, were seen or taken at night, being drawn chiefly to lights but also to uprooted H. indicum. Occa- sionally taken during the day and then usually in the early morning or late afternoon, within two hours of dawn and dusk. Seen from Decem- ber to May. The numbers drop off in April and May. Rest Position.— Wings 80° back, outline curved, overlapping medially, only thorax ex- posed. In this position the light yellow midwing markings form a continuous transverse band. Antennae 45° forward and out, tips slightly curved. When walking the moth waves its an- tennae alternately up and down. This species appears to be unusually lethargic and slow to take flight. Flight.—Slow and fluttering. Palatability.— (Table 1). Ants ate three speci- mens, ignored two; eaten by mantids, spider, frog and lizard; Moriche Oriole ignored three, approached one hesitantly but would not take it. M/m/cry.— Several authors, especially Kaye (1913), have commented on the remarkable resemblance that exists between members of this genus and lycid beetles, such as the genus Lyco- morpha in North America. Lycid beetles are general in the Neotropics and are known to be protected (Forbes, 1930, p. 27). There are three Simla insects which might be considered as models for C. assimilis: a small arctiid moth, Lycomorphodes aracia, and two lycid beetles, a smaller and a larger species (Plate II, figs. 18, 19). The arctiid and the smaller beetle are noc- turnal, the larger beetle is a day-flyer. The smaller lycid is common and frequently found associated with the ctenuchid. The larger beetle and C. assimilis are similar in color and pattern and in their slow, lethargic movements. They also resemble one another in appearance and movements of the antennae, and in the mutual habit of slowly lifting and lowering the forewings (elytra in the case of the beetle) as 1957] Beebe & Kenedy: Habits, Palatability & Mimicry in Ctenuchid Moths 155 they walk. The average in both is seven waves in five seconds. During the same length of time 54 arctiids, 37 lycids and 130 specimens of the ctenuchid were recorded. Kaye (1913) made several interesting obser- vations on mimicry in Correbidia which we have also noted in C. assimilis in Trinidad: they are sluggish in the early morning and, beetlelike, drop to the ground when disturbed, drawing their legs in; the shape and color of wings are like a lycid beetle. To heighten the deception, the abdomen of the moth is slightly flattened, the legs are short and the heavy pectinations of the antennae are carried to the tip. To check palatability, the two beetles and the arctiid were tested with ants. Three specimens of the arctiid were eaten; one large and two small beetles were untouched. Ctenucha andrei Rothschild (Plate I, Figure 13) Range.— been recorded also from British Guiana. Field Name.— Large White-banded Black. Field Characters.— A good-sized moth; fore- wings blue-black with broad, oblique, white band; hindwings steel blue. $$ with white on venter, ?S black. Frequency.— $$ common, $$ rare. Of 89 speci- mens, 63 were S3, 2 59 and 24 undetermined. Active Period.— A diurnal species, flying at any hour of the day from dawn to dusk. A single exception was a $ taken at the night light in early evening. Observed from December to June. Rest Position.— Wings 60“ back, meeting over dorsum; white bands on forewings not meeting. When walking the moth waves its wings slowly through a 45° arc. Antennae 70° forward, 20° apart. Flight.— Rather slow and fluttering. In flight the wings become a black haze, with curved ring of white above and below. Wings move through an arc of 130°. Palatability. — {Table 1). This species was eaten by mantid, tarantula, frog and lizard. In 60 per cent, of the tests with ants the moth was eaten. Army ants ignored one specimen, killed but did not eat another. In two tests the Moriche Oriole took the moth but dropped it imme- diately; it ignored four others. Mimicry.— The most reasonable “model” for this species is Cecropterus bipunctatus (Hes- periidae). In abundance, pattern and color, gen- eral motions and habitat this approximates the ctenuchid. However, this skipper was found to be acceptable to the Moriche Oriole. IV. Discussion Table 2 is a recapitulation of data, most of which are already given in the text under in- dividual species, with the addition of positive or negative results with the use of Heliotropium indiciim as an attractant. These results bring out several interesting problems: although the at- traction of the weed is highly selective, being almost exclusively limited in the Heterocera to the family Ctenuchidae, yet some of the most common Trinidad ctenuchids, such as Macroc- neme thyra intacta, have never been found on it. Pseudosphex kenedyae is a common visitor to FI. indicum, but it has not been seen elsewhere nor had we ever seen a specimen before using the attractant. Other ctenuchids are common both at lights and on the weed, such as Corre- bidia assimilis and Eucereon maia; however, an- other member of the same genus, E. obscurum, is fairly common at lights but extremely rare on H. indicum. Frequency.— The species are listed in order of abundance in Table 2. The sex ratio of 11 species shows a predominance of males. How- ever, females slightly outnumber males in Calonotos craneae and Aethria carnicauda. As the latter flies by day and as the larval foodplant is a low-growing sedge (see Beebe, 1953) and as eggs are laid singly (Kenedy, unpub.), it is perhaps reasonable to suppose that ovipositing females would be seen and netted more often than males. But in the case of C. craneae all the females seen or taken at Simla were attracted to night lights. Active Period.— Six species were found to be exclusively nocturnal and four exclusively di- urnal; three species were active both by day and at night, although chiefly nocturnal. Rest Position.— The backward angle of fore- wings at rest ranges among the 13 species from 20° in Macrocneme thyra intacta to an almost parallel 80° in Eucereon obscurum and Corre- bidia assimilis. The wasp “mimics,” M. thyra intacta and Pseudosphex kenedyae, hold their wings at a slightly upward angle when walking; the other species hold them either parallel to the surface or angled slightly downward so that the apex is almost touching the surface. Flight.— There is great variation in habits of flight, from slow and wavering {Correbidia assimilis, Episcepsis hypoleuca) to rapid and direct {Aethria carnicauda, Eucereon maia). Palatability and Mimicry.— In Table 1 the re- sults of palatability tests show that the only 156 Zoologica: New York Zoological Society [42: 13 Table 2. FREQtrENCT, Active Period and Attraction to Heliotropium indicum Species are listed in order of abundance. Totals are for three-year period 1953-1955, principally during January to April. Extreme limits of observation and collection were December to June. Species Frequency Active Period Attraction to H. indicum Total Seen or Taken % Sex Ratio % 9 9 % Un- determined % Diurnal % Nocturnal Conebidia assimilis 337 61 9 30 5 95 + Macrocneme tkyra Intacta 326 97 3 4 96 - Eucereon maia 260 75 5 20 100 -1- Saurita clusia 156 97 3 100 - Episcepsis hypoleuca 128 64 27 9 20 80 -1- Pseudosphex kenedyae 114 >99 <1 100 + Ctenueha andrei 89 71 2 27 100 -1- Eucereon obscurum 87 67 11 22 100 -f (rare) Histiaea cepheus 81 68 22 10 100 - Calonotos craneae 78 45 49 6 100 - Aethria carnicauda 66 45 55 100 -t- Napata walkeri 25 88 12 100 + (rare) Agyrta dux 23 83 13 4 100 -1- species which appears to be well protected from potential enemies is the vespid wasp “mimic” Pseudosphex kenedyae and therefore is an ex- cellent example of Miillerian mimicry. Although tests with the other species are inconclusive, it seems that they are at least partially protected against possible predators. V. Summary Field characters, sex differences, frequency, active period, rest position and flight are re- corded in 13 of the more common species of ctenuchid moths in Trinidad. Interesting habits include extremely wasplike flight as well as facies in two species, and in two others a beetlelike dropping to the ground when disturbed. In two species only, more females were seen or taken than males; in the others, males far outnumbered females. There is great variation in habits of flight and in rest position among the 13 species. Their palatability to various vertebrates and invertebrates was tested under controlled condi- tions. They appear to be at least partially un- acceptable to possible predators. It is suggested that a fluid exuded from the thorax of Histiaea cepheus serves as an effective defense. Mimicry in several species is discussed. Pseudosphex kenedyae appears to be an almost perfect example of Mullerian mimicry. 1957] Beebe & Kenedy: Habits, Palatability &. Mimicry in Ctenuchid Moths 157 VI. References Beebe, W. 1952. Introduction to the ecology of the Arima Valley, Trinidad, B.W.I. Zoologica, 37: 157-184. 1953. A contribution to the life history of the euchromid moth, Aethria carnicauda But- ler. Zoologica, 38: 155-160. 1955. Two little-known selective insect attraa- ants. Zoologica, 40: 27-32. Fleming, H. 1957. The Ctenuchidae (moths) of Trinidad, B.W.I. Zoologica, 42: 105-130. Forbes, W. T. M. 1930. Insects of Porto Rico and the Virgin Islands. Sci. Surv. Porto Rico and Virgin Isl. (New York Acad. Sci.), 12 (1): 1-171. Kaye, W. J. 1913. A few observations on mimicry. Trans, ent. Soc. London, 1913: 1-10. Kaye, W. J., & N. Lamont 1927. A catalogue of the Trinidad Lepidoptera Heterocera (moths). Memoirs of the De- partment of Agriculture, Trinidad and Tobago. No. 3. 158 Zoologica: New York Zoological Society [42: 13: 1957] EXPLANATION OF THE PLATES Plate I Fig. 1. Pseudosphex kenedyae. Fig. 2. Saurita clusia. Fig. 3. Histiaea cepheus. Fig. 4. Macrocneme thyra intacta. Fig. 5. Calonotos craneae. Fig. 6. Aethria carnicauda. Fig. 7. Episcepsis hypoleuca. Fig. 8. Napata walkeri. Fig. 9. Agyrta dux. Fig. 10. Eucereon maia. Fig. 11. Eucereon obscurum. Fig. 12. Correbidia assimilis. Fig. 13. Ctenucha andrei. Plate II All photographs are of living specimens in natural rest position, taken in a confined area. Fig. 14. Pseudosphex kenedyae feeding on Helio- tropium indicum seed panicle. Note wasp- like pattern, constriction of abdomen and curve of forewings; also reduction of pec- tinations on distal third of antennae. Fig. 15. Macrocneme thyra intacta, showing long, heavily scaled hindlegs and white tarsi. Fig. 16. Napata walkeri. Fig. 17. Mapeta xanthomelas (Pyralidae). Fig. 18. Correbidia assimilis and small lycid beetle. Fig. 19. Same, with large lycid beetle. BEEBE a KENEDY PLATE I FIG. 3 FIG, 4 FIG. 5 FIG. 6 FIG. 7 FIG. 9 HABITS, PALATABILITY AND MIMICRY IN THIRTEEN CTENUCHID MOTH SPECIES FROM TRINIDAD. B.W.I. BEEBE a KENEDY PLATE II FIG. 14 FIG. 16 FIG. 15 FIG. 17 FIG. 18 FIG. 19 HABITS, PALATABILITY AND MIMICRY IN THIRTEEN CTENUCHID MOTH SPECIES FROM TRINIDAD, B.W.I, 14 Serological Relationships among Members of the Order Carnivora^ Ludwig K. Pauly- & Harold R. Wolfe Department of Zoology, University of Wisconsin, Madison, Wisconsin The serological technic has been used in taxonomic studies for more than fifty years. The discovery of precipitins by Krause in 1897 and the publication in 1904 of Nultall’s book applying the new technic to problems of animal relationships gave taxonomists an ap- proach which showed great promise of clarify- ing disputed or undetermined relationships. The precipitin technic has been used to a limited ex- tent for the latter purpose, but has been quite extensively applied for verification of existing relationships based on morphological criteria. The usefulness and trustworthiness of the tech- nic have been shown by many workers. Studies have chiefly been made with vertebrates but there has also been research with a few of the invertebrate phyla. The following is a brief list of some representative papers in the field: Boyden (1926, 1934, 1943), Boyden & Noble (1933), Wolfe (1936), Brown & Helfron (1928), Eisenbrandt (1938), Wilhelmi (1940), Martin & Cotner ( 1934) , Baier & Wolfe (1942) , Gemeroy (1943) and Leone & Pryor (1954). It seems to us that the greatest value of the sero- logical technic now lies in its possibility of set- tling questionable animal or plant relationships. Perhaps no other order of the Class Mam- malia contains such a diversified group of ani- mals as does the Order Carnivora. This diversi- ty has been responsible for considerable conflict and uncertainty in attempts at classification. The problem lies not so much in the over-all picture of Carnivora classification as in the grouping of related forms in the suborders and superfamilies. Most authors seem to agree that the order should ^Supported in part by the Research Committee of the University of Wisconsin Graduate School from funds supplied by the Wisconsin Alumni Research Foundation. ^Present address: University of Wisconsin, Milwaukee. be divided into two suborders: the Fissipedia and the Pinnipedia. There are disagreements as to the relationships and inter-relationships among the families of the Fissipedia and the re- lationships of these families to the Pinnipedia. Winge (1923-24) and Scott (1937) both seem to favor combining the Ursidae, Canidae and Procyonidae into one group, with the Musteli- dae as a distinct but associated family, and the Viverridae and Hyaenidae into another group, with the Felidae as a distinct but associated family. Winge associates the Pinnipedia with the Ursidae and Canidae. Beddard (1902), Romer (1933) and Simpson (1945) favor combining the Canidae, Ursidae, Procyonidae and Mus- telidae into one superfamily and the Viverridae, Hyaenidae and Felidae into another superfamily. Beddard apparently believes the Pinnipedia closely associated with the Mustelidae, whereas Romer seems to indicate that the Pinnipedia are most closely associated with the Ursidae and Canidae. Other authors have opinions varying slightly from the above. Since their discovery in the late Nineteenth Century, little use has been made of serological reactions in the taxonomic study of the Carni- vora. Nuttall (1904) summarized the results of his flocculation tests with the sera of 56 different species of Carnivora. In general his results agreed with the systematic position of the species tested. With this lone exception no other exten- sive work has been done on the serological taxonomy of the Carnivora. Brief mention of them is made in several papers: Boyden (1926, 1942), Boyden & Gemeroy (1950) and Wolfe (1936). An extraordinary study on the taxonomy of the Carnivora was included in the book by Reichert & Brown (1909) in which the rela- tionships between various members were deter- 159 160 Zoologica: New York Zoological Society [42: 14 mined by the resemblances and differences of hemoglobin crystals formed from the different species. On the basis of their study Reichert & Brown claimed that the Ursidae and Mustelidae showed a closer relationship to the Pinnipedia than did any of the other Fissipedia. A peculiar finding was that the hemoglobin crystals of the skunk more nearly resembled those of the Pro- cyonidae, which in turn did not resemble those of the Pinnipedia. The hemoglobin crystals of the Canidae resembled those of the Pinnipedia less, while the crystals of the Felidae and Viver- ridae resembled those of the Pinnipedia least of all. The classification and nomenclature used in this paper follow those of Simpson (1945) as much as possible. Materials and Procedures Many of the blood sera used as antigens were received from the New York Zoological Park through the courtesy of Dr. L. J. Goss and from the San Diego Zoo and the Serological Museum of Rutgers University through the courtesy of Drs. C. R. Schroeder and Alan A. Boyden respectively. Table 1 presents a list of animals from which sera were obtained. Both chickens and rabbits were used for the production of antisera. Three different technics of precipitin testing were employed. These were the ring (inter facial) test, the photronreflectom- eter method of Libby (1938) and the micro- densitometer method of Baier (1943). The latter two are turbidimetric methods; the pho- tronreflectometer measures scattered light pro- duced by the flocculating particles and the microdensitometer measures transmitted light. The authors found it advisable to use different injection procedures in order to obtain antisera of different precipitating ability, for it was neces- sary to have quite heavy precipitates when the microdensitometer was used, and weaker pre- cipitating sera when the photronreflectometer was employed. With one exception, all antisera used in the ring tests were produced in chickens. Each chicken was given a single intravenous inocula- tion of 1 ml. of a 2 per cent, solution of blood serum (the antigen). This injection procedure is the best for production of antiserum of low precipitating power, a high interfacial titer and good specificity (Wolfe, 1936). The birds were bled eight to ten days after the injec- tion. The antisera were allowed to stand for at least seven days in the refrigerator before use, as in vitro changes occurred in the serum upon such standing (Wolfe, 1942). The one rabbit used was treated in the same manner Table 1. Animals Used in Study Order CARNIVORA Suborder Fissipedia Family Canidae Canis familiaris (dog) Canis lupus (timber wolf) Vulpes fulva (red fox) Family Ursidae *Ursus americanus (black bear) ■• Thalarctos maritimus (polar bear) Family Procyonidae Procyon lot or (raccoon) *-'Potos caudivolvuliis (kinkajou) **Nasua narica (coati-mundi) Family Mustelidae Mustela furo (ferret) Mephitis mephitis (skunk) Taxidea taxus (badger) Mustela vison (mink) ■•■Tayra sp. (tayra) Family Felidae ■•'Felis concolor (mountain lion) Felis doinesticus (house cat) Panthera pardus (leopard) *Panthera tigris (tiger) *Acinonyx jubatus (cheetah) Family Hyaenidae Hyaena hyaena (striped hyaena) Suborder Pinnipedia Family Otariidae ■■■**Eiimetopias jiibata (Steller’s sea lion) ***Zalophus californianus (California sea lion) '^Zalophus californianus (California sea lion) Family Odobenidae *Odobenus rosinanis (walrus) Family Phocidae *-'*Phoca vitulina richardii (harbor seal) Order ARTIODACTYLA Family Bovidae Bos taurus (cattle) Bison bison (American bison) Order PRIMATES Family Hominidae Homo sapiens (man) *Sera furnished by Dr. L. J. Goss. **Sera furnished by Dr. A. A. Boyden. * * •^■'Sera furnished by Dr. C. R. Schroeder. All other sera collected from local sources, in- cluding Madison Zoo. 1957] Pauly & Wolfe: Serological Relationships among the Carnivora 161 described for production of antisera used in the photronreflectometer method. Both chiekens and rabbits were used in the production of antisera for the photronreflectom- eter studies. The rabbits were given a single series of three intravenous injections on alter- nating days, totaling 3 ml. of undiluted serum; the first injection was 0.5 ml., the second 1 ml. and the third 1.5 ml. The chickens received the same number of injections and on similar days but the solution was a 2 per cent, solution of the antigen rather than undiluted serum. The rabbits and chickens were bled on the seventh day after the last injection. The antisera used in the microdensitometer studies were produced in chickens and rabbits. The increased precipitating power of the anti- sera needed in such studies was produced by increasing the amount of antigen inoculated into the animals. The rabbits were given two more series of three injections each at an interval of approximately 30 days. Each series consisted of a total of 3 ml. of undiluted serum. The chick- ens were given only one injection series, totaling 3 cc. of undiluted serum in three injections. The rabbits and chickens were bled seven days after the last injection. It should be emphasized that the above injection procedures did not always result in antisera of sufficient potency, and rather than reinject the animals that were poor antibody producers they were discarded. Chick- ens were found to be much better producers of antibody than the rabbits. The chickens and rabbits were starved for 18 to 24 hours before bleeding. The blood was removed by cardiac puncture and allowed to clot. The serum was removed after centrifugation and stored in the refrigerator. The ring test was performed in a 1.8 per cent, saline solution when chicken antiserum was used and in a .9 per cent, buffered saline so- lution when rabbit antiserum was employed. Serial dilutions of the antigen were made from a 2 per cent, solution which in turn was made from the undiluted antigen. One-tenth of a milliliter of antisera was layered below the anti- gen solutions (.5 ml.). Readings were made at 5, 10, 20, 30 and 60 minutes, but only the 60- minute readings are recorded in this paper. The test antigens employed in the photron- reflectometer and microdensitometer studies were also serially prepared. The final reaction mixtures of antigen and antisera were approxi- mately .9 per cent, for the rabbit system and 8 per cent, for the chicken system, since Goodman, Wolfe & Norton (1951) showed that 8 per cent, was the optimum for the chicken antiserum system. In order to conserve anti- serum, only alternate dilution tubes were used in the microdensitometer and photronreflectom- eter tests. The photronreflectometer tests were conduc- ted according to a procedure modified slightly from that outlined by Baier (1947). The cells used were standardized as to thickness. This means that the light beam passes through the same distance in the liquid of each set of cells. This is important in measuring the light-scat- tering effect of particles in a suspension. Com- parative results cannot be obtained if this dis- tance varies among the individual cells of each set. The microdensitometer tests were conducted according to the procedure outlined by Baier ( 1 947 ) . The tubes used in these tests were stand- ardized according to diameter and transmission of light beams. The photronreflectometer and microdensitom- eter tests were made with antigen dilutions so chosen that the final readings for the highest and lowest dilutions (antibody excess and anti- gen excess) equalled the control reading at those two points. This was not always possible at the antigen excess region because of the small amount of antigen that was available in several cases. Values of the relationships are expressed in percentage. The homologous reaction is con- sidered to be 100 per cent, and the heterologous reactions are related to this. The galvanometer readings were summated for all the antigen concentrations used when the photronreflectom- eter and microdensitometer technics were used. Results Table 2 presents a summary of the results obtained with 15 different antisera using the ring test technic. The two anti-Mustelidae sera (anti-mink and anti-ferret) gave high cross-re- actions not only with other mustelids but also with the dog, fox, black bear and raccoon. The degree of cross-reactions with members of the Felidae were considerably lower in two of the three tests made and the reactions with Bovidae and Hominidae were very low. Only one Canidae antiserum was produced. The anti-red fox serum reactions indicated that the Ursidae and Procyonidae were more closely related to the Canidae than to the Mustelidae or Felidae. The anti-black bear serum was a very specific serum and gave a high cross-reaction only with the raccoon serum. Much weaker reactions oc- curred with other families of Fissipedia and also with man. 162 Zoologica: New York Zoological Society [42: 14 Table 2. Serological Relationships among the Carnivora — Ring Test Technic. Relationship Values Expressed in per cent. Vh yi C < Antigen Source Mink-67 (12800)* Ferret- 107 (51200) Fox-165 (25600) Black bear-162 (12800) Cat-65 (51200) Leopard-139 (25600) Mountain lion-111 (25600) Kinkajou-PC-10 (51200) Coati-mundi-PC- 1 0 (51200) Raccoon-C-1416 (51200) Steller’s sea lion-C-1 55 (51200) Steller’s sea lion-C- 1 63 (51200) } Walrus-PC-88 (6400) Harbor seal-C-165 (25600) Harbor seal** 12800 Suborder Fissipedia Family Procyonidae Kinkajou 100 6.3 6.3 Coati-mundi 100 6.3 Raccoon 50 100 25 25 6.3 2.3 50 50 100 12.5 0 25 0 0 Family Ursidae Polar bear 12.5 50 6.3 12.5 25 0.4 Black bear 100 25 100 3.1 6.3 100 50 6.3 25 0.8 Family Canidae Red fox 50 25 100 1.6 0.4 12.5 25 0 25 0 Dog 25 50 50 3.1 6.3 1.6 0.8 0 Family Mustelidae Ferret 100 100 12,5 4.7 1.6 25 Skunk 50 50 3.1 3.1 12.5 3.1 0.4 6.3 12.5 50 12.5 1.6 12.5 0 Mink 100 100 3.1 1.6 1.6 Badger 50 50 3.1 3.1 3.1 3.1 1.6 Family Felidae Cat 9.4 18.8 3.1 1.6 100 12.5 50 6.3 0 Leopard 3.1 3.1 100 Mountain lion 25 6.3 100 50 100 Suborder Pinnipedia Family Otariidae Steller’s sea lion 50 100 25 100 100 50 3.1 25 Family Odobenidae Walrus 25 50 25 50 100 3.1 25 Family Phocidae Harbor seal 25 50 25 50 100 100 Order Artiodactyla Cattle 6.3 3.1 0 1.2 0.8 3.1 0.8 0 0 Bison 6.3 0 1.6 6.3 0 Order Primates Man 6.3 0.9 0.8 3.1 3.1 2.3 * Ring test titer. ** Produced in rabbit; all others produced in chickens. Three anti-Felidae sera were tested. All of these gave strong cross-reactions with the cat, leopard and mountain lion, but a low order of reactions with the members of other families. The intra-family results were very peculiar in one test. The reactions of the anti-leopard serum indicated that the leopard was more closely re- lated to the mountain lion than to the house cat. Such unexpected results warrant further investigation. The three Procyonidae sera were all quite aspecific in their cross-reactions. All these sera gave large reactions with the Pinnipedia and the anti-kinkajou serum indicated that these Pin- nipedia were more closely related to the kin- kajou than were the Mustelidae, Canidae and Ursidae. The one test made with a representa- tive of the Felidae showed the cat to be more distantly related to the raccoon than to other Fissipedia. 1957] Pauly & Wolfe: Serological Relationships among the Carnivora 163 Table 3. Serological Relationships among the Carnivora-Microdensitometer and Photronreflectometer Technics. Relationships Expressed in per cent. Microdensitometer Photronreflectometer ci u o C c Antigen Source *Steller’s sea lion-157 *Steller’s sea lion-CU Steller’s sea lion-3698-3796 *Black bear-3659-3660 1 *Raccoon-CX 1 Steller’s sea lion-5 UJ Oh tJL "o pq C/2 * *Cat-PA Suborder Fissipedia Family Procyonidae Kinkajou Raccoon 11.6 42 23.4 49.2 34.9 65.7 100 18.2 16.7 0 1.37 0.7 Family Ursidae Polar bear Black bear 34.9 26.7 39 35.9 55.6 59.5 93 100 53.6 26.9 15.7 18.4 21.6 18.8 15.9 11.2 0 22 Family Canidae Red fox Dog Timber wolf 20.5 13.4 47.3 38.4 18.6 2.9 52.5 88.1 100 5. 0 0 1.3 Family Mustelidae Skunk Ferret Tayra 30.7 55.3 45.3 37.3 20.1 13.1 12.1 100 11.3 15.1 2 1 Family Felidae Cat Tiger Mountain lion Cheetah 6.1 0 1 3.2 100 50.4 47.8 76.6 Family Hyaenidae Hyaena 0 18.3 Suborder Pinnipedia Family Otariidae Steller’s sea lion Cal. sea Uon 100 100 100 45.3 32.7 100 82.6 9.3 6 0 .9 Family Odobenidae Walrus 45.5 8.9 8.1 Family Phocidae Harbor seal 81.5 52.4 76.6 44.2 47.4 9 10.3 Order Artiodactyla Cattle 14.4 17.7 5.2 4.5 * Antisera produced in chickens; other produced in rabbits. Five anti-Pinnipedia sera were tested. They all gave distinct suborder reactions. One of the two anti-sea lion sera was more specific than the other and its reactions indicated that the Ursidae were more closely related to the sea lion than was the raccoon, fox or skunk. The anti- walrus sera was aspecific; it gave fairly dis- tinct subordinal reactions but the degree of reac- tions with representatives of the Fissipedia showed a similarity of the relationships of all four families of this suborder to the Pinnipedia. The reactions of the two harbor seal antisera were interesting. The antiserum (C-165) pro- duced in the chicken was highly specific and its 164 Zoologica: New York Zoological Society [42: 14 reactions with the blood of the sea lion and walrus were very low. It gave no reactions with the raccoon, fox or cat and only slight reactions with the polar bear and black bear. On the other hand, the antiserum produced in the rabbit gave reactions with the kinkajou and coati-mundi but not with the raccoon, dog or cat. No explanation suggests itself for the differences shown by these two antisera. Table 3 records the results secured with the turbidimetric technic. Nine antisera were used and of these four were produced against the sea lion. These four antisera gave definite subordinal reactions and the three that were tested by the microdensitometer method indicated that the Ursidae were probably more closely related to the Pinnipedia than were the red fox or raccoon. On the other hand reactions with antiserum-CU were slightly higher for the kinkajou than with the Ursidae and antiserum 3698-3796 gave reac- tions with the skunk sera that were of similar magnitude to that of the Ursidae. The black bear antiserum gave distinct famil- ial reactions and the cross-reactions with the Pinnipedia were slightly higher than with the raccoon, red fox and skunk. The high degree of cross-reaction with the polar bear could indicate that a closer relationship exists than the present classification of these animals indicates. It would be unsafe to argue this question on the evidence of but one antiserum. The anti-raccoon serum results showed the expected close relationship between the raccoon and the kinkajou. The Ursidae, Mustehdae and Pinnipedia showed a closer relationship to the Procyonidae than did the Canidae. In this par- ticular series of tests the great difference between the polar bear result and the black bear result is strange considering the close relationship shown between these two animals in other tests. The authors cannot account for this variation but future tests may give a satisfactory answer. The anti-wolf serum results gave an excellent example of intra-family relationships. The dog showed an 88 per cent, relationship to the wolf, while the red fox showed only a 52 per cent, relationship. Of the families tested, the Ursidae and Procyonidae seemed to be more closely re- lated to the Canidae than were the mustelids. The representatives of three families of Pinni- pedia showed a remarkably similar relationship to the wolf, while the cat (Felidae) was most distantly related. The anti-skunk serum had relatively low pre- cipitating power and was highly specific. The ferret and the tayra are members of the sub- family Mustelinae while the skunk belongs to the subfamily Mephitinae. The ferret and tayra showed no closer relationship to the skunk than did the Ursidae. The Pinnipedia, Canidae and Felidae showed distant relationship while the raccoon and hyaena gave no cross-reactions. The anti-cat serum results showed strong in- tra-family relationships but extremely weak in- ter-family relationships with all but the Hyaeni- dae. The intra-family results were peculiar. The cheetah showed a 76 per cent, relationship to the house cat while the tiger showed a 50 per cent, relationship and the mountain hon showed a 47 per cent, relationship. From these limited data one gets the impression that the tiger and mountain lion might belong to a different genus than the house cat. These results warrant much more research. Discussion The findings set forth above are at least a beginning in the serological study of the rela- tionship of the Carnivora. The species used in the study were chosen primarily because of their availability. As indicated previously, authorities disagree about the relationship of the families of the Fissipedia to each other and to the Pinnipedia on the basis of comparative anatomy and paleon- tology. This has resulted in great variations and confusion in the classification and nomenclature of the Order Carnivora. On the basis of the sero- logical data presented in this study, the following tentative conclusions can be drawn: 1. The Pinnipedia are more closely related to the Ursidae, Canidae, Mustelidae and Pro- cyonidae than to the FeUdae and Hyaenidae. 2. The closest serological resemblance exists between the Ursidae and Pinnipedia. 3. The Felidae and Hyaenidae are more closely related to each other than to any other family. 4. The polar bear seems to be quite closely related to the black bear and possibly should be included in the same genus. If additional data support present findings, it might mean that the nomenclature of members of the family Felidae should be revised. Simpson in his work on the classification of mammals came to this same conclusion in his discussion of the taxonomy of the Felidae. To attempt a serological classification of the Carnivora on the basis of the results presented in this paper might be premature. However, certain indications are given by these somewhat limited data. The tendency to classify the Cani- dae, Ursidae, Procyonidae and Mustelidae to- gether into one superfamily (Canoidea) is per- fectly valid serologically. Whether the Felidae, 1957] Pauly & Wolfe: Serological Relationships among the Carnivora 165 Hyaenidae and Viverridae can be classified to- gether into another superfamily serologically will have to be determined in the future. There are very strong indications that the suborders Fissipedia and Pinnipedia do not exist serologi- cally. Future research may show that there are three or even four serological suborders of the Carnivora. The authors realize the shortcomings of this paper as well as other serological research in which animal relationships have been consid- ered. The shortcomings are due to the lack of materials to make more complete studies. It would be much more preferable to concentrate on fewer species but with larger numbers of specimens so that a number of antisera could be produced against each species and tested against several members of each studied. In this way it might be possible to make a statistical analysis of the data and one that might yield significant results rather than mere indications. Summary 1. Serological tests were carried out using the blood sera of 23 different species from nine fam- ilies of the Order Carnivora, two sera of the Order Artiodactyla, and human serum (Order Primates) . 2. The ring test, the Baier microdensitometer and the Libby photronreflectometer were all used in performing these tests. 3. Where comparable reactions were made, the results of the microdensitometer and pho- tronreflectometer tests paralleled those of the ring test. 4. A serological basis for the classification of the Carnivora was indicated. The authors are especially grateful to Dr. Joseph G. Baier of the University of Wisconsin in Milwaukee for the use of his microdensi- tometer. Bibliography Baier, J. G., Jr. 1943. A null reading photoelectric microden- sitometer for use in turbidimetry and abridged spectrophotometry. Indust. Eng. Chem., 15: 144-148. 1947. An analysis of photoelectric instruments for measurement of turbidity with refer- ence to serology. Physiol. Zool., 20: 172- 186. Baier, J. G., Jr. & H. R. Wolfe 1942. Quantitative serologic relationships within the Artiodactyla. Zoologica, 27: 17-23. Beddard, F. E. 1902. Mammalia. Macmillan & Company, New York. Boyden, a. a. 1926. The precipitin reaction in the study of animal relationships. Biol. Bull., 50: 73- 107. 1934. Precipitins and phylogeny in animals. Amer. Nat., 68: 516-536. 1942. Systematic Serology: A critical apprecia- tion. Physiol. Zool., 15: 109-145. 1943. Serology and animal systematics. Amer. Nat., 77: 234-255. Boyden, A. A. & D. Gemeroy 1950. The relative position of the Cetacea among the orders of Mammalia as indicated by precipitin tests. Zoologica, 35: 145-151. Boyden, A. A. & G. K. Noble 1933. The relationship of some common Am- phibia as determined by serological study. Amer. Mus. Novitates, No. 606. Brown, F. M. & H. M. Heffron 1928. Serum diagnosis and Rhopalocera. Journ. N. Y. Ent. Soc., 36: 165-170. Eisenbrandt, L. L. 1938. On the serological relationship of some helminths. Amer. Joum. Hyg. 27: 117- 141. Gemeroy, G. G. 1943. On the relationship of some common fishes as determined by the precipitin reaction. Zoologica, 28 : 109-123. Goodman, M., H. R. Wolfe & S. Norton 1951. Precipitin production in chickens. VI. The effect of varying concentrations of NaCl on precipitate formation. Joum. Immun., 66: 225-236. Leone, C. A. & C. W. Pryon 1952. Serological correspondence among three species of Penaeid Crustacea. Journ. Elisha Mitchell Scientific Society, 68: 27-31. Libby, R. L. 1938. The photronreflectometer— an instmment for the measurement of turbid systems. Joum. Immun., 34: 71-73. Martin, S. & F. Cotner 1934. Serological studies of moth proteins with special reference to their phylogenetic significance. Ann. Ent. Soc. Amer., 27 : 372-379. Nuttall, G. H. F. 1904. Blood Immunity and Blood Relationship. Cambridge University Press. Reichert, E. T. & A. P. Brown 1909. The crystallography of hemoglobins. Publ. 116, Carnegie Instit. Wash., D. C. Romer, a. S. 1933. Vertebrate Paleontology. Univ. of Chi- cago Press. 166 Zoologica: New York Zoological Society [42: 14: 1957] Scott, W. B. 1913. History of Land Mammals in the Western Hemisphere. The Macmillan Company, New York. Simpson, G. G. 1945. The principles of classification and a classi- fication of mammals. Bull. Amer. Mus. Nat. Hist., 85: 1-350. WiLHELMI, R. W. 1940. Serological reactions and species specificity of some helminths. Biol. Bull., 79: 64. Wince, H. 1923-24 Interrelationships of the Mammalia Genera. Vol. 2. Pattedyr-Staeger, Copen- hagen, H. Hagerups Forlag. Wolfe, H. R. 1936. The specificity of precipitins for serum. Joum. Immun., 31: 103-116. 1942. Precipitin production in chickens. I. Inter- facial titers as affected by quantity of antigen injected and aging of antisera. Journ. Immun., 44: 135-145. [1957] Zoologica: Index to Volume 42 167 INDEX Names in bold lace indicate new genera, species or subspecies,- num- bers in bold face indicate illustra- tions; numbers in parentheses are the series numbers ol papers con- taining the plates listed immediately following. A Achlya ambisexualis, 132 bisexualis, 132 glomerate, 132 Klebsiana, 132 sp., 132 Sparrowii, 132 Acinonyx jubatus, 160 Aethria carnicauda, 152, (13) PI. I Agyrta dux, 153, (13) PI. I Anlichloris trinitatis, 109 Aphanomyces laevis, 132 Astridia, 108 Autochloris trinitatis, 109 B Bathygobius soporator, 99, (9) Pis. I, II Bison bison, 160 Bos laurus, 160 Bufo boreas boreas, 63, 65 boreas halophilus, 63, 65 C Calonotos, 127 chalcipleurus, 128, 129 craneae, 128, (10) Pis. I, II, 152, (13) PI. I helymus, 127 hoffmannsi, 109 tiburlus, 127, (10) PI. Ill tripunotatus, 129, (10) PI. Ill Calyplralegnia achlyoides, 132 Canis familiaris, 160 lupus, 160 Chrostosoma, -116 viridipunctatum, 116, (10) PI. Ill Collybus drachma, 52, (4) PI. Ill Correbidia assimilis, 154, (13) Pis. I, II Cosmosoma, 117 achemon, 118, (10) PI. Ill anoxanthia, 118, (10) PI. Ill gemmata, 112 klagesi, 118, (10) PI. Ill melalhoracia, 117, (10) PI. Ill pytna, 118 remote, 118 rubriscapulae, 117, (10) PI. Ill subfiamma subflamma, 117, (10) PI. Ill Cricosaura, 85 Ctenucha andrei, 155, (13) PI. I D Dixophlebia, 118 holophaea, 118, (10) PI. Ill Dryas julia julia, 135, (12) PI. I Dycladia, 122 basimaculata, 122, (10) PI. Ill correbioides, 122 E Episcepsis hypoleuca, 152, (13) PI. I Eucereon maia, 153, (13) PI. I obscurum, 154, (13) PI. I Euchromiinae, 106 Eumetopias jubala, 160 Eupemphix pustulosus, 63, 64 F Felis concolor, 160 domeslicus, 160 H Heliconius Isabella Isabella, 135, (12) PI. I erato hydara, 135, (12) PI. I melpomene euryades, 135, (12) PI. I ricini insulana, 135, (12) PI. I Sara rhea, 135, (12) PI. I Heliotropium indicum, 148, (13) PI. II Histiaea, 123 cephus, 123, (10) PI. Ill, 151, (13) PI. I meldolae, 123, (10) PI. Ill Homoeocera, 108 magnolimbata, 108 Homo sapiens, 160 Hyaena hyaena, 160 Hyla crepitans, 63, 64 crucifer, 63, 65 maxima, 63, 64 I Isanthrene, 108 tryhanei, 108 Isoachlya monilifera, 132, (11) PI. I K Klauberina, 83, 86 I Lepfodactylus typhonius, 63, 64 Leucotmemis, 116 albigutta, 109 hoffmannsi, 109 nexa, 116 thoracica, 109 Loxophlebia, 113 bisigna, 113, (10) PI. Ill diaphana, 113 klagesi, 114 postflavia, 114, (10) PI. Ill M Macrocneme, 123 alesa, 126 eacus, 126 nigritarsus, aberration Irinitatensis, 126 plumbea, 123, (10) PI. Ill spinivalva, 125, (10) Pis. I, II thyra intacla, 124, (10) PI. Ill, 151, (13) Pis. I, II thyra, 124 vittala, 126, (10) PI. Ill Mapeta xanthomelas, (13) PI. II Mephitis mephitis, 160 Mesothen, 114 aurantegula, 114 desperate, 114 endoleuca, 114 (10) PI. Ill pyrrha, 114, (10) PI. Ill Methysia, 110 Mustela furo, 160 vison, 160 Mydropastea chrysonota, 109 N Napala walkeri, 152, (13) Pis. I, II Nasua narica, 160 Nyridela, 116 chalciope, 116 O Odobenus rosmarus, 160 P Panthera pardus, 160 tigris, 160 Paramya chrysonota, 109 ? hoffmannsi, 109 Phaeo, 110 Pheia, 115 beebei, 115, (10) Pis. I, II Phoca vilulina richardii, 160 Phoenicoprocla, 109 astrifera, 112 atrapennis, 112 aurantipatagiata, 112 auriflua, 112 biformata, 112 capistrata, 112 chrysorrhoea, 109 cubana, 112 exima, 113 flavipicta, 112 haemorrhoidalis, 112 insperata, 112 intermedia, 112 jamaicensis, 113 latimarginata, 113 lydia, 113 melachrysea, 109 mexicana, 112 168 Zoologica: Index to Volume 42 [1957] nigricoxa, 112 nigri venter, 112 nigropeltata, 109 paucipunctata, 112 punicea, 112 rubriventer, 109 sanguinea, 112 schreiteri, 113 selecta, 112 sieboldi, 113 teda, 113 thomae, 112 trinitatis, 109 vacillans, 109, (10) PI. Ill variabilis, 112 Pleurosoma, 108 trinitatis, 108, (10) PI. Ill Poliopastea, 124 Potos oaudivolvulus, 160 Procyon lotor, 160 Prostherapis trinitatis, 63, 64 Protoachlya paradoxa, 132 Psarocolius decumanus, 87, 92, 93, (8) Pis. I, II virens, (8) PI. II Pseudomya, 118 melanthus, 119, (10) PI. Ill sanguiceps, 119 Pseudosphex, 106 deceptans, 107 kenedyae, 106, (10) Pis. I, II, 148, (13) Pis. I, II laticincta, 106, 107 melanogen, 107 Psoloptera, 122 leucosticta, 122 R Rana catesbeiana, 33, 36-43, 48 clamitans, 63, 65 pipiens, 33, 36-43, 48 sylvatioa, 63, 65 Rhynohopyga, 119 flavicollis, 119, (10) PI. Ill S Saprolegnia delica, 132 ferax, 132 megasperma, 132, (11) PI. I mixta, 132 parasitica, 131, 132 Saurita, 119 afflicta, 121, (10) PI. Ill arimensis, 120, (10) Pis. I, II Cassandra, 119, (10) PI. Ill clusia, 119, (10) PI. Ill, 150, (13) PI. I concisa, 122 lacteata, 120 nox, 120 perspicua, 121 salta, 121 temenus, 121 Sphecops, 108 aurantiipes, 108, (10) PI. Ill Steinegeria rubescens, 52 Syntomeida, 122 melanthus, 122 T Taractes asper, 56 breevoorti, 56 longipinnis, 52, (4) Pis. I-III miltonis, 56 platycephalus, 56, (4) PI. Ill princeps, 56 raschi, 56, (4) PI. II saussuri, 56, (4) PI. Ill steindachneri, 56 Taxidea taxus, 160 Tayra, sp., 160 Thalarctos maritimus, 160 Thraustotheca clavata, 132 primoachlya, 132 Tursiops truncatus, 11, 13, (2) Pis. I-III U Uca annulipes, 78, (6) PI. I batuenta, 78 beebei, 78 cumulanta, 78 deichmanni, 78 dussumieri, 78, (6) PI. I festae, 78 gaimardi, 78 galapagensis, 78 heterochelos, 78 heteropleura, 78 inaequalis, 78 insignis, 73, 78 inversa, 78 ischnodactyla, 78 lactea, 75, 76, 78 lalimana, 78 latimanus, (6) PI. I leptodactyla, 78 limicola, 78 longidigitum, 78 manii, 79 maracoani, 79 marionis, 79, (6) PI. I minax, 79 mordax, 79 oerstedi, 79 olympioi, 79 panamensis, 79 princeps, 79 pugilator, 79 pugnax, 79 rapax, 75, 79 rathbunae, 79 rhizophorae, 72, 76, 79 rosea, 79 saltitanta, 79 signata, 72, 79 speciosa, 79 stenodactyla, 79 stylifera, 79 terpsichores, 79 letragonon, 79 Ihayeri, 79 zamboangana, 72, 76, 79 Ursus americanus, 160 V Vulpes fulva, 160 X Xantusia, 86 riversiana, 83, 84, 85 Xiphophorus maculatus, 133 Z Zalophus calif ornianus, 160 Zygaena parthenii, 112 NEW YORK ZOOLOGICAL SOCIETY GENERAL OFFICE 30 East Fortieth Street, New York 16, N. Y. PUBLICATION OFFICE The Zoological Park, New York 60, N. Y. OFFICERS PRESIDENT VICE-PRESIDENTS SECRETARY TREASURER Fairfield Osborn Alfred Ely Harold J. O’Connell David H. McAlpin Laurance S. Rockefeller SCIENTIFIC STAFF: Zoological Park and Aquarium John Tee-Van General Director ZOOLOGICAL PARK Leonard J. Goss Assistant Director and Veterinarian John L. George Associate Curator of Mammals William G. Conway . . Associate Curator of Birds Grace Davall Assistant Curator, Mammals and Birds James A. Oliver Curator of Reptiles Charles P. Gandal. . .Associate Veterinarian Lee S. Crandall General Curator Emeritus William Beebe Honorary Curator, Birds AQUARIUM Christopher W. Coates . Director James W, Atz Associate Curator Carleton Ray Assistant to the Director Ross F. Nigrelli Pathologist Myron Gordon Geneticist C. M. Breder, Jr Research Associate in Ichthyology Harry A. Charipper. . .Research Associate in Histology Homer W. Smith Research Associate in Physiology GENERAL William Bridges Editor & Curator, Publications Sam Dunton Photographer Henry M. Lester. . . Photographic Consultant DEPARTMENT OF TROPICAL RESEARCH William Beebe Director Emeritus Jocelyn Crane Assistant Director David W. Snow Resident Naturalist Henry Fleming Entomologist John Tee-Van Associate William K. Gregory Associate AFFILIATES C. R. Carpenter Co-ordinator, Animal Behavior Research Programs L. Floyd Clarke Director, Jackson Hole Research Station SCIENTIFIC ADVISORY COUNCIL A. Raymond Dochez Alfred E. Emerson W. A. Hagan Caryl P. Haskins K. S. Lashley John S. Nicholas EDITORIAL COMMITTEE Fairfield Osborn, Chairman James W. Atz William Beebe William Bridges Christopher W. Coates William G. Conway Lee S. Crandall Leonard J. Goss James A. Oliver John Tee-Van o