jL/r/-, PSYCHE A JOURNAL OF ENTOMOLOGY Vol. 73 March, 1966 No. 1 CONTENTS The Habits of Pheidole ridicula Wheeler, with Remarks on Habit Pat- terns in the Genus Pheidole (Hymenoptera : Formicidae). William S. Creighton 1 Revisionary Notes on the Genera of Eucheilini (Coleoptera, Carabidae). Hans Reichardt 8 Notes on Neotropical Tabanidae VIII. The Species Described by J. C. Fabricius. G. B. Fairchild 17 Two New American Aradidae (Hemiptera: Heteroptera) . Nicholas A. Kormilev 26 Revision of the Genera Gnotus and Fabrasia (Coleoptera: Ptinidae). John F. Lawrence and Hans Reichardt 30 The Lower Permian Insects of Kansas. Part 11. The Orders Protorthoptera and Orthoptera. F. M. Carpenter 46 CAMBRIDGE ENTOMOLOGICAL CLUB Officers for 1966-67 President J. Reiskind, Harvard University Vice-President C. C. Porter, Harvard University Secretary F. Coyle, Plarvard University Treasurer F. M. Carpenter, Plarvard University Executive Committee W. L. Brown, Jr., Cornell University K. W. Cooper, Dartmouth College EDITORIAL BOARD OF PSYCHE F. M. Carpenter (Editor), Professor of Entomology , and Alexander Agassiz Professor of Zoology , Harvard University P. J. Darlington, Jr., Alexander Agassiz Professor of Zoology, Harvard University W. L. Brown, Jr., Associate Professor of Entomology , Cornell University ; Associate in Entomology, Museum of Comparative Zoology E. 0. Wilson, Professor of Zoology, Harvard University H. W. Levi, Associate Curator of Arachnology, Museum of Com- parative Zoology H. E. Evans, Curator of Insects, Museum of Comparative Zoology PSYCHE is published quarterly by the Cambridge Entomological Club, the issues appearing in March, June, September and December. Subscription price, per year, payable in advance: $4.50 to Club members, $5.00 to all other subscribers. Single copies, $1.25, Checks and remittances should be addressed to Treasurer, Cambridge Ento- mological Club, 16 Divinity Avenue, Cambridge, Mass. Orders for back volumes, missing numbers, notices of change of address, etc., should be sent to the Editorial Office of Psyche, Biological Laboratories, Har- vard University, Cambridge, Mass. IMPORTANT NOTICE TO CONTRIBUTORS Manuscripts intended for publication should be addressed to Professor F. M. Carpenter, Biological Laboratories, Harvard University, Cambridge, Mass. Authors contributing articles over 4 printed pages in length may be required to bear a part of the extra expense, for additional pages. This expense will be that of typesetting only, which is about $10.00 per page. The actual cost of preparing cuts for all illustrations must be borne by contributors: the cost for full page plates from line drawings is ordinarily $12.00 each, and the full page half-tones, $18.00 each: smaller sizes in proportion. AUTHORS SEPARATES Reprints of articles may be secured by authors, if they are ordered at the time proofs are received for corrections. A statement of their cost will be furnished by the Editor on application. The December, 1965 Psyche (Vol. 72, no. 4) was mailed March 10, 1966. The Lexington Press. Inc., Lexington, Massachusetts PSYCHE Vol. 73 March, 1966 No. 1 THE HABITS OF PH El DOLE RIDICULA WHEELER WITH REMARKS ON HABIT PATTERNS IN THE GENUS PILEID OLE (HYMENOPTERA: FORMICIDAE) By William S. Creighton'1 Department of Biolog}7, City College, New York During 1965 the writer was able to study seven colonies of Pit. ridicula at La Feria, Texas. The observations in this paper were made on these colonies or on individuals transferred from them to observation nests. There are few North American ants as poorly known as Ph. ridicula. When W. M. Wheeler described this species in 1916 he had seen three specimens, all majors (1). One of these (the type) was taken by C. L. Scott at Brownsville, Texas. The other two, in the collection of the U. S. National Museum, came from San Diego, Texas, a town about 140 miles northwest of Brownsville. Except for these locality records no field data for ridicula were available and, as far as can be determined, no additional records have been published for this species. The nests of ridicula are surprisingly difficult to find and this seems to be the reason why the species, which is a door-yard ant in the lower Rio Grande Valley, has escaped observation for the past fifty years. To judge from the La Feria colonies, ridicula prefers to nest in areas where there is a heavy cover of weeds, often nettles, common sunflower, Johnson grass and careless weed. These weeds not only conceal the nests but also the foragers which come from them. During December 1964 I made repeated visits to an area where there were two flourishing colonies of ridicula. It is now apparent that I often stood directly above these nests but neither was discovered until the covering weeds were removed. There are other features which make the nests of ridicula hard to find. A mature colony of this ant contains at least seventy-five majors and three Trofessor Emeritus, City College, New York Manuscript received by the editor February 11, 1966 Published with a Grant-in- Aid of Research from the Society of the Sigma Xi. I 2 Psyche [March hundred minors. These figures are based on totals secured by pro- longed baiting of two of the colonies. Since neither nest was put out of action by the baiting it seems certain that the estimate is on the conservative side, yet there would be justification for the view that a much smaller population was present. There is rarely a con- spicuous accumulation of excavated soil or of chaff around the nest entrance, for both are brought to the surface gradually and in small quantities. Even when a crater is built its diameter seldom exceeds three or four centimeters. Moreover, the crater is a transient struc- ture for, since ridicula mixes the chaff with the excavated soil particles, the texture of the crater is loose and light and it is easily scattered by wind or rain. Hence much of the time the only indica- tion of a ridicula nest is the nest entrance itself. This is never more than five millimeters in diameter and, more often, its diameter is about two millimeters. In addition, the nest entrance is frequently blocked up and drifted over with windblown dust and detritus. Early in this study the writer found it necessary to mark the nest entrances in order to be sure of their exact position. There is a simple explanation for most of the above features. While ridicula will sometimes bring in other seeds, it is mainly interested in those of the careless weed, Amaranthus palmeri. These seeds are matured throughout the year, hence there is an ample supply of them at all times and large numbers are not garnered seasonally. More- over, a great many of the palmeri seeds are free of any covering when they are brought to the nest. As a result there is no occasion for the production of a large chaff pile or an extensive crater since, in the genus Pheidole, both these features usually result from a seasonal excess of grass seeds which must be stripped and stored2. The soil in which ridicula nests is the Victoria loam, a fine-tex- tured, compact soil which is virtually stone free. As the walls of excavations made in it are slow to crumble, there was every reason to expect that a ridicula nest could be fully exposed. Actually this proved to be impossible. It was easy to trace the main nest passage, which consists of an unbranched shaft of remarkably uniform diam- eter (about 2 mm.) that descends vertically through the soil to a depth of about thirty-two inches. It was not difficult to demonstrate the existence of lateral passages leading away from the main shaft, for the workers would open up the transected ends of these passages 2Similar considerations apply to Ph. cerebrosior Wh. which mainly garners the seeds of desert portulacas ( P . oleracea Linne and P. retusa Engl.). This crop is seasonal but, since the seeds are bare when brought to the nest, no chaff pile results. 1966] Creighton — Pheidole 3 in the walls of the excavation. But to follow these passages to their ends was quite another matter. Although some of them were traced to a distance of two feet from the main shaft, no brood chamber or seed chamber was found. Nevertheless, it is possible to state that the nest of ridicula is diffuse with the seed and brood chambers widely separated. The foraging responses of ridicula are flexible and this makes them difficult to describe. Much of the foraging occurs at night but it is misleading to characterize ridicula as a nocturnal forager. If the nest area is shaded, or if the day is overcast, foraging may occur over a twenty-four hour period. Even when the nest is not shaded foraging ordinarily continues until mid-morning. Foraging may be diffuse or the foragers may form columns. Most of the seeds brought in are picked up from surface litter and in this type of foraging columns rarely form. But when a concentrated food source is at hand, the foragers will converge on it and a column may result. The foragers seldom get far from the nest. In each of the seven colonies most of the foraging was done within ten feet of the nest entrance. The majors take no part in the foraging although they will leave the nest with the minors. When they do so they restrict their patrol to the area around the nest entrance and it is exceptional to find them more than a foot from it. During vigorous foraging there may be several majors outside the nest and their activities effectively clear the area of other ants. There is little about the appearance of the major of ridicula to suggest its lethal behavior. From a structural standpoint Wheeler’s choice of name is apt enough, for it would be hard to imagine a more top-heavy ant. But there is nothing ridiculous about the major in action. When it is on guard, either within the nest entrance or out- side it, it attacks other ants with such efficiency that the victim usually has no chance to defend itself. The basic features of this attack are the same as those described for the major of militicida in 1959 (2). As the ridicula major approaches its victim the man- dibles are opened to their widest extent. This is followed by a rapid lunge towards the victim during which the mandibles are snapped shut. Unlike the militicida major the major of ridicula does not hold the antennae close to the head during attack. They are usually held with the scapes at right angles to the long axis of the head and with the funiculi flanking the open mandibles (See Fig. 1). Because of the lateral expansion of the genae the mandibles can be opened to an astonishing extent, with the result that the major of ridicula seldom has difficulty getting the head or the thorax of its victim 4 Psyche [March Fig. 1. Major of Pheidole ridicula in the position it assumes when about to attack another ant. between the mandibles. Perhaps this is why the ridicula major is much less deliberate in its attack and will strike the victim anywhere. Moreover, when the major of ridicula is aroused it will sometimes charge its own minors by mistake. It seldom kills them for the minor usually dodges under the closing mandibles to a safe position below the major’s head. Nevertheless minors of ridicula are sometimes crushed by their own majors. In the encounters which occurred daily around each of the nests, the majors of ridicula rarely got the worst of it. They disposed of the majors of Ph. floridana , metallescens and dentata with ease and had little trouble with those of Solenopsis geminata. They would occasionally kill workers of Pogonomyr?nex harhatus although this species was more often driven away than killed. There can, however, be no doubt about the high efficiency of the major of ridicula as a guard. In addition to its activities as a guard the major of ridicula also functions as a seed crusher. Each of the flowers of Amaranthus palmeri produces a single, shiny, black seed about one millimeter in diameter, which resembles a tiny, blunt-edged discus. When these seeds are ripe they may fall out through a transverse slit which de- velops in the ovary wall or the ovary may be shed with the seed inside it. The minors strip the ovary wall away from such seeds after they are brought to the nest and the small amount of chaff which accumulates around the nest entrance is the result of this stripping. As far as could be determined the minors of ridicula can- not open the palmeri seeds; at least they never did so in the observa- tion nests. When the major opens one of these seeds it picks it up by closing the mandibles on the lateral surfaces of the seed. Increas- ing pressure on these surfaces ultimately shatters the seed. In the observation nests the majors opened a number of seeds in quick 1966] Creighton — Pheidole succession. Their contents were then gradually eaten away by the minors. The majors take little interest in the seeds after they have opened them. The behavior of ridicula runs counter to the “classic” view of the habits of Pheidole in several important respects. This view states that most species of Pheidole gather large quantities of grass seeds during a harvest period in late summer or early fall. These seeds are carried to the nest, stripped, and stored in seed chambers. The discarded hulls are built into a chaff pile. As a result of this the colony is provided with an abundant store of seeds which carries it over the time when no seeds are available. The account is usually rounded off with the statement that the stored seeds are opened by the major, whose large head and powerful jaws adapt it for seed crushing. There is nothing illogical in the above view. The only trouble is that, as the habits of the genus Pheidole become better known, it seems to fit fewer and fewer species. Let us look for a moment at the matter of the use of stored seeds during periods when none are available. It is possible that a few species of Pheidole whose ranges extend into the northeastern United States (Ph. hicarinata, davisi, morrisi and pilifera) may behave in this fashion, for climatic conditions there often prohibit foraging over a period of five or six months. But this is assuredly not true of the bulk of our species, most of which forage all year long or at least for the greater part of the year. In addition, it can often be shown that there is no harvest period in the sense that the seeds are garnered when they have matured. Many species of Pheidole collect their seeds from surface litter and this litter furnishes a supply of seeds that may be worked for months after the seeds have ripened. The “harvest” may thus proceed throughout the entire winter and into the spring. Ph. macclendoni, militicida and ridicula all behave in this way. It seems plain enough that these species are not storing seeds against a period when seeds are not available, for there is either no such period or, if one exists, it is too brief to be of any significance. There is the even more disturbing fact that many species of Pheidole do not confine themselves to a diet of seeds. No other North American species of Pheidole gathers greater quantities of seeds than does Ph. (M.) rhea. A large nest of this species may have several bushels of chaff around the nest entrances. But, when the foraging columns of rhea are observed it may be seen that the foragers often bring in seeds and insect remains in equal numbers. Allowing for 6 Psyche [March the far smaller size of its colonies, the same "behavior is true of Ph . creightom. The matter becomes even more complex when it is neces- sary to deal with species which bring in insect remains during most of the year and gather seeds only at intervals. Such species are ex- ceptionally difficult to handle for, unless they make a conspicuous chaff pile, which they often fail to do, the only way toi prove that they have gathered seeds is to expose the seed chambers in the nest. This behavior is found in Ph. bicarinata, cerebrosior , sitarchesf rugulosa and xeroplula. It is only by stretching a point that these five species can be considered as harvesters, since their main reliance is on insect food. This leads directly to the problem of the strictly carnivorous species of Pheidole. In the writer’s opinion there are considerably more of these than has been supposed. As early as 1908 W. M. Wheeler had recognized that Ph. dentata and hyatti are carnivorous and predacious (3). In 1955 Creighton and Gregg showed that Ph. titanis is termitophagus (4). In 1964 the writer pointed out that Ph. (C.) clydei is an entomophagus scavenger (5). But there are other species which can be added to this list. It should certainly include Ph. grallipes and vallicola , both of which are in- sectivorous and predatory. It also appears that Ph. floridana and metallescens belong here. In 1958 Van Pelt showed that both species are attracted to a variety of baits (6). But when they are not baited or allowed access to kitchen scraps, each brings insect remains to the nest. They have not been reported as seed collectors and the writer has been unable to find stored seeds in the nests. The above discussion should show why it is misleading to char- acterize Pheidole as a genus of harvesters. There is obviously no possibility of applying such a designation to the growing number of carnivorous species, nor is the situation much better in the equally large number of species which utilize insect food at least as often as they do seeds. For the truth of the matter appears to be that species which subsist mainly on seeds are in the minority in the genus Phei- dole. One further detail is pertinent in this connection. It now seems probable that the major of Pheidole functions more often as a guard than it does as a seed crusher. The writer has been able to observe the guarding function in the majors of Ph. clydei , dentata , floridana, macclendoni, metallescens , militicida and ridicula. Only in ridicula has the major also functioned as a seed crusher. It is ob- vious that the major of a carnivorous species can have no occasion to crush seeds and the fact that the guarding function cuts across food preferences would seem to indicate that it, rather than seed crushing, is the basic response of the major of Pheidole. 1966] Creighton — Pheidole 7 Literature Cited 1. Wheeler, W. M., Proc. New Eng. Zool. Club, 6: 29-35 (1916). 2. Creighton, W. S. & Creighton, M. P., Psyche, 66: 1-12 (1959). 3. Wheeler, W. M., Bull. Amer. Mus. Nat. Hist. 24: 399-485 (1908). 4. Creighton, W. S. & Gregg, R. E. Univ. Colo. Studies, 3: 1-46 (1955). 5. Creighton, W. S., Psyche, 71: 169-173 (1964). 6. Van Pelt, A., Amer. Mid. Natural., 59: 1-57 (1958). REVISIONARY NOTES ON THE GENERA OF EUCHEILINI (COLEOPTERA, CARABIDAE) By Hans Reichardt1 Departamento de Zoologia, Secretaria da Agricultura, Sao Paulo, Brazil In the course of my studies on Neotropical Carabidae I have dis- covered several interesting taxonomic novelties about the endemic tribe Eucheilini which seem to be important enough to be reported on. Even though I have examined the types of the species of Euchei- lini which are preserved in the Museum National d’Histoire Naturelle, Paris, in July, 19642, it is as yet imposible to revise the tribe at the species level, since the species of Inna, one of the two genera of Eucheilini, are very poorly understood at present. Material in collections is very scarce. I hope that in the near future accu- mulation of enough specimens will allow a specific revision of this interesting tribe of Carabidae. The material on which this revision is based has been borrowed (and partly also studied in loco) from the Departamento de Zoologia, Sao Paulo (CDZ) , the Museu de la Universidad de La Plata, Argentina (MLP), the Museum of Comparative Zoology, Cam- bridge, Mass. (MCZ), the Museum National d’Histoire Naturelle, Paris (MNHN) and the United States National Museum, Wash- ington, D. C. (USNM). The loan of this interesting material is gratefully acknowledged. The tribe and the genera included in this revision are not formally redescribed, since it seems to me that for the time being the char- acterization presented below is enough. Tribe Eucheilini Eucheilinae Bates, 1883, Biol. Centr. Amer., Col., 1 ( 1 ) :1 68. Periglossinae Liebke, 1929 Ent. Anz., 9:247. new synonymy. Euchilini; Csiki, 1932, Col. Cat., 124:1585; Blackwelder, 1944, Bull. U.S.N. Mus., 185:70. Periglossini ; Csiki, 1932, Col. Cat., 124:1585; Blackwelder, 1944, Bull. U.S.N. Mus., 185:70. Eucheilini; Ball, 1960, Beetles of the U.S.:164. Currently at the Museum of Comparative Zoology, Harvard University Manuscript received by the editor December 1, 1965. 2A trip supported by the Evolutionary Biology Committee at Harvard University; this support is gratefully acknowledged. 8 3.966] Reichardt — Eucheilini 9 This small Neotropical tribe includes only two genera of small, Lebiini-like Carabidae. Their systematic position has been uncertain for a long time. Even though Eucheila , the type-genus, was described as early as 1829 (in Lebiini), the true relations to Helluonini were only recognized in 1883, when Bates erected the subfamily Eucheilinae to incorporate Eucheila and Inna (the latter also described in Lebiini, and at first considered as related to Eucheila by Bates), and placed it in the vicinity of the Helluonini. In 1929 Liebke described the subfamily Periglossinae for a new Central-American genus, Periglossium. From his description and illustrations of this beetle, it is evident that Periglossium is a synonym of Inna, and consequently the name Periglossinae has to be suppressed. The characters which link the Eucheilini to the Helluonini are the strangely modified mouthparts (Figs. 1-8). In spite of sim- ilarities, the Eucheilini are undoubtedly a distinct tribe, easily dis- tinguished from the Lebiini by the completely different mouthparts and from the Helluonini by their general Lebiini-habitus, as well as by the antennae which are pubescent from the 4th segment on in Eucheilini (pubescent from base in Helluonini). The aedeagus of the Eucheilini was unknown up to now. I was able to dissect one male of Eucheila strandi (Liebke) and one of Inna boyeri (Solier) (see Figs. 11 and 12). The two aedeagi are very similar, and this fact strengthens the supposed relation between the two genera. The left paramere of the aedeagus is reduced, but still present, being somewhat lobate in the two species. The genitalia of Neotropical Helluonini are also unknown, so that no comparison can be made now. Geographic distribution: The tribe is typically Neotropical, ex- tending from Argentina to the southern United States (Texas). No species has yet been reported from the Antilles. Key to genera 1. Labrum convex, covering apex of mandibles, with short setae only on lateral margins (Fig. 8) ; lateral margins of pronotum smooth, not crenulated; pronotum with basal setae only; tarsal claws pectinate Eucheila Dejean Labrum flat, not covering apex of mandibles, with 4 long setae on anterior margin, 2 longer ones and a series of short ones laterally (Fig. 3); lateral margins of pronotum crenulated; pronotum with basal and latero-median setae; tarsal claws simple Inna Putzeys IO Psyche [March Eucheila Dejean Eucheyla Dejean, in Dejean and Boisduval, 1829, Icon. Col. Eur., 1 :60, 176- 177 (type-species, by monotypy, Eucheyla flavilabris Dejean). Eucheila; Dejean, 1831, Spec. gen. Col., 5:455-456; Chaudoir, 1848, Bull. Soc. Nat. Moscow, 21 (1) :124; Lacordaire, 1854, Gen. Col., 1:148. Euchila Dejean (nec Euchila Billberg) ; Agassiz, 1846, Nomencl. Zool. (em- mendation) ; Gemminger and Harold, 1868, Cat. Col.. 1:155; Csiki, 1932, Col. Cat., 124:1585; Blackwelder, 1944, Bull. U.S.N. Mus., 185:70. The genus was originally spelled Eucheyla by Dejean, who in the original description gave the Greek derivation of the name. It is obvious that this spelling was an incorrect transliteration. Dejean himself must have realized this, and in 1831 used the name Eucheila instead, without any mention of Eucheyla. According to article 32 of the International Code of Zoological Nomenclature, Eucheyla should be accepted as the “correct original spelling”, since incorrect transliteration is not to be considered an inadvertent error (article 32, section a, ii). However, Eucheyla has remained unused as a senior synonym since 1829, and must, therefore, be considered a nomen oblitum (article 23, section b). Eucheila Dejean must, there- fore, be considered the correct and valid name for the genus. Eucheyla Berlese, 1913, proposed as a subgenus of Cheyletia Haller, 1884 (Arachnida, Acari, Cheyletidae) is a junior homonym of Eucheyla Dejean, and has been replaced by Neoeucheyla Radford, 1950. Agassiz (1846) emmended the name to Euchila , which is, however, a junior homonym of Euchila Billberg, 1820 (Insecta, Lepidoptera) . Eucheila Dejean is easily distinguished from Inna Putzeys by the characters given in the generic key. The genus was described for a single species, flavilabris Dejean; however, material of Inna strandi Liebke from the type-locality, proves that Liebke’s species is congenic with flavilabris. Key to species of Eucheila 1. Metallic-brown species with dark brown appendages and lighter labrum; elytra 9-carinate strandi (Liebke) Metallic-green species (sometimes very dark), with yellow ap- pendages and labrum; elytra with vestigial carinae flavilabris Dejean Eucheila strandi (Liebke), new combination. (Figs. 5-8, 11) Inna strandi Liebke, 1939, Festschr. Emb. Strand, 5:121 (type from Jatai, Brazil, in Liebke’s collection; probably destroyed). 1966] Reichnrdt — Eucheilini ] i Figs. 1 — 4, Inna boyeri (Solier) : Fig. 1, maxilla; Fig. 2, labium; Fig. 3, labrum; Fig. 4, mandibles; Figs. 5 — 8, Eucheila strandi (Liebke) : Fig. 5, mandibles; Fig. 6, maxilla; Fig. 7, labium; Fig. 8. labrum. Psyche, 1966 Vol. 73, Plate 1 Reichardt — Eucheilini 1966] Reich ardt — Eucheilini 13 The pectinate claws, the very typical labrum and labium and the non-crenulated margin of the pronotum put this species without doubt in the same genus as flavilabris. It is easily distinguished by the dif- ferent color and the well developed elytral carinae. Examined specimens (6): Brazil: Sao Paulo , Guatapara (i ex., CDZ) ; Golds ; Jatai (3 exx., CDZ, MCZ) ; Bahiaj no locality (1 ex., MNHN) ; Ceard, no locality (1 ex., USNM). Euchila flavilabris Dejean (Fig. 9) Eucheyla flavilabris Dejean, in Dejean and Boisduval, 1829. Icon. Col. Eur., 1:178, pi. 8, fig. 3 (type from “environs de Rio-Janeiro”, MNHN; examined). Eucheila flavilabris; Dejean, 1831, Spec. gen. Col., 5:456-457; Lacordaire, 1854, Gen. Col., 1, pi. 4, fig. 4; Putzeys, 1863, Mem. Soc. Sci. Liege, 18:72. pi. 2, figs. 75-77. Eucheila flavilabris is easily distinguished from strandi by the com- pletely different color, especially that of the appendages. A few specimens are very dark, almost as brown as strandi; however, the elytral carinae are always vestigial and the appendages always yellow. Examined specimens (14): Brazil: Bahia , Salobro (1 ex., MNHN); Minas Gerais Matusinhos (1 ex. MNHN); Serra do Caraca (1 ex., MNHN); Rio de Janeiro , Nova Friburgo (6 exx., MNHN); Guanabara , Rio de Janeiro (1 ex., MNHN) ; Sao Paulo , Estacao Biologica de Boraceia, Salesopolis (2 exx., CDZ) ; Santa Catarina , no locality (1 ex., MNHN). Argen- tina: Santiago del EsteroJ near Icano (1 ex., MNHN). Inna Putzeys Injia Putzeys, 1863, Mem. Soc. Sci. Liege, 18:71 (type-species, by monotypy, Inna punctata Putzeys) ; Chaudoir, 1872, Rev. Mag. Zool., (2) 23 :2 19- 221 (redescription). Periglossium Liebke, 1929, Ent. Anz., 9:246-247 (type-species, by original designation, Periglossium nevermanni Liebke). NEW synonymy. Fen species of Inna are presently known, their distribution ranging from Argentina ( atrata Dejean) to southern Texas ( texana Schaef- fer). My notes on the types suggest that some of the described species are synonyms, e.g., Inna costulata Chaudoir is differentiated from granulata Chaudoir only by color: costulata is coppery-metallic, Explanation of Plate 1 Fig. 9, Eucheila flavilabris Dejean, head and pronotum; Fig. 10, Inna me gala, n. sp., head and pronotum; Fig. 11, aedeagus of Eucheila strand * (Liebke) ; Fig. 12, aedeagus of Inna boyeri (Solier). H Psyche [March while granulata is very dark brown, almost non-metallic. This color difference seems to be a very weak character, but since only the types are known, further decisions cannot be made now. It is also possible that nevermanni , from Costa Rica, is a synonym of costulata , which ranges from Colombia to Guatemala. The type of Inna nevermanni (Liebke) has probably been destroyed with part of Liebke’s collection and no material is presently available from Costa Rica, so that nothing else can be said here. The generic description of P eriglossium Liebke does not differ* in any respect from that of Inna Putzeys. There seems to be no reason to maintain P eriglossium , which was probably described by Liebke when he did not know Inna. Inna is easily distinguished from Eucheila by the characters given in the key. Even though I have seen the types of most species, as mentioned above, I am presenting below new data only on two of the older species, of which material was available and could be identified. The recognition of the new species is based on comparison with the orig- inal descriptions of all the older species as well as on my notes on their types. Ina boyeri (Solier) (Figs. 1-4, 12) Polystichus boyeri Solier, 1835, Ann. Soc. Ent. France, 4:111 (holotype male from “Colombia”, MNHN ; examined). Inna boyeri; Chaudoir, 1872, Rev. Mag. Zool., (2) 23:241-242 (redescrip- tion). I am referring to this species, originally described from Colombia, a series of io specimens from Barueri in the state of Sao Paulo, Brazil (CDZ, MCZ), which agree with the description and my notes on the type. Inna boyeri is very similar to costulata; however, it has more densely punctate pronotum and head, and is slightly larger in size. Inna atrata (Dejean) Cymindis atrata Dejean, 1831. Spec. gen. Col., 5 :327 (holotype from “Buenos- Ayres”, MNHN; examined). Inna atrata; Chaudoir, 1872, Rev. Mag. Zool., (2) 23:243-244. The type-specimen in the Paris Museum is very damaged : the left elytron and the left antenna are missing, as well as parts of several legs. The species is very characteristic, having a densely punc- tate head and being the smallest species of the genus. 1966] Reichardt — Eucheilini 15 Examined specimens (3) : Argentina: Buenos Aires , Buenos Aires (1 ex., MLP); Isla Martin Garcia (1 ex., MLP) ; Misiones f no locality (1 ex., MLP). Inna megala, n. sp. (Fig. 10) Description: Reddish-brown, with light, almost yellow appendages; elytral margin very light brown. Head — densely punctate dorsally, with longitudinal striation on antennal tubercules; whole surface micro-reticulate. Pronotum — wider than long, slightly wider than head; densely punctate on surface, with somewhat granulate aspect; posterior angles more or less square; lateral margins turned upwards, crenulated; median line in a slight depression which continues on each side anteriorly (forming a Y). Elytra — with 8 irregularly punctured sulci; 7 discal interstices more or less smooth, convex; 8th and 9th interstices very slightly indicated only, mainly posteriorly; almost twice as wide as pronotum, less than twice as long as wide; elytral margin with setose punctures. Measurements — holotype, 3.7 X 10.3 mm; paratype, 4.1 X 10.9 mm. Types: Paraguay: holotype female, Villarrica, F. Schade col. (MCZ n. 3 1 197); paratype female, Amambay, A. Schulze col. (CDZ). Inna megala is very similar to planipennis Bates, which is only known from Mexico. The two species are of about the same size; planipennis has a less densely punctured head, especially between the eyes; the pronotum of megala is more transverse than that of plani- pennis. Specific name: megala is derived from the Greek adjective megas , meaning large. References Ball, G. E. 1960. Carabidae, in Arnett, The Beetles of the United States, pp. 55- 182, 63 figs., Washington, D. C. Bates, H. W. 1881-1884. Carabidae in Biologia Centrali-Americana, Coleoptera, 1, 299 pp., 13 color plates. Blackwelder, R. E. 1944. Checklist of the Coleopterous Insects of Mexico, Central Amer- TSa, the West Indies and South America. Part 1. Bull. U.S.N. Mus., 185:1-88. Chauboir, M. Memoire sur la famille des carabiques. Bulb Soc. Nat. Moscow, 21:3-134. i6 Psyche [March 1872. Descriptions d’especes nouvelles de carabiques de la tribu des troncatipennes, et remarques synonymiques. Rev. Mag. Zool., (2) 23:219-221. Csiki, E. 1932. Carabidae, Harpalinae VII, in Coleopterorum Catalogus, pars 124:1279-1598. Dejean. P. F. M. A. 1829. In Dejean and Boisduval, Iconographie et histoire naturelle des Coleopteres d’Europe, 1, 400 pp., 60 pis., Paris. 1831. Spec. gen. Col., 5, 883 pp., Paris. Gemminger, M. and E. von Harold 1868. Catalogus Coleopterorum . . . , 1, 424 pp., Monachii. Lacordaire, J. T. 1854. Genera des Coleopteres . . . , 1. 486 pp., Paris. Liebke, M. 1929. Laufkaeferstudien. VI. Ent. Anz., 9:245-247, 261-265, figs. 1939. Neue Laufkaefer. Festschr. Embr. Strand, 5:91-130, 21 figs. Putzeys, J. A. A. H. 1863. Postscriptum ad clivinidarum monographiam atque de quibusdam alliis. Mem. Soc. Sci. Liege, 18:1-78, figs. SOLIER, A. J. J. 1835. Description de quelques especes nouvelles de la famille des cara- biques. Ann. Soc. Ent. France, 4:111-121. NOTES ON NEOTROPICAL TABANIDAE VIII. THE SPECIES DESCRIBED BY J. C. FABRICIUS1 By G. B. Fairchild Gorgas Memorial Laboratory, Panama, R. de. P. During the summer of 1964 I had the opportunity of studying the types of Neotropical Tabanidae in Copenhagen with the financial aid of a travel grant from the Bache Fund of the National Academy of Sciences,. The existing type specimens of all insects described by Johann Christian Fabricius have recently been reviewed by Dr. Ella Zim- sen (1964), of the Universitetets Zoologiske Museum in Copen- hagen. The present notes concern only the neotropical Tabanidae. Studies of the Wiedemann types in Copenhagen will appear elsewhere. Fabricius type material was for a long time present in the Museums at Kiel and Copenhagen, but the Kiel collection has recently been deposited in Copenhagen. The Kiel material was much damaged by pests, while the Copenhagen material is generally quite well pre- served. I am greatly indebted to Dr. S. L. Tuxen for permission to study this material, and for comments on the results, and to Dr. Leif Lyneborg and Dr. Ella Zimsen, who helped greatly with advice and hospitality during my visit in Copenhagen. Wiedemann (1828) discussed and redescribed most of Fabricius’ species, and later authors have largely depended on his interpretations, as he studied the Fabrician specimens. More recently Philip (i954? i960) has reported on a number of the Fabrician types, selected lectotypes, and corrected certain misidentifications. In the following list all the supposedly Neotropical Tabanidae described by Fabricius are cited, together with brief notes on their present condition, tax- onomic status and whether deriving from the Kiel (K.) collection or the Copenhagen (C.) collection. Types of all but two species were seen. The species are listed alphabetically by modern genera, or subgenera, the genus in which they were placed by Fabricius added in parentheses where this differs. As to the localities from which Fabricius’ material came, I do not have information more definite than given in his descriptions. It is noteworthy, however, that all of his species from “America meridionali”, except Chrysops moerens and possibly Fidena analis, have since been taken in Surinam or the Guianas. The localities of 1 Manuscript received by the editor February 12, 1966 17 i8 Psyche [March the remainder, Brasilia, Cajennae, and Americae insulis, have also been confirmed by subsequent collecting. List of Species A canthocera longicornis (Fab.) (Tabanus), 1 77 5 > Syst. Ent., p. 790, Brasilia. 1794, Ent. Syst., IV p. 371. Brasilia Mus. Dom. Banks. The species should date from 1775, not x794> as given by Wiedemann and most subsequent authors. The descriptions are iden- tical. I did not see the type, which should be in B.M., but is ap- parently lost. The species is the type of A canthocera Macq. 1834. Catachlorops rufescens (Fab.) (Tabanus), 1805, Syst. Antliat., p. 100, Amer, Merid. Dom. Smidt Mus. de Sehestedt. The single type (C.) is intact though faded and dusty. It is labelled “T. rufescens ex Am: Mer: Schmidt”. Scutellum white-haired and white transverse bands on at least second to fourth abdominal tergites. Legs entirely dark. Beard, pleura and venter of abdomen entirely dark- haired. Labella large and wholly sclerotized, palpi slender, brown, black-haired. I give here figures of wing and head structures (Fig. 1). Krober’s (1939) C. rufescens is entirely different, while the type of his C. scuteliatus in British Museum agreed closely with notes and figures of type of rufescens Fab. (New synonymy) . Chlorotabanus inanis (Fab.) (Tabanus), 1787, Mantissa Insect. II, p. 356. Cajennae Dom. v. Rohr. The name should date as above not 1794 as given by subsequent authors. Philip and Fairchild (1956) have discussed the material in Copenhagen, which consists of but the pin and labels of the original specimen (K.) and another specimen (C.) in fair condition det. by Fabricius from Amer. Merid. Chrysops costatus (Fab.) (Tabanus) 1794, Ent. Syst. IV, p. 373, without locality; 1805, Syst. Antliat. p. 112, in America meridionali Dom. Smidt. Mus. Dom Lund. There are three specimens (C.) all in good condition, one of which is labelled type. This agrees with current interpretations. One of the others is C. ecuadoriensis Lutz or a related form. The name has been shown to be a synonym of C. variegatus (De G.). Chrysops laeta Fab. 1805, Syst. Anti., p. 112, in America merid- ionali Dom. Smidt. Mus. Dom. Lund. Philip (1955) has dis- cussed the types and shown that laeta auct. is not Fabricius’ species. There are 3 specimens, two (C.) in good condition, the other (K.) represented by wings only. The specimen described as tuxeni by Philip is formosa Krob., (New synonymy ) while the Kiel specimen is probably C. varians var. tardus Wied. True laeta differs from most allied species by the much shortened discal cell. Specimens in 1966] Fairchild — Tabaniclae 19 Vienna labelled types are those studied by Wiedemann, and not, of course, true types. They are varians var. tardus. C. tenuistria Krob., whose type is lost, is very probably a synonym of laeta Fab. (new synonymy). Chrysops tristis (Fab.) (Tabanus) 1798, Ent. Syst. Suppl., p. 567* Cajennae Dom. v. Rohr. Of the type (K) only the name label remains. Two others specimens (C.) are in the collection, one of which bears a Lectotype label by Philip. Although not strictly types, these specimens were surely studied by Fabricius and Philip’s action seems justified in the interests of stability, especially since subsequent workers have applied the name to at least 3 other species. I have an agreeing specimen from Surinam. Chrysops moerens (Fab.) (Tabanus) 1787, Mantiss. Insect. II. p. 356 Cajennae Dom. v. Rohr. The type now consists of a single wing (K.). It is not a Neotropical species and is discussed elsewhere (Fairchild, 1966 in press). Diachlorus bicinctus (Fab.) (Tabanus) 1805, Syst. Antliat. p. 102, America meridionali Dom. Smidt. Mus. Dom. Lund. The single type (C.) lacks one wing. It agrees with current interpreta- tions, and is type of the genus. Diachlorus curvipes (Fab.) ( Haematopota) 1805, Syst. Antliat., p. 107. America meridionali Dom. Schmidt. Mus. Dom de Sehestedt. There are 2 females labelled type (C.) in good condition, and another (K.) not labelled type lacking head and abdomen. The two types agree with current interpretations, the other specimen is too damaged for certainty, but is probably the same. Diachlorus podagric, us (Fab.) (Haematopota) 1805, Syst. Antliat. p. 108. America meridionali. Dom. Smidt. Mus. Dom. de Sehestedt. Two females (C.) labelled type in good condition. The type of D. nigrithorax Krob. 1930 in British Museum agrees with my notes on podagricus, and I believe is a synonym, (new synonymy). Neither Krober (1928) nor Lutz (1913) recognized the species. Dichelacera cervicornis (Fab) (Tabanus) 1805, Syst. Antliat. p. 100. America meridionali. Dom. Smidt. Mus. Dom. Lund. Two females in good condition (C.) are labelled types, another (K.) lacks head and abdomen, but is probably the same. All agree with current concepts (Fairchild and Philip i960). Dichelacera damicornis (Fab.) (Tabanus) 1805, Syst, Antiliat. p. 1 01. America meridionali. Dom. Smidt. Mus. Dom. Lund. Two females in fair condition (C.) are labelled types. Another (K.) con- sists of but 2 legs and the wings. The types are as treated by Fair- child and Philip (i960) ; the Kiel specimen is probably the same. [March Fig. 1. Catachlorops rufescens (Fab.), palpi and proboscis, frons, an- tenna and wing. Proboscis and frons are to same scale, antenna at twice this magnification and wing about half magnification of frons. Holotype. Dichelacera T. nigrum (Fab.) (Tabanus) 1805, Syst. Antliat. p. 101. America meridionali. Dom. Smidt. Mus. Dom. Lund. Two females in fair condition (C.) labelled type. Another (K) consists only of fragments of thorax and 1 wing. At least the types agree with current interpretations. Fidena analis (Fab.) (Pangonia) 1805, Syst. Antliat. p. 91. America meridionali. Dom. Smidt. Mus. Dom. Lund. The type 1966] Fairchild — T ahanulae 21 Figs. 2 and 3. Fig. 2. Fidena analis (Fab.), head in side view. Holotype. Fig. 3. Tabanus trivittatus Fab., palpus, frons. antennae, all to same scale. Lectotype. is a male, labelled “P. analis ex Am: Mer: Schmidt.” It lacks most of third antennal segments, but is otherwise in good condition (C.). This is not the species so identified by Lutz (1909) or Krober (1930). The legs are bicolored, femora black, tibiae & tarsi yellow, beard and all thoracic hairs black. Clypeus pollinose dorsally, shiny laterally. Wings entirely glass clear, veins yellowish, blackish only at extreme base. Abdomen plump and inflated, the first two seg- ments blackish, dark-haired, the remainder yellowish or yellowish brown, wholly yellow-haired. Remains of antennae yellow. First posterior cell narrowly open. It is structurally similar to F. rhino- 22 Psyche [March phora Bell, and F. basilaris Wied., differing in lacking darkened basal cells and coloring. It is possible that Fidena oldroydi Barr, from British Guiana is the female of cinalis, but direct comparison of specimens has not been made, and I had not seen specimens of the former when I studied the type of analis. Oldroydi seems to differ in having the wing veins brownish and patches of silvery hairs on sides of tergite 2. A figure of the head of the type of analis is in- cluded here (Fig. 2). Lepiselaga crassipes (Fab.) (Haematopota) 1805, Syst. Antliat. p. 108. America meridionali Dom. Smidt. Mus. Dom. Lund. A single female (C.) labelled type is in good condition except for lack of antennae. It agrees with current interpretations. Phaeotabanus cajennensis (Fab.) Tabanus) 1787, Mantissa In- sect., II, p. 355. 1794, Ent. Syst., IV, p. 366-367, Cajennae Dom. v. Rohr. The type (K.) is fragmentary, with little remaining but shell of thorax and wings, and not determinable with certainty. An- other specimen (C.) has an old determination label and a Metatype label by Philip. It is well preserved, the fore femora and basal halves of others black, agreeing with the darker specimens of the species, as currently understood, common in northern S. America. Selasoma tibiale (Fab.) (Tabanus) 1805, Syst. Antliat. p. 102. America meridionali. Dom. Smidt. Mus. Dom. de Sehestedt. The type is intact, in good condition, and agrees with current concepts. Stenotabanus stigma (Fab.) (Tabanus) 1805, Syst. Antliat. p. 104. Americae insulis. Dom. Smidt. Mus. Dom. Lund. Bequaert (1940) examined and reported on the types, which consist of a male and female (C.) in excellent condition. Another specimen (K.) is fragmentary, only body and one wing remaining. I can add nothing to Bequaert’s comments. Tabanus indicus Fab. 1805, Syst. Antliat., p. 103, America meri- dionali Dom. Smidt. Mus. Dom. Lund. Two specimens labelled type. One bears a label with “T. indicus ex Tranquebar”, the other unlabelled, both are in fair condition and from the Copenhagen collection. The specimen from Tranquebar is an oriental species (Tranquebar is a town on the east coast of India). The unlabelled type is the common neotropical species currently being treated as T. lineola var. carneus Bell., but also represented by the types of T. dorsiger Wied. and T. dorsovittatus Macq. Philip (in press) has selected the Tranquebar specimen as lectotype, so the name will apply to an oriental species. Tabanus lineola Fab. 1794, Ent. Syst. IV p. 369, America boreali Mus. Dom. Bose. As previously noted by Philip (1942), a type 1966] Fairchild — T abanidae 23 specimen in the Kiel collection is fragmentary, consisting of wings, fore legs, part of thorax and first two abdominal segments almost completely denuded. Specimen bears an old label with “lineola” on one side and “49” on the other. The wings are glass clear, the costal cell not tinted and without appendix on third vein. Fore coxae pale grey pollinose, white haired; femora black, pale haired; tibiae with basal half white, apical half and tarsi black. "I horax blackish with notopleural lobes reddish, as are sides of mesonotum. Scutellum destroyed by pests. First abdominal tergite yellow with a pair of small round submedian dark integumental spots nearly touching in mid-line. Between them there are the remains of a whitish pollinose streak. Second tergite with a broad median dark stripe which is somewhat forked at its posterior end and fails to reach hind margin. This median dark integumental stripe is over- laid by vestiges of a narrow pale pollinose stripe. Nothing in the type contradicts present assignment of the name to the Nearctic form, though absence of scutellum prevents certainty that it is not sub- similis Bell. Wholly clear wings rule out Neotropical forms except colombensis Macq., which has darker and less contrastingly marked tergites. Use of this name by Krober (1932) for southern Neotrop- ical material is unwarranted. The Bose specimen, not found in Paris, is probably lost. Tabanus pellucidus Fab. 1805, Syst. Antliat., p. 97, America meri- dionali Dom. Smidt. Mus. Dom. de Sehestedt. The type (C) bears an old label with “T. pellucidus ex Am: Mer: Schmidt”. It is on an old short pin, is extensively denuded, lacks antennae, mouth parts, palpi, and all legs; hole in subcallus and base of abdomen beneath. The beard and fore coxae white. Abdomen above reddish with traces of white sublateral patches, beneath with broad sublateral white- haired longitudinal bands. Wing veins brown-margined, first pos- terior cell closed and petiolate. It agreed closely with a homotype of T. senior Wlk. in shape of frons and all else remaining, except that wing cell closed further from margin. Another specimen (K.) not labelled type, lacks head, legs and one wing and is very dirty and denuded. In this the cell is open, but the specimen too fragmen- tary for certain placement. I believe the following names, whose types I have seen, to be synonymous as noted elsewhere (Fairchild 1966a in press). T. crassicornis Wied. 1821, T. albibarbis Wied. 1824, T'. angustifrons Macq. 1847, T. alboater Wlk. 1850, T. senior Wlk. 1850, T. atricornis Big. 1892, and probably also Chelommia amazonensis Barr, 1949, of which I have not seen the type. The species is variable as to color, brown to black, and the first posterior 24 Psyche [March cell varies from open to long petiolate. The species was unrecognized by Wiedemann, whose description (1828) differs in notable respects from that of Fabricius, suggesting that he did not see the present type. Subsequent students have done no better with the brief and characterless description. Tcibanus quadripunctatus Fab. 1805, Syst. Antliat. p. 99. Amer- ica meridionali Dom. Smidt. Mus. Dom. de Sehestedt. The type (C.) bears an old handwritten label “T. 4 punctatus ex Am: Mer: Schmidt” and is on an old short pin. Palpi, proboscis and outer half of left wing missing, but otherwise the specimen in fair condition. I have nothing to add to Philip’s (i960) comments on it. T ah anus trivittatus Fab. 1805, Syst. Antliat., p. 104, America meridionali Dom. Smidt. Mus. Dom. de Sehestedt. The type (C.) is labelled “T. 3 vittatus ex Am: Mer: Schmidt.” and bears a red lectotype label of Philip’s, who discussed the types (1954). The other specimen labelled type (C.) is det. Philip as dorsiger Wied. It lacks the shiny subcallus, but through oversight I did not compare it with the dorsiger type in Copenhagen. A final specimen (K) is fragmen- tary, lacking head and much of abdomen. What remains agrees with the lectotype. The lectotype has orange antennae, the style slightly darker. All femora black. Fore tibiae two-thirds, the others nearly entirely, white. Palpi yellowish, almost wholly black-haired. Frontal callus rugose. Wings very lightly tinted. The rather denuded abdomen shows an even middorsal stripe. I give here a drawing of the type (Fig. 3), which matches closely in proportions of frons and antennae my homotype of T. callosus Macq., though slightly larger. Specimens from Surinam are an even closer match to this figure, though I did not have them available in Copenhagen, unfortunately. References Bequaert. J. 1940. The Tabanidae of the Antilles. Rev. Ent. 11 (1-2): 253-369, figs. 1-32. Fairchild, G. B. and C. B. Philip 1960. A revision of the Neotropical genus Dichelacera subgenus Diche - lacera Macq. Studia Ent. 3 (1-4): 1-86, PI. 1-7. Fairchild, G. B. 1966. Some new synonymies in Tabanidae. Proc. Ent. Soc. Washington, in press. Fairchild, G. B. 1966a Notes on Neotropical Tabanidae VII. The species described by C. R. W. Wiedemann. J. Med. Ent. In press. Krober, O. 1925-1926. Die Chrysofs- arten Sud-und Mittelamerikas nebst den arten den Inselwelt und Mexikos. Konowia 4:210-256, 319-375, PI. 1-5. 1966] Fairchild — T abanidae 25 1928. Die Araerikanischen Arten der Tabanidensubfamilie Diachlorinae End. Beih. z, Arch. f. Schiffs-u. Tropen-Hyg., etc., 3 2 (2) :7 3-123, figs. 1-26. 1930. Die Tabanidengattung Sackenimyia Big. Zool. Anz., 90(1-2) :1-12, figs. 1-6. 1933. Das Subgenus Neotabanus der Tabanidengattung Tabanus s. lat. Rev. Ent. 3 (3) :337-367. 1939. Das Tabanidengenus Catachlorops Lutz. Verhoff. deutsche Kol. Mus. Bremen, 2(3) :211-232, pi. 13-16. Lutz, A. 1909. Tabaniden Brasiliens und einiger Nachbarstaaten. Zool. Jahrb., Jena, Suppl., 10 (4) :619-692. PI. 1-3. 1913. Tabanidas do Brasil e alguns estados visinhos. Mem. Inst. Osw. Cruz, 5(2) :142-191, PI. 12-13. Philip, C. B. 1942. Notes on Nearctic Tabaninae. Part III. The Tabanus lineola complex. Psyche, 49 (1-2) :25-40. 1954. New North American Tabanidae, VII. Descriptions of Tab- aninae from Mexico. Amer. Mus. Novitates, No. 1695, pp. 1-26, figs. 1-15. 1955. The types of Chrysops laeta Fabr. and a new species of Neotrop- ical deerfly in the Copenhagen Zoological Museum. Ent. Medd., 27:70-75. 1960. Further records of Neotropical Tabanidae, mostly from Peru. Proc. Calif, Acad. Sci., Ser. 4, 31(3) :69-102, fig. 1. 1966. New North American Tabanidae XVIII. New species and ad- denda to a Nearctic catalogue Ann. Ent. Soc. Amer.. in press. Philip, C. B. and- G. B. Fairchild 1956. American biting flies of the genera Chlorotabanus Lutz and Cryptotylus Lutz. Ann. Ent. Soc. Amer., 49 (4) :3 13-324. fig. 1 and PI. 1. Wiedemann, C. R. W. 1828. Aussereuropaische zweiflugelige Insekten, 1 :1-XXXII, 1-608, PL I-VIb. Zimsen, Ella 1964. The type material of J. C. Fabricius. Copenhagen, 1964, pp. 1-660. TWO NEW AMERICAN ARADIDAE (HEMIPTERA-HETEROPTERA) By Nicholas A. Kormilev1 By the kind offices of Dr. John F. Lawrence, Museum of Com- parative Zoology, Cambridge, Mass., I have had the opportunity to study several lots of Aradidae, collected by him on bracket fungi ( Basidiomycetes : Polyporaceae) in various parts of North America. Two of the species were found to be new, and are described below, in order to facilitate a study of fungus-feeding insects now being undertaken by Dr. Lawrence. In the measurements, 25 units — 1 mm. Aradus oviventris, new species (Figs. 1 — 4) type data : Holotype, cf , Rustler Park, 8 mi. W Portal, Cochise Co., Arizona, VIII. 8. 1961, J. F. Lawrence, coll. (Lot No. 918), ex Fomes subroseus [— Forties cajanderi Karsten] on Pinus sp. ; de- posited in the American Museum of Natural History, New York. Paratypes, 6cf cf and 6$?, same data; in the Museum of Compara- tive Zoology (MCZ No. 3 1201) and the author’s collection. description : Male ovate, with rather broad abdomen ; female more narrowed posteriorly. Head (Fig. 1) slightly longer than its width through eyes ( cf 32.5:28, $ 33:30); anterior process stout, rather short, slightly widened in middle, reaching to basal 1/5 of antennal segment II; antenniferous spines short and stout, acute, and slightly divergent; lateral tooth minute, but distinct; eyes sub- globose. very prominent; preocular tubercles small, acute, the post- ocular blunt; vertex deeply and narrowly, semicircularly depressed, with two rows of rough granules in middle. Antennae (Fig. 2) less than twice as long as head ( cf 57:32.5, $ 57-5:33), and much nar- rower than fore femora; segment II subcylindrical, slightly widened in apical third, as long as head width including both eyes ( cf ) , or one eye (?) ; HI subcylindrical, slightly widened toward apex; IV elongate-ovate; proportions of antennal segments; cf 6:28:13:10, $ 6.5:27:13:11; rostrum reaching to middle coxae, or, at most, to hind border of prosterum. Pronotum (Fig. 1) widest just behind middle, less than half as long as its maximum width ( cf 23 :53, ? ^3 65 Lincoln PI., Apt. 2F, Brooklyn, New York Manuscript received by the editor March 15, 1966. 26 1966] Kormilev — A radidae 27 21:55); anterior border subtruncate; anterolateral borders straight, with irregular teeth ; the posterolateral slightly rounded ; fore and hind lobes rather flat; interlobal depression shallow; carinae mod- erately prominent, granulate. Scutellum subtriangular, much longer than its width at base ( c? 30:23, $ 29:23) ; lateral bordersi barely convex, almost straight, feebly reflexed ; disc flat, finely granulate in basal 2/5, more roughly so in apical 3/5; depressed at base, and moderately raised around depression; apex narrowly rounded. Hem- elytra narrow, leaving entire connexivum and parts of tergum ex- posed, reaching genital lobes (cf), or tergum VII (?) ; corium extending to suture between terga IV and V, its basolateral border expanded, rounded, and more or less reflexed. Abdomen longer than its maximum width across segment IV ( cf 90:75, 9 io5 :$5) > lateral borders strongly convex in both sexes, very finely serrate, and slightly notched at PE angles of connexiva; genital plates (Figs. 3 — 4) slightly convex posteriorly, and slightly notched at spiracles in male, more convex and less notched in female. Color testaceous; depres- sion on vertex, apex of antennal segment II, basal 2/3 of III, entire IV, some spots on pronotum, and lateral borders of scutellum, darker, brown to piceous; posterolateral borders of pronotum, basolateral expansions on hemelytra, oval spots on terga III to VI, and round, callous spots on connexiva III to VI whitish; in some specimens posterior borders of connexiva II to VI, and R + M veins of corium, pinkish; ventral side testaceous, with some darker spots; hind borders of connexiva pink. Total length: cf 6.48 mm., 9 7-20 mm.; width of pronotum: cf 2.12 mm., 9 2.20 mm.; width of abdomen: cf 3.00 mm., 9 3-40 mm. Aradus oviventris keys out to Aradus basalis in Parshley2, but it may be separated from the latter species by the relatively shorter rostrum, reaching at most to hind border of prosternum, different shape of pronotum (with stronger teeth), relatively shorter antennae (only slightly longer than head and pronotum together), and color. Aradus lawrencei, new species (Figs. 5 — 8) TYPE DATA: Holotype, cf , 7 mi. NW Wilton, Hillsboro Co., New Hampshire, VII. 30. 1965, J. F. Lawrence coll. (Lot No. 1559), ex Fomes cajanderi Karsten on Picea sp. ; deposited in the American Museum of Natural History, New York. Paratypes, 2cfd' and Tarshley, H. M., 1921. Essay on the American species of Aradus (Hemip- tera). Trans. Amer, Ent. Soc., 47: 1-106. 7 Pis. 28 [March Psyche Figs. 1 — 4. Aradus oviventris n. sp. Fig. 1, head and pronotum; Fig. 2, antenna; Fig. 3, tip of abdomen ( $ ) ; Fig. 4, tip of abdomen ( 9 ). Figs. 5 — 8. Aradus lawrencei n. sp. Fig. 5, head and pronotum; Fig. 6, antenna; Fig. 7, tip of abdomen ($); Fig. 8, tip of abdomen ($). 399, same data; in the Museum of Comparative Zoology (MCZ No. 31202) and the author’s collection. One nymph collected with this lot. It is a pleasure to dedicate this species to its collector, Dr. John F. Lawrence. description: Male elongate-ovate; pronotum flat; finely granu- late. Head (Fig. 5) slightly longer than its width through eyes ( cf 29:27, 9 32:31) ; anterior process with parallel sides, but very slightly enlarged at base, reaching a little over 1/5 of antennal segment II; antenniferous spines strong, acute, slightly divergent; lateral tooth minute, sometimes obsolete; eyes subglobose, very prominent; pre- ocular teeth diminutive, acute, the postocular acute ( cT ) or blunt (9) ; depressions of vertex moderately deep, slightly convergent pos- teriorly. Antennae (Fig. 6) moderately strong, narrower than fore femora; segment II slightly enlarged at base, more so at apex, its length equal to or slightly greater than interocular space plus one eye; proportions of antennal segments: cf 6:24:12:11, 9 7:27:13.5: 1966] Kormilev — A radidae 29 12; rostrum reaching to middle of prosternum. Pronotum (Fig. 5) less than half as long as its maximum width ( cf 20:47, 9 25:56) ; anterior border sinuate; anterolateral borders straight, with mod- erately strong, irregular teeth; posterolateral borders barely convex, slightly convergent posteriorly; disc flat; interlobal depression some- times obsolete in middle or very shallow; carenae thin and semi- obliterated. Sc.utellum triangular, longer than its width at base ( cf 26:21, 9 32:23), shorter than ( cf ) or as long as (9) head; lateral borders low, very slightly convex; apex narrowly rounded; basal elevation obsolete. Hemelytra narrow, leaving entire connexivum and parts of tergum exposed ; reaching genital lobes ( cf ) or base of tergum VII (9) ; corium extending a little beyond ( cf ) or to (9) suture between terga IV and V, its basolateral borders expanded, rounded, and moderately reflexed. Abdomen ovate, longer than its maximum width across segment IV ( cf 80:66, 9 106:84); lateral borders convex, very finely serrate, and slightly notched at PE angles of connexiva; genital lobes (Figs. 7 — 8) slightly convex posteriorly, more so in female. Color greyish brown; antennae, head, pronotum, scutellum, and fore half of connexiva slightly darker; lateral borders of pronotum in middle whitish; basolateral expansions of corium, and round, callous spots on connexiva III to VII also whitish; some specimens almost uniformly dark grayish-brown. Total length: cf 5.75 mm., 9 7-20 mm.; width of pronotum: cf 1.88 mm., 9 2.24 mm.; width of abdomen: cf 2.64 mm., 9 3-36 mm. Aradns lawrencei is related to both Aradus basalis Parshley and Aradus oviventris , n. sp. From the former it differs by the shorter antennal segment II, which is equal in length to the interocular space and one eye only, the much shorter rostrum, reaching only to the middle of the prosternum, the smaller size and different coloration. From the latter species it differs by the shorter rostrum, relatively shorter head (only slightly longer than its width through eyes), relatively shorter antennal segment II, much flatter pronotum, with obsolete, or very shallow interlobal depression, and different colora- tion. REVISION OF THE GENERA GNOSTUS AND FA BRA SI A (COLEOPTERA: PTINIDAE) By John F. Lawrence1 and Hans Reichardt2 The species of Gnostus and Fabrasia are the only myrmecophilous Ptinidae occurring in the New World, although the family is rep- resented in South Africa and Australasia by at least 12 genera of inquiJines. Gnostus comprises three species ranging from Bolivia and central Brazil to Florida and the Bahamas and all apparently associated with ants of the genus Crematogaster. The three species of Fabrasia are all South American, and two of these (described below) have been taken with ants of the genus Camponotus. Gnostus jormicicola was described by Westwood (1855) from material collected with ants in Brazil. Although the species bore a distinct resemblance to the Paussidae in the structure of the legs, antennae, and prothorax, Westwood concluded, after a thorough comparative study, that it was “most nearly allied to such of the Xylophaga of Latreille as possess five-jointed tarsi . . Later au- thors were more impressed by the specialized features of the species, and it was placed in a distinct family by Gemminger and Harold (1868). Wasmann (1894) and others placed it near the Paussidae, Pselaphidae, Scydmaenidae, or Ectrephidae. Gnostus remained a genus of dubious affinities until Forbes (1926) pointed out the similarities in wing venation and wing-folding pattern to members of the family Ptinidae. In the present study, it was found that the male genitalia of G. jormicicola and G. floridanus Blatchley (Figs. 13-14) are of the same general type as that of Fabrasia , Ftinus, Pseudeurostus , Gibbium , and several of the Old World Ectrephinae (Hinton, 1941; Sharp and Muir, 1912). The ectrephines have al- ready been placed in the Ptinidae by Wasmann (1916) and will be treated in a forthcoming paper (in preparation). The genus Fabia was proposed by Martinez and Viana (1964) for a Brazilian species (F. alvarengai) , but no mention was made of myrmecophilous habits in the species; the name was later changed to Fabrasia (Martinez and Viana, 1965) because of homonymy. Last year we found specimens of a second species of Fabrasia among the William Morton Wheeler inquiline collection in the Museum of Comparative Zoology, and shortly after this Father Thomas Borg- Aluseum of Comparative Zoology, Harvard University, Cambridge, Mass. 2Departamento de Zoologia, Secretaria da Agricultura, Sao Paulo, Brazil. Manuscript received by the editor January 6 , 1966 30 1966] Lawrence and Reichardt — Ptinidae 31 meier presented us with a specimen representing another undescribed species. This led to a comparative study of ptinid myrmecophiles, part of the results of which are presented below. Since both Gnostus and Fabrasia have been made the types of subfamilies, it might be well to briefly discuss the classification and possible evolutionary origins of the two groups. Gnostus and Fabrasia differ considerably from any other group myrmecophilous ptinids in general form and in the presence of fully developed wings and hu- meral callosities; in this respect they are both similar to certain species of the genus Ptinus. They do resemble the Old World myrmecophiles in the modifications of the antennae, presence of glandular areas or trichomes, elongation of the clypeus, and reduc- tion of body hairs or scales, so that the surface appears shiny; most of these similarities probably represent adaptations associated with myrmecophilous habits, and similar trends are found in the Paus- sidae, the clavigerine Pselaphidae, and other groups of inquilines. Although Gnostus and Fabrasia are set apart from the other genera, Figs. 1 — 5. Gnostus flondanus Blatchley (from Plantation Key, Florida), Fig. 1, dorsal view; Fig. 2, lateral view; Fig. 5, antenna; Gnostus meinerti Wasmann (from Barro Colorado Island, Panama), Fig. 3, hind leg; Gnostus jormicicola Westwood (lectotype), Fig. 4, hind leg. 3 2 Psyche [March they are also quite distinct from one another, and it would be difficult to derive them from a common stem. The kinds of adaptive struc- tures occurring in the two groups are entirely different. The glandu- lar structures in Gnostus consist of trichomes on the pronotum, while those of Fabrasia consist of scattered pores on the1 elytra and a brush of hairs on either side of the distal articulation of the hind femur. The legs of Gnostus may be flattened and expanded like those of certain Paussidae, while the species of Fabrasia have pecul- iarly enlarged and club-like femora, which to our knowledge are unknown in any other beetle group. Characters of the head, antennae, prosternum, abdomen, and genitalia show similar differences. We think that the two genera were probably independently derived from Ptinus- like ancestors, perhaps early in the evolution of the family. Since most of the ant host species appear to nest in woody habitats, the beetles may have evolved from subcortical scavengers which had originally been derived from wood-boring ancestors (Crowson, 1955). It appears reasonable to retain the subfamilies Gnostinae and Fabrasiinae for the present, but in a revision of the family they may well be reduced to tribal rank within the Ptininae. In 1962, Costa Lima described another apparently myrmecophilous ptinid, Plaumanniola sanctaecatharinae, which he made the type of another subfamily. The type of this species has been examined, and it was found to belong to the family Scydmaenidae (Lawrence and Reichardt, 1966). Gnostus Westwood (Figs. 1 — 5, 13 — 15) Gnostus Westwood, 1855, Trans. Ent. Soc. London, (2)3: 90. (type species: Gnostus formicicola Westwood, by monotypy). description: Form elongate, convex; vestiture of short, suberect, sparsely distributed, fine hairs; surface shiny. Head strongly de- clined, only partly visible from above; vertex concave, forming a broad transverse impression, joined laterally on each side to a deep and narrow, oblique groove extending to the eye below; eyes lateral, relatively small, elongate and carinate ; antennal fossae relatively deep, located well above the eyes, separated by a sharp median carina; clypeus subtriangular, almost as long as wide, strongly tumid, and emarginate at apex; labrum inflexed and concave. Antennae (Fig. 5) 3-segmented, relatively short (not reaching base of pronotum) ; seg- ment I elongate, slightly curved, expanded towards apex; II about half as long as I, inserted beneath the latter, terminal segment almost twice as long as I and II together, expanded ventrally at middle and truncate apically, the tip excavate and filled with short hairs. 1966] Lawrence and Reichardt — Ptinidae 33 Pronotum longer than wide, constricted basally; anterior angles rounded ; disc somewhat flattened dorsally, with two more or less parallel longitudinal grooves, extending from two anterior pores pos- teriorly to a broad and deep, transverse impression, which is bordered on each side by two opposing, curved trichomes. Scutellum small, semicircular. Elytra more than twice as long as pronotum, almost twice as wide at humeri as pronoum at base; sides more or less parallel, expanded and broadly rounded apically; humeral callosities distinct; disc slightly convex dorsally, steeply sloping laterally; each elytron with 9 rows of fine punctures bearing short, suberect, fine, yellow hairs; apices of elytra individually rounded but only barely divergent. Hindwings fully developed. Prosternum before coxae flat, fairly long, intercoxal process laminate and depressed, so that the elongate coxae are almost touching. Mesosternum shorter than pros- ternum; mesocoxae separated by less than 0.25 times coxal width. Metasternum less than 0.50 times as long as wide and slightly longer than mesosternum, convex, median suture absent; metacoxae strongly transverse, separated by less than 0.20 times coxal width. Abdomen about as long as wide at base, strongly convex anteriorly; first 3 segments fused, the sutures between them indicated only laterally; 4th segment very narrow. Legs (Figs. 2 — 4) relatively short and simple; trochanters elongate-oval; femora and tibiae narrow and slightly flattened ( G. floridanus) or strongly flattened and expanded (G. formicicola and G. meinerti) ; tarsi short, the first 4 tarsomeres subequal. Male genitalia (Figs. 13 — 14) of the general type of Ptinus with median lobe articulating dorsally with bases of para- meres; basal piece relatively large, parameres very narrow, curved, converging apically. Total length : 1.70 — 2.53 mm. distribution: (Fig. 15): South America (Brazil, Bolivia, and Venezuela), Panama, Florida, and the Bahamas. The relatively narrow and elongate body form, presence of dis- tinct humeral callosities, and well developed scutellum and hindwings, will distinguish Gnostus from the majority of ptinid genera. From Ptinus and Niptinus it differs by the structure of the clypeus (which is more elongate), the lack of a metasternal suture, narrower pro- notum, and the modifications of legs and antennae. Gnostus is easily separated from Fabrasia by the 3-segmented antennae, position and size of eyes, longer and narrower prothorax, shorter legs, absence of metasternal suture, and partial obliteration of the first two ab- dominal sutures. The genus is most closely related to Ptinus and Fabrasia. All of the species of Gnostus appear to be associated with ants of the genus Crematogaster, and in the compilation below, at least five 34 Psyche [March species of this genus are involved. To our knowledge, no Gnostus has been collected with ants of any other group, and the record of G. meinerti from a termite nest needs verification. According to Professor E. O. Wilson (personal communcation) , only one speci- men of G. floridanus was found in a colony of C. cishmeadi Mayr on Plantation Key, although the ant is very common at that locality. Bates, on the other hand, as quoted in Westwood (1855), states that the formicaria of Crematogaster victima Smith almost invariably contained one, or at most two, Gnostus formicicola. Key to the species of gnostus 1. Tibiae only very slightly compressed, about 4 times as long as wide at apex (Fig. 2) ; longitudinal grooves of pronotum deep; Florida and Bahamas floridanus Blatchley — Tibiae strongly compressed and expanded apically, only 2 times as long as wide at apex (Figs. 3 — 4) ; longitudinal grooves of pronotum very shallow; South America and Panama .... 2 2. Posterior tarsus, when retracted, almost completely hidden in apical groove of tibia (only tarsal claws visible) ; longitudinal grooves of pronotum converging apically; Venezuela and Panama meinerti Wasmann — Posterior tarsus, when retracted, not completely hidden in apical groove of femur (5th tarsomere and claws visible); longitu- dinal grooves of pronotum subparallel; Brazil and Bolivia formicicola Westwood Gnostus formicicola Westwood (Fig. 4) Gnostus formicicola Westwood, 1855, Trans. Ent. Soc. London, (2)3: 92, pi. 8, figs. 1-21. type data: “Habitat in Brasilia, apud Santarem, in nidis Myrmicae ( Crematogastris) victima , Smith. D. Bates. In Mus. Britann., &c.” Lectotype, by present designation, $, with the fol- lowing labels: 1) “Co-type, Westwood Trans. Ent. Soc. 1855. p. 90. T. 8. Coll. Hope Oxon.” (red border), 2) “W Amazon Bates” (diamond-shaped), 3) “Gnostus formicicola Westwood B2 Bates,” 4) “Type Col: 235 Gnostus formicicola Westw. Hope Dept. Ox- ford” (Black border) ; deposited in the Hope Department of En- tomology, Oxford, England. Specimens Examined: Bolivia: La Paz, Mapiri, iv — ix. 1925, G. L. Harrington col. (i$, USNM). brazil: Mato Grosso , Utiar- iti, 325 m, viii.5.1961, K. Lenko col., with Crematogaster prope 1966] Lawrence and Reichardt — Ptinidae 35 brasiliensis Mayr (W. W. Kempf det.) (2 exx., CDZ) ; Para, San- tarem, 54 63 (1$, BM) ; Villa Nova, 55 75 ( 1 cT , BM) ; Amazonas , Sao Paulo [de Olivenca] (1 ex., BM) ; “Amazon, Bates” (1$, BM). This species was previously known only from the several specimens examined by Westwood; the only available specimen which can be assumed to be from the original type series is designated as lectotype above. Four other specimens in the British Museum are apparently not from the Bates material examined by Westwood. Gnostus formi- cicola seems to have a fairly wide distribution south of the Rio Amazonas. It appears to be very closely related to the northern G. meinerti , and only a few minor characters separate them. Since the two species are allopatric on the basis of presently known material, it is quite possible that they represent geographical races. Because of the paucity of available specimens, the lack of intermediate forms, and the absence of noticeable variation within either population, we prefer to retain them as distinct species. The original series of G. formicicola was taken from the nests of Crematogaster victima Smith, but Lenko has recently collected it in the colony of a Crematogaster near brasiliensis Mayr. Gnostus meinerti Wasmann (Fig. 3) Gnostus meinerti Wasmann, 1894, Kritisch. Verzeich.: 121, 216. type data: “Ein Exemplar von Dr. F. Meinert am Rande des Weges von las Trincheras nach Valencia (Venezuela) im Nest von Cremastogaster limaia Sm. am 5. Nov. 1891 entdeckt.” According to Wasmann the type is deposited in “Collect. Mus. Univ. Hafn. not examined. Specimens Examined: panama: Canal Zone , Barro Colorado Island, vii.17.1924, G. C. Wheeler col., with Crematogaster limata dextella Santschi (Santschi det.) (2 exx., USNM, CNHM) ; same locality, vi.1923, R. C. Shannon col., ex stomach of anteater (T amandua tetradactyla chiriquensis) (2$$, USNM) ; Erwin Island, vi. 17. 1923, R. C. Shannon col. (1 ex., USNM); Gatun Lake, Barro Colorado Island, Wheeler col., with Coptotermes niger Snyder (1 ex. USNM). Although Wasmann’s type was not examined, specimens of the Panama population agree well enough with the description to be considered conspecific with meinerti from Venezeula. The species seems to be restricted to northern South America, but further rec- ords will be needed to ascertain the limits of its distribution. Both of the known populations of this beetle have been collected in associa- Psyche, 1966 Vol. 73, Plate 2 Lawrence and Reichardt — Ptinidae 1966] Lawrence and Reichardt — Ptinidae 37 tion with Crematogaster limata Smith. The close relationship to G. formicicola has been discussed under that species. Gnostus floridanus Blatchley (Figs, i, 2, 5, 13, 14) Gnostus floridanus Blatchley, 1930, Ent. News, 36: 111-112, fig. 1. type data: “Type a unique (sex undetermined) in the author’s collection, taken near Dunedin, Florida, March 7, 1927” Location of type apparently in Blatchley Collection, Purdue University, Lafay- ette, Indiana; not examined. Specimens Examined: Florida: Plantation Key, vi. 17.58, E. O. Wilson col., with Crematogaster ashmeadi Mayr. Bahamas: New Providence, Clifton Point, iv.2.65, R. W. Hamilton & B. D. Valentine col., with Crematogaster sanguinea lucayana Wheeler (compared with type in the M. C. Z. collection). Gnostus floridanus wTas previously known only from the type locality, but the data presented above indicates that it may be more widespread in Florida and the West Indies. Blatchley’s specimen was beaten from Spanish moss, but this species, like its congeners ap- pears to be found only with ants of the genus Crematogaster . It is easily distinguished from, formicicola and meinerti by the characters given in the key, although all three species are fairly similar. Fabrasia Martinez and Viana (Figs. 6—12, 15) Fabia Martinez and Viana, 1964, Neotropica, 10(31): S, not Dana, 1851. (type species: Fabia a'lvarengai Martinez and Viana, by original des- ignation) . Fabrasia Martinez and Viana, 1965, Neotropica, 11(34): 18. description : Form elongate, strongly convex ; vestiture of short, suberect, sparsely distributed bristles; surface shiny. Head strongly declined, only partly visible from above ; vertex slightly convex, with elongate, narrow, median impression; eyes lateral, well developed, rounded; antennal fossae broad and shallow, located between eyes, separated by a slight tumidity (F. alvarengai) or a distinct carina (F. borgmeieri and F. wheeleri) ; clypeus subtriangular, almost as Explanation of Plate 2 Fabrasia borgmeieri. n. sp. (holotype), Fig. 6, dorsal view; Fig. 7, lateral view; Fabrasia wheeleri, n. sp. (paratype $), Fig. 8, dorsal view; Fig. 9, lateral view. 38 Psyche [March long as wide, strongly tumid, sharply indexed and emarginate at apex; labrum flat. Antennae 1 1 -segmented, extending at least to elytral hu- meri, moniliform to subserrate, the distal segments slightly wider at apices; segment I oval to oblong, 2 times as long as II which is sub- globular, III narrow and elongate, IV — VIII decreasing in length, IX — XI increasing in length; terminal segment distinctly expanded and truncate apically, the tip excavate and filled with short hairs. Pronotum slightly transverse, constricted basally; anterior angles pro- duced laterally forming 2 sharp flanges which partly conceal eyes ; disc with 6 medial tubercles in a transverse row. Scutellum small, semicir- cular. Elytra more than 3.50 times as long as pronotum; sides subpar- allel, but distinctly constricted just anterad of middle; humeral callosites well developed; disc somewhat flattened dorsally on either side of suture, steeply sloping laterally ; each elytron with 6 or 8 deep lateral striae and 2 dorsal striae, which may be obsolete except at base; punctation dual, consisting of small, setiferous punctures, which are seriate and located in the interstices, and larger, shallow ones, which contain glandular pores and are located within the striae; glandular pores in the lateral constrictions larger and closer together, giving the appearance of a transverse band ; apices of elytra slightly to strongly divergent. Hindwings fully developed. Prosternum before coxae tumid, fairly short, intercoxal process long, narrow but not laminate; procoxae separated by less than 0.25 times coxal width. Mesosiernum subequal to prosternum; mesocoxae separated by about 0.33 times coxal width. Metasternum about 0.50 times as long as wide and 1.50 times as long as mesosternum, strongly convex and rounded, median suture present; metacoxae transversely oval, sep- arated by 0.50 times coxal width. Abdomen 1.25 times as long as wide at base, strongly convex anteriorly, the sutures all distinct. Legs (Figs. 7, 9) long and highly modified; trochanters elongate-oval; femora enlarged and club-like distally, the hind femur greatly en- larged and irregularly formed, its tibial articulation with a dense brush of yellow hairs on each side; front and middle tibiae elongate, the hind tibia short and formed as in figure; first tarsomere more than 2 times as long as 2nd in front and middle tarsi, more than 3 times as long in hind tarsus. Male genitalia (Figs. 10 — 12) of the general type of Ptinus with median lobe articulating dorsally with Explanation of Plate 3 Fabrasia wheeleri , n. sp. (paratype$), aedeagus, Fig. 10, lateral view; Fig. 11, ventral view; Fig. 12, ventral view (with basal piece removed); Gnostus floridajius Blatchley (from Clifton Point, Bahamas), aedeagus, Fig. 13, ventral view; Fig. 14, ventral view (with basal piece removed). Psyche, 1966 Vol. 73, Plate 3 Lawrence and Reichardt — Ptinidae 40 Psyche [March bases of parameres; basal piece relatively small, parameres fairly broad and straight, slightly expanded at apices. Total length: 2.87 — 3.32 mm. distribution (Fig. 15) : South America (Brazil and Colombia). Fabrasia may be distinguished from the majority of ptinid genera by the body form, distinct humeral callosities, scutellum, metasternal suture, and fully developed hindwings. It differs from Ptinus and Niptinus by the clypeal structure, elytral constriction, glandular pores, and modifications of legs and antennae. It may be separated from Gnostus by the characters given under that genus. This genus, like Gnostus, seems to be related to Ptinus. Although the type species has not been collected with ants, both t . borgmeieri and F. wheeleri have been taken in the nests of species of Camponotus. It is interesting to note that the two ant hosts belong to the same subgenus ( Myrmothrix) , but more material will be needed to consider this anything but an accident of sampling. Camponotus bugnioni, the host of Fabrasia wheeleri , is known only from near the type locality in northern Colombia, while C. renggeri, the host of F . borgmeieri, is a common, widespread species, occurring from northern Argentina to northern Brazil. Nothing is known at present of the habits of these myrmecophiles ; a study of their be- havior might be particularly fruitful, considering the unique special- ized structures which they possess. Key to the species of Fabrasia 1. Elytra with 8 lateral striae, 2 dorsal striae visible for their entire lengths ; discal tubercles of pronotum conical and sharp ; apices of elytra only slightly divergent; hind tibia without pencil of black hairs; elytral bristles very short; larger species (3.17 — 3.30 mm.) ; Colombia wheeleri n. sp. — Elytra with 6 lateral striae, 2 dorsal striae obsolete except at base ; pronotal tubercles low and rounded ; apices of elytra strongly divergent ; hind tibia with 1 or 2 pencils of black hairs; elytral bristles longer; smaller species (2.87 — 3.15 mm.) 2 2. Hind tibia with a single pencil of hairs on outer side; clypeus distinctly swollen; antennal fossae joined by a transverse chan- nel; Brazil (Rio Grande do Norte) alvarengai Martinez and Viana — Hind tibia with 2 pencils of hairs, one on inner and one on outer side ; clypeus not swollen ; antennal fossae separated by a median carina; Brazil (Goias) borgmeieri n. sp. 1966] Laivrence and Reichardt — Ptinidae 4i Fabrasia alvarengai (Martinez and Viana) Fabia alvarengai Martinez & Viana, 1964, Neotropica, 10(31): 9, figs. 1-3. Fabrasia alvarengai (Martinez & Viana), 1964, Neotropica. 11 (34): 18. type data: “. . . i holotipo de Brasil, Rio Grande do Norte, Natal, VI-1951 (M. Alvarenga-leg.) , en la coleccion Moacir Al- varenga de Rio de Janeiro.” Holotype examined. This species is represented by a single specimen whose abdomen is missing so that the determination of sex is impossible. Fabrasia borgmeieri, new species (Figs. 6 — 7) description : Reddish-brown with dark brown spots on elytra ; clothed with short yellowish bristles, which are longer on elytra and more concentrated in median constriction, giving the appearance of a transverse band on each side. Plead with vertex shallowly im- pressed, post-antennal ridges weakly developed, antennal fossae rel- atively shallow and separated by a weak, somewhat flattened ridge; clypeus tumid, but not swollen as in alvarengai ; surface coarsely, densely and shallowly punctate and granulate. Antennae almost 3 times as long as pronotum; segment I 1.57 times as long as wide, expanded at apex. Pronotum 0.79 times as long as wide at base, an- terior angles moderately produced, the sharp flanges only slightly longer than medio-lateral tubercles, which are sharp and conical, the other discal tubercles low and rounded; surface rugosely punctate and granulate, as is scutellum, and clothed with short yellowish bristles. Elytra 1.94 times as long as wide at base and 3.81 times as long as pronotum ; base with large and rounded humeral callosities and a low rounded elevation on each side between callosity and suture; dorsal striae obsolete except at base, lateral striae 6 in num- ber, very deep in the vicinity of the medial constriction, where the glandular pores are larger and closer together; elytral apices strongly divergent and narrowed, forming 2 triangular processes; surface smooth and shiny, clothed with suberect yellowish bristles, which form brushes laterally in between the pores. Hind legs as in alvar- engai, but the tibia with a longer pencil of black hairs on outer side and a shorter one on inner side. Abdomen strongly convex anteriorly, but not impressed mesiaily. Total length: 2.87 mm.; greatest width (across humeral callosities) : 1.12 mm. type data: Holotype, probably 9, Campinas, Goias, Brazil, ii.1964, Schwarzmaier col., with Gamponotus ( Myrmothrix) renggeri Emery (T. Borgmeier, det.). Holotype deposited in the Departamento de Zoologia, Secretaria da Agricultura, Sao Paulo, Brazil. 42 Psyche [March F. borgmeieri is very similar in general appearance to F. alvarengai, sharing with it several characters (given in key) which are not found in F. wheeleri. F. borgmeieri is easily distinguished from alvarengai by the presence of a ridge between the antennal fossae, 2 pencils of hairs on the hind tibia, somewhat shorter antennae, and a less swollen clypeus. Since the specimen of alvarengai cannot be sexed, it has occurred to us that the two forms may represent male and female of a single species; the two forms are, however, separated by several characters of the type that are not involved in sexual dimorphism elsewhere in the Ptinidae. The two species1 are closely related, and when more material is studied, they may prove to be geographic races, which, like Gnostus formicicola and G. meinerti , occur on either side of the Rio Amazonas. Fabrasia wheeleri, new species (Figs. 8— 12) description : Reddish-brown with blackish elytral spots, which are larger and fewer in number than in borgmeieri ; clothed with very short yellowish elytral bristles; elytra with accumulations of white Fig. 15. Geographic distribution of the species of Gnostus and Fabrasia. 1966] Lawrence and Reichardt — Ptinidae 43 waxy secretion oil each side in the constriction, giving the appearance of a transverse band. Head with vertex deeply impressed, post-an- tennal ridges well developed and sharply defined, the antennal fossae deeper and separated by a sharp carina; clypeus tumid but not swol- len ; more distinctly indexed than in bor gmeieri ; surface very coarsely, densely, and deeply punctate. Antennae about 2.50 times as long as pronotum; segment I 1.33 times as long as wide, subglobular. Prono- tum 0.76 times as long as wide at base, anterior angles strongly produced, the sharp flanges distinctly longer than the medio-lateral tubercles, which are sharp and conical, the other medial tubercles higher and sharper than in borgmeieri , 2 comma-shaped antero-mesal longitudinal ridges bordering a median impression which extends to anterior edge; surface very coarsely and densely punctate, as is scutellum, and clothed with yellowish bristles, which are finer and longer than those on elytra. Elytra 1.93 times as long as wide at base and 3.77 times as long as pronotum; base with humeral cal- losities strongly developed and somewhat carinate, without antero- mesal elevations, but with a deep puncture on either side of the suture; dorsal striae distinct for their entire lengths, lateral striae 8 in number, containing glandular pores as in borgmeieri which are also concentrated in the constriction ; elytral apices only slightly diver- gent and rounded. Hind legs as in borgmeieri , but lacking a pencil of black hairs on the tibia. Abdomen of male with a distinct median impression on the first two visible sternites. Aedeagus (of paratype) elongate, basal piece subquadrate, subangulate basally, tegmen deeply cleft apically forming 2 fairly narrow and straight lateral lobes or parameres which are expanded and setose at their tips; median lobe narrow and elongate, curved dorsally at apex and at base, where it is attached to base of tegmen (Figs. 10 — 12). Total length: 3.17 — 3.30 mm. (type 3.17 mm.) ; greatest width (across humeral cal- losities) : 1.27 — 1.38 mm. (type 1.27 mm.). type data : Holotype, c? , paratype, Chubut, Patagonia (W. F. H. Rosen- berg) [USNM], Remarks. — The close similarity of the male genitalia of jaffueli and bergi indicates conclusively that the presence or absence of rugae on the female propodeum should not be regarded as a generic char- acter. The description and drawing of the genitalia of jaffueli were made from the specimen from Valle los Piuquenes. The genitalia of the allotype we-re extracted some years ago by R. R. Dreisbach, but the slide which he prepared contains only a broken fragment of the subgenital plate, the genitalia presumably having been lost. AMERICAN SPIDER GENERA THE RID ULA AND PA RA T IIE RID ULA (ARANEAE: THERIDIIDAE)1 By Herbert W. Levi Museum of Comparative Zoology Since my revision (1954) of the genus Theridula in North Amer- ica, no additional species have been found. In the meanwhile, how- ever, I have been able to examine the types of South American species (except that of Theridula polita , which appears to be lost). Addi- tional records of Paratheridula perniciosa have been obtained since I described the genus (1957), but no additional species. That Para- theridula perniciosa is a pantropical species is suggested not only by its wide distribution in the Americas, but also by the absence of related species. Its small size, inconspicuous appearance, and the fact that its epigynum is always covered by a secretion (which has to be removed to examine it), suggest that it may have been overlooked in collections. I would like to thank Dr. G. Owen Evans and his staff at the British Museum (Natural History) for making the Keyserling types available for study. A National Science Foundation grant (G-4317) permitted initial examination of theridiid types in European museums. The further researches were supported in part by Public Health Service Research Grant AI-01944 from the National Institute of Allergy and Infectious Diseases. Misplaced species: Theridula polita Mello-Leitao, 1947, Papeis Avulsos, Dept. Zool., Sao Paulo, 8(11): 127, fig. 1, 9- Female holotype from Santa Cruz, Est. Parana, Brazil, apparently lost, is probably a Dipoena judging by coloration, shape, and proportions of the species. The black stripe on the legs and the illustration of the epigynum suggest that it is D. militaris Chickering. The holotype of Theridula polita is not in the Mello-Leitao collection in Rio de Janeiro nor in Sao Paulo. Some Mello-Leitao types are alleged to be in the Museum Paranaense, but according to Mr. P. de Biasi it is not there either. The illustration of Mello-Leitao has been copied (Fig. 6). Theridula Emerton Theridula Emerton, 1882, Trans- Connecticut Acad. Sci. 6:25. Type species Theridion sphaerula Hentz, 1850 D=T. opulenta ( Walckenaer) ], how- ever Emerton misidentified the specimens which he examined; they belong to T. emertoni. 1 Manuscript recived by the editor May 11, 1966 123 124 Psyche [June Map 1. Distribution of Theridula emertoni Levi and Theridula opulenta (Walckenaer) . Carapace as in Theridion, not modified. Chelicerae with two teeth on anterior margin, none posterior. First leg longest with patella and tibia 1.2 to 1.5 times carapace length. Abdomen wider than long in females, and lacking a colulus. Male palpus very simple, without conductor, median apophysis or radix. Haematodocha fastens both ends of tegulum to cymbium. Key to Theridula species in the Americas References to “figures” in lower case signify illustrations in Levi 1954, to “Figures” in upper case signify the ones in this paper. 1 a. Females 2 ib. Males 9 2a. Epigynum with two distinct circular openings (Figs. 2, 5) •■••• ••••■•• 3 2b. Epigynum with humps or otherwise no distinct opening visible 5 3a. Connecting duct with one loop (fig. 19) gonygaster 3b. Connecting ducts not looped (Figs. 1, 4) 4 4a. Openings connected by a posterior lip (Fig. 2).... multiguttata 4b. Openings on a hump not connected by a lip (Fig. 5) nig err im a 1966] Levi — Theridula and Paratheridula 25 5a. Canada, United States 6 5b. Latin America 7 6a. Epigynum with a dark anterior semicircle (fig. 3) emertoni 6b. Epigynum with a dark posterior transverse area (fig. 11) opulent a 7a. Epigynum with two projecting bars (fig. 8) casas 7b. Epigynum with humps 8 8a. Abdomen brown, usually with a black dorsal spot faceta 8b. Abdomen black with white spots puebla 9a. Palpal embolus as long as tegulum is high IQ 9b. Palpal embolus shorter than tegulum high 11 10a. Embolus tightly coiled (Fig. 3) nigerrima 10b. Embolus loosely coiled (fig. 18) gonygaster 1 1 a. Canada, United States 12 11b. Latin America 13 12a. Palpal embolus straight with a notch on tip (fig. 9) opulenta 12b. Palpal embolus undulating (fig. 1) emertoni 13a. Embolus as wide as long (fig. 23) puebla 13b. Embolus longer than wide 14 14a. Total length 1.4 to 1.9 mm faceta 14b. Total length about 2.5 mm casas Map 2. Distribution of Theridula faceta (O. P. - Cambridge) and Theri- dula gonygaster (Simon). 126 Psyche [June Theridula emertoni Levi Map i Additional record. Newfoundland. Junction Pond, Notre Dame Camp, 13 July 1961, 9 (C. P. Alexander). Theridula opulenta (Walckenaer) Map 1 Additional record . United States. Oregon: Crater Lake Natl. Park, 15 June 1947, $ (B. Malkin and S. Sargent). Theridula faceta (O. P. - Cambridge) Map 2 Additional record. Mexico. Tabasco: 3.2 km NE of Comalcalo. Theridula puebla Levi Map 3 Additional records. Panama. Canal Zone: Madden Dam, 8 Aug, 1939 (A. M. Chickering) ; Barro Colorado Island, several records (A. M. Chickering). Boquete, several records (A. M. Chickering). Figs. 1, 2. Theridula multiguttata Keyserling. 1. Female genitalia, dor- sal view. 2. Epigynum. Figs. 3-5. T. nigerrima (Petrunkevitch) . 3. Left palpus. 4. Female geni- talia, dorsal view. 5. Epigynum. Fig. 6. T. polita Mello-Leitao, epigynum after Mello-Leitao. 1966] Levi — Theridula and Paratheridula 127 Theridula gonygaster (Simon) Map 2 Thcridium gonygaster Simon, 1873, Mem. Soc. Roy. Sci. Liege, (2)5:109. Male lectotype designated by Levi, 1954, in the Museum National d’His- toire Naturelle, Paris. Theridula gonygaster y — Levi, 1954, Trans. Amer. Microscop. Soc., 73:340, figs. 18-22, $ $. Distribution. Worldwide tropical. In America from Arizona, southern Florida, eastern Mexico, West Indies to Rio de Janeiro, Brazil. Additional records. United States. Arizona: Cochise Co. Cave Creek Canyon. Columbia. Valle: Buenaventura, 4 Nov. 1950 (E. I. Schlinger, E. S. Ross). Ecuador. Guavas: Milagro (H. E., D. L. Frizzell); Guayaquil (Landes, H. E., D. L. Frizzell). El Oro: 5 km above Pasaje (R. Walls). Peru. Piura: Sulana, banana grove (H. E., D. L. Frizzell) ; Higueron, Las Lomas (H. E., D. L. Friz- zell). Luna: San Antonio (E. I. Schlinger, E. S. Ross). Theridula nigerrima (Petrunkevitch) Figures 3-5, Map 3 Dipoena nigra Keyserling, 1886, Die Spinnen Amerikas, Theridiidae, 2(2): 43, pi, 12, fig. 158, $ $. Female lectotype here designated from Pumamarca [PPumamanta, Junin], Peru, in the Polish Academy of Sciences, Warsaw; examined. Name preoccupied by D. nigra Emerton, 1882. Dipoena nigerrima Petrunkevitch, 1911, Bull. Amer. Mus. Nat. Hist., 29: 174. New name for D. nigra Keyserling, preoccupied. Diagnosis. The openings of the epigynum are on the anterior edge of a sclerotized plate (Fig. 5), quite different from T. gonygaster which lacks this plate. The connecting canals seem straight and short (Fig. 4). The shape of the narrow pointed embolus (Fig. 3) separates T. nigerrima from T. gonygaster. Records. Ecuador. Azuay: Cuenca, 3 Apr. 1942, $ (H. E., D. L. Frizzell). Peru. Cuzco: Machu Picchu, 3000 m elev., 6 Mar. 1947, 2 cT (J. C. Pallister) . Theridula multiguttata Keyserling Figures 1, 2. Map 3 Theridula multiguttata Keyserling, 1886, Die Spinnen Amerikas, Theridiidae, 2: 258, pi. 21, fig. 306, $. Female holotype from Blumenau, [Santa Catarina], Brazil, in the British Museum (Natural History) ; examined. The shorter straight connecting ducts (Fig. 4) separate this species from T. gonygaster. 128 Psyche [June Map 3. Distribution of T heridula puehla Levi, T. casas Levi, T. niger- rima (Petrunkevitch), and T. multiguttata Keyserling. Record. Brazil. Pernambuco: Recife (in the Senckenberg Museum). Paratheridula Levi Paratheridula Levi, 1957, Trans. Amer. Microscop. Soc., 76:105. Type species by original designation, Mysmena quadrimaculata Banks [ =Paratheridula perniciosa (Keyserling)]. Carapace as in Theridion, not modified. Chelicerae with two teeth on anterior margin, a small tooth on posterior of female. First leg longest, patella and tibia 1.3 to 1.4 times carapace length. Abdomen subspherical and without colulus. Male palpus very simple without conductor, median apophysis or radix, and having only a basal hema- todocha. The single species known is widely distributed ; it is suspected that it is introduced. Paratheridula perniciosa (Keyserling) Map 4 T heridium perniciosum Keyserling, 1886, Die Spinnen Amerikas, Theridiidae, 2 (2): 233, pi. 20, fig. 288, $. Female type from Blumenau, [Est. Santa Catarina], Brazil, in the British Museum (Natural History) ; examined. 1966] Levi — Theridula and Paratheridula 129 Map 4. Distribution of Paratheridula perniciosa (Keyserling) . 130 Psyche [June Mysmena quadrimaculata Banks, 1896, Trans. Amer. Entomol. Soc., 23 : 66. Female holotype from Punta Gorda, Florida, in the Museum of Com- parative Zoology. Paratheridula quadrimaculata , — Levi, 1957, Trans. Amer. Microscop. Soc., 74; 106, figs. 1-6, map. Distribution. Gulf states, Mexico, Central America, West Indies, South America to Chile, possibly pantropical. Additional records. Mexico. San Luis Potosi: Valles, July 1956 (W. J. Gertsch, V. Roth). Cuba. Villas: Soledad. Bahama Island. New Providence (J. Greenway). Ecuador. Napo-Pastaza: Rio Topo, June 1943 (H. E., D. L. Frizzell). Peru. Junin: San Ramon, 800 m (H. W. Levi). Brazil. Pernambuco : Recife. Chile. Osorno: Valley Forest, 18 km W of Purranque, Jan. 1951 (E. S. Ross, A. E. Schlinger) ; 10 km E of Puyehue, Jan. 1951 (E. S. Ross, A. E. Schlinger). References Cited Levi, H. W. 1954. The spider genus T heridula in North and Central America and the West Indies. Trans. Amer. Microscop. Soc. 73: 331-343. 1957. The North American spider genera Faratheridula, Tekellina, Pholcomma , and Archerius. Ibid. 76: 105-115. COLONIZING FLIGHTS AND ASSOCIATED ACTIVITIES OF TERMITES. I. THE DESERT DAMP-WOOD TERMITE PA RA NE O TERM ES SIMPLICICORNIS ( KALOTERM ITIDAE ) 1 By W. L. Nutting Department of Entomology, University of Arizona Introduction. — The swarming period provides the only oc- casion when observations can be made undisturbed on, many species of termites. Flights involving large numbers of alates, apparently synchronized with particular seasons, are a conspicuous phenomenon in many parts of the world and well known to local inhabitants. Yet only the most limited information is available on the flights of a small percentage of the commoner species, i.e., the months and time of day during which flights are staged, and perhaps an associa- tion with rainfall. Not until recent years have many real efforts been made to correlate the dispersal flights of termites (Snyder, ’6i) or the nuptial flights of ants (Kannowski, ’59; Talbot, ’56; ’64) — which appear to be similarly related to definite weather patterns — with even the most obvious environmental factors. The difficulties involved in gathering more detailed intelligence, however, become apparent after a season or two of field study. The following account is the result of an increasingly successful series of observations made on one of twelve different species in a single area over a period of seven years. It summarizes field notes on 78 separate flights together with a preliminary analysis of accom- panying weather data. The ultimate objective of such a study should be the elucidation of the physical and physiological factors which trigger the flights of a species, both seasonally and daily, over its entire range. At present the bulk of the data is climatological and far outweighs that on the behavior of the termite itself; the study is thus perforce largely descriptive. Succeeding papers in this series should gain in significance from data being gathered concomitantly on additional species in the families Kalotermitidae, Hodotermitidae, Rhinotermitidae and Termitidae. Although the actual flight stimuli may eventually prove very different from those which suggest them- selves from time to time, this descriptive state should provide a start- ing point for other more profitable approaches to the subject. Arizona Agricultural Experiment Station Journal Article No. 1112. Manuscript received by the editor February 25, 1966 132 Psyche [June Paraneotermes simplicicornis is apparently unique among the dry- wood termites (Kalotermitidae) for, although the reproductive cen- ter has never been found, it is fundamentally subterranean in habit. Over its known range in the deserts from southern California and northern Baja California to southern Nevada, southwestern Texas and northern Sinaloa, Mexico, it typically attacks moist wood in or on the ground in washes and canyons. Small living trees are even cut off' just beneath the soil on occasion. A detailed account of the biology, distribution and taxonomic relations of this most interesting termite was published by Light in 1937. At that time swarming had never been observed, nor has it been reported since. Study area and methods. — Nearly all the flights were ob- served within an area of less than an acre in north-central Tucson, Arizona, at an altitude of 2400 ft. This part of the valley floor is now about equally divided between creosote bush desert (Larrea tridentata) and small residential developments. Three to four feet of fine-textured alluvium overlies a narrow zone of friable caliche or hardpan. The mean annual precipitation of 10.9 inches is almost equally divided between a summer (July-Sept.) and a winter (Dec.- March) rainy season, characteristic for most of the Sonoran Desert. Although the mean annual temperature is I9.6°C, it is of little sig- nificance unless it is realized that daily fluctuations of 150 are common and of 2i°C not unusual. In 62 years the highest recorded temperature was 44.40 (July, 1953, and several other dates), the lowest — I4.4°C (Jan., 1913) (Sellers, ’60). Considering our scant knowledge of the types of sense organs with which the various castes of termites are equipped, much less the in- formation which they are collecting, it is still far from clear what environmental factors should be measured, to say nothing of the necessary degrees of accuracy. Hence, many of the methods used here are admittedly exploratory and probably crude. Certain refine- ments have already been made and others are continually suggesting themselves. For example, it would probably be desirable to measure many microclimatic parameters; however, their usefulness will de- pend upon a much greater knowledge of the termites and their pre- flight activities within and near the nest than is now available. Starting in January, i960, continuous records of air temperature and relative humidity have been made in the study area with a Ben- dix-Friez Hygrothermograph, Mod. 594 (maintained to an accuracy within approximately z±i°C from —12 to +43°C and ±5% from O to 100% RH). The instrument is sheltered six feet above the ground. Weekly mean temperatures and relative humidities were 1966] Nutting — Paraneotermes 33 figured from the hygrothermograph records with the aid of a plani- meter (Cutright, ’27). Evaporation data are available (U. S. De- partment of Commerce, ’62; ’63) from measurements made daily at 5 :00 PM on the campus of the University of Arizona, Tucson, 2.5 mi. SSW of the study area. Saturation deficit was calculated for time of sunset plus 15 minutes at Tucson, and for certain other times, on the basis of hygrothermograph records. Rainfall was meas- ured daily with a simple plastic collecting gauge near the center of the area from midnight to midnight. Illumination at the zenith was measured several different times with a Weston photographic ex- posure meter (Master II, Mod. 735). During 1962 barometric pressure was continuously recorded on a temperature-compensated Taylor recording barometer. Some observations were made on flights from a laboratory colony in an 8 X 8-foot walk-in refrigerator which was equipped with a Partlow recording temperature program control (Mod. RCS) capable of providing any desired daily or weekly temperature pattern. All times are Mountain Standard and based on the 24-hour clock. Development of alates and pre-flight behavior. — Light (’37) termed the large aggregations of nymphs and soldiers of this species “temporary outposts” or “foraging subcolonies” ; no repro- ductives of any form, eggs or small nymphs have yet been found in them. Samples from 19 subcolonies collected in or within 25 miles of Tucson have revealed subimaginal nymphs (with large wing pads) to be present between August 30 (1962) and May 2O1 (1961). On the latter date an entire subcolony was removed from a large timber beneath the steps of a farm building and set up in a laboratory ob- servation nest. It contained 39 soldiers and 760 nymphs, many of them pre-alates. Between June 2 and 15, all of these had developed into 130 alates at laboratory temperatures only slightly less variable than, and averaging a few degrees below, those outside. By a most fortunate coincidence another subcolony was found on June 14, 1961, in the roots of a dead palo verde tree ( Gercidium microphyllum) . It contained 104 soldiers, at least 1100 nymphs (no pre-alates) and 190 alates. No flights were observed in the Tucson area in 1961 until July 5, although it is likely that this was not the first flight of the season. The above evidence shows an abrupt and rapid mass develop- ment of alates within three weeks, probably less, of the first flight. Although no local soil temperature records are available it seems reasonable to assume that this maturation is a direct result of the intensei heating of the soil which occurs during the very clear, dry months of May and June. Monthly mean air temperatures increase 134 Psyche [June rapidly from i8.i°C in April to 22.40 in May, 27.6°* in June and 30. 1 0 in July (Sellers, ’60). Light (’37) reported alates in colonies in California from late October into November and in Nevada during May. Snyder ( 54) noted that alates were taken from a colony in Texas in mid-April. In the absence of flight records from these regions speculation is unprofitable, although flight schedules are probably shifted in con- formity with local weather patterns which result from altitudinal and latitudinal differences over its entire range. Once alates are present in a colony, the influence of biological factors on the initiation of flights must also be considered. Although the behavioral interactions among the members of a colony have only begun to be fathomed, the feverish activity exhibited within a colony at the onset of conditions favoring flight has probably been noted many times (e.g., in Kalotermes [= Incisitermes ] minor Hagen by Harvey, ’34, p. 221). In this connection some early attempts were made to induce flights from a large sub-colony which had been in the laboratory for about a month. It was confined in an observa- tion nest of seven plastic petri dishes interconnected with polyethylene Figure 1. Flight season of Paraneotermes simplicicornis in relation to rainfall, weekly mean temperature and relative humidity at Tucson, Arizona, in 1962. Dates showing 0.1" actually indicate amounts from a trace to 0.1" of rain. Flights began when mean temperature remained above about 23°C. The upper section shows the numbers of alates observed on each date during the flight season from June 9 to Sept. 7. 1966] Nutting — Paraneotermes 135 tubing and placed on a large tray of moist soil. ( Paraneotermes has been maintained for more than a year in this type of nest with the petri dishes containing a layer of 4% agar and thin slices of wood.) The colony was held in the dark, walk-in refrigerator for about a week at 27°C and 30% RH prior to a brief, simulated flight season. Observations were made with the program control set to duplicate the temperature cycle of five consecutive days of typical summer weather. Daily temperatures varied between 22° and 43°C, relative humidities between 28 and 78%. Temperature peaks occurred be- tween 1500 and 1600 each day. During the middle of these days the temperature sometimes rose at a rate of about 3.3 °C per hour and fell later by as much as 9°C per hour for short periods. No measurements were made within the nest, although it is certain that the relative humidity was continuously at or close to saturation. Most cf the observations were made under a red light. Under such conditions and below about 34°C, activity in the colony appeared to be normal, with the alates remaining almost motionless in one or more tightly packed clusters. As the temperature rose between about 340 and 36°C, the general activity gradually in- creased; the alates began to run about and the clusters broke up. Above about 36°C the entire colony became more and more agitated and the alates leaped and fluttered their wings in frenzied excitement. As the temperature dropped this behavioral sequence was reversed. In the range from 38° to 36°C, the alates again became gregarious — in one instance within 1 5 minutes from their peak of activity. On three days, small numbers of these excited alates (15, 13, 2) actually emerged from the nest through a small hole drilled in the cover of one petri dish as well as from under the edge of the cover. In the latter case two soldiers were obviously and attentively stationed near the point of emergence although they did not appear to be regulating the exit of the alates as suggested for Kalotermes (— In - cisitermes) minor by Harvey (loc. cit.) . The emerging alates rapidly gained high points on the soil and small stones in the tray, then took flight after much moving of heads from side to side and waving of antennae. The door of the chamber was opened and observations were terminated after the alates had made rapid, erratic flights into the day-lighted room outside. These limited observations show that the phases of flight activity, even within the colony, are amenable to laboratory study. Although the preceding manipulations did not duplicate the environmental conditions under which this essentially subterranean termite stages its flights, they do suggest that high temperature (probably reached 136 Psyche [June later in the day near the soil surface) may provide an important cue for the initiation of daily flight periods. As more data become avail- able further refinements in this type of experimentation are planned with several other species of termites in Arizona. Flight season. — According to legend in the Sonoran Desert, the summer rainy season begins on San Juan’s Day, June 24. The rains are also believed to initiate the swarming season of the termites in this region ; indeed, in many Latin American localities with a similar weather pattern, the swarming alates are called “Palomitas de San Juan” (Light, ’34). Legend and meager published informa- tion led to some early and rather casual observations which resulted in one flight record for Parccneotermes on August 12, 1956; three in 1957, on July 10, 14 and 16; and one on July 17, 1959. More persistent observation, still within the rainy season, yielded 1 1 flights from July 5 through 29, 1961. Guided by experience of previous seasons and new knowledge that alate production may begin in early June, intensive observations were made from mid-May to mid- September in 1962 and, with the exception of six weeks of June and July, again in 1963. Thirty-one flights were thus recorded from June 9 through Sept. 7, 1962 (fig. 1), while the first and last flights in 1963 were recorded on June 15 and August 19 (fig. 2). These records spanning 91 and 66 days attest to a considerably protracted flight season for this species, beginning well in advance of the rainy season. A flight on May 29, 1965, further expands the cumulative seasonal record to 102 days within five consecutive months. T. E. Snyder has informed me (’64, in litt.) that the U. S. National Museum contains alates caught by light trap on July 6 and August 15, I947» at Blythe, Riverside Co., California. Influence of environmental factors on time of flight. — The flights of termites are considered to provide for the foundation of new colonies and hence the maintenance and dispersal of the species. Extensive observations on the termite fauna in Mexico and the southwestern United States have led the writer to conclude that matings between sexes from different colonies, in synchronous flights ever limited areas, are presumably usual. The idea that sibling matings are the rule has been advanced by several authors (Weesner, ’60, p. 1 61) although, admittedly, either position would be equally difficult to prove. Since alates of Paraneoterrnes have never been observed emerging from a nest in nature, the following factors ac- tually describe the conditions under which flights are known to have occurred. A more refined analysis does not yet seem warranted on the basis of such data, limited as they are to only two seasons. A 1966] Nutting — Paraneotermes 137 Figure 2. Flight season of Paraneotermes simplicicornis in relation to rainfall, weekly mean temperature and relative humidity at Tucson, Ari- zona, in 1963. Dates showing 0.1" actually indicate amounts from a trace to 0.1" of rain. Flights began when mean temperature remained above about 23 °C. The upper section shows the numbers of alates observed on each date during the flight season from June 15 to Aug. 19. No observa- tions were made between June 24 and August 1. Data on the maturation of alates were obtained from two colonies, one in the laboratory and one in the field. few simple correlations are apparent, however, and may provide a basis for determining the cues which are actually used by Paraneo- termes for initiating both its annual flight season and daily flight periods. T emperature The timing of the flight season in relation to weekly mean tempera- ture and relative humidity for two years is plotted in figures 1 and 2. Flights began when the weekly mean temperature remained con- sistently above about 23CC, while later in the season means reached as high as 29.90 (’62) and 30.6°C (’63). The seasons ended during weekly means of 25.90 (’62) and 25.4°C (’63), leaving three and seven weeks in these years before falling to 23 °C. At this point it might be mentioned that the areas under the curves for the two years studied — taken arbitrarily from the lowest weekly mean of the winter to the date of the first flight (figs. 1 and 2) — are very close as measured with a planimeter. This has suggested that it Relative Humidity, % (weekly mean) 138 Psyche [June might eventually be profitable to explore modifications of the tem- perature-summation method in connection with alate development both in nature and in controlled environments, and then to relate results to the initiation of the flight season. However, the list of unknowns concerning this species is so great — location and environ- ment of the nest, favorable temperature ranges for the development of any caste, etc. — that meaningful calculations are hardly possible at present. The simplest and perhaps most plausible explanation for the termination of the flight season might lie in the fact that all alates had flown. Figure 3 shows the flight periods in relation to the following daily temperatures during the season in 1962; minimum, maximum and temperature 15 minutes after sunset. Daily flight periods (29 in 1962, 11 in 1963) occurred on days with a mean minimal tempera- ture of 20.1° (range, 12.0 — 24.8) in 1962, and 19.O0 (range, 13.7 — 22.0) in 1963. The mean maximal temperature for the same days was 33. 30 (range, 26.9 — 37-6) in 1962, and 32. 30 (range, 29.5 — 36.7) in 1963. Flights began each day at approximately 15 minutes after sunset; the mean temperature for this time, on the days when flights occurred, was 28.2° (range, 23.4 — 32) in 1962, and 26.O0 (range, 23.7 — 29) in 1963. Within these limits there is no readily apparent relation to daily temperature patterns. The temperature cycle of a “typical day,” together with curves Figure 3. Flight dates of Paraneotermes simplicicornis in relation to daily temperature fluctuations in Tucson, Arizona, in 1962. Flights began at approximately 15 min. after sunset. Rain or wind may have prevented flights on at least 15 evenings. 1966] Nutting — Paraneotermes 139 for relative humidity and saturation deficit, is presented in figure 4. The combinations of temperature and relative humidity at the time of sunset plus 15 minutes are plotted in figure 6. Daily flights almost always begin while temperature is falling steadily and generally near or shortly after the low in humidity for the day. Moisture Rainfall and Soil Moisture. — The relation of the flight season and the daily flights to rainfall for 1962 and ’63 is shown in figures 1 and 2. For these years, where the first and last flights were estab- lished with reasonable certainty, the flight season began well in ad- vance of the summer rains. In 1962 the first flight was staged 81 days after the last measurable rain of 0.09" was recorded on March 20. Eight flights had occurred before the next measurable rain (0.02") had fallen 99 days later. The next year, 51 days elapsed between the last measurable rain of 0.58" on April 25 and the first flight, with at least three flights occurring before the first measurable rain (0.01") in 69 days. In 1965 the first flight took place 17 days after 0.0 1" was measured on May 12. These records established beyond any doubt that Paraneotermes does not begin swarming in response to seasonal rainfall. This is apparently the case with many dry-wood termites, such as Cryptotermes (Weesner, ’60, p. 160). Although a few of the larger flights have taken place on days fol- lowing substantial rains, the general pattern seems to be one in which rain is actually avoided (figs. 1 and 2). Of 29 flights observed in the study area during 1962, only eight occurred on rainy days, and on each of these days only 0.05" or less was measured. Further, flights did not take place if any rain occurred within an hour before or after flight time. A similar pattern is evident for other years, although the data are much less complete. No measurements of soil moisture have been made in connection with the flights of this termite. It is certain, however, that the gen- eral floor of the desert is hard-packed and that the soil is extremely dry to a depth of many feet during the early part of the season. The matter of the selection of nesting site is briefly covered under Post- Flight Behavior. Relative Humidity. — Figures 1 and 2 show the relation of the total flight season to the weekly mean relative humidity for two years. The seasons took place within means of 50. 1 and 79.1% (’62) and 40.6 and 86.9% relative humidity (’63). These ranges include very nearly the extremes for each year, and cover the summer trend from very low to high mean humidities. 140 Psyche [June Figure 4. Relationship between approximate flight time (sunset plus 15 min.) of Paraneotermes simplicicornis and temperature, relative humidity, saturation deficit and evaporation. June 20 was taken as a typical day during the 1962 flight season in Tucson, Arizona. In relation to daily fluctuations of relative humidity, individual flight periods (29 in 1962, 11 in 1963) occurred as follows: on days with a mean minimum humidity of 45.8% (range, 37 — 65) in 1962, and 47.4% (range, 30 — 65) in 1963; on days with a mean maxi- mum humidity of 85.9% (range, 62 — 100) in 1962, and 91.4% (range, 65 — 100) in 1963. Mean humidity at flight time (sunset plus 15 minutes) was 56.8% (range, 38 — 92) in 1962, and 71.3% (range, 36 — 100) in 1963. When the rains begin, usually in July, fluctuations in relative humidity vary widely during each day as well as from day to day. There were several days during the two years studied when the relative humidity at the time of sunset was at or close to 100%. Daily flights of Paraneotermes almost never occurred on such days; indeed, probably as a result of its avoidance of rain, it rarely flew 1966] Nutting — Paraneotermes 141 when the relative humidity was as high as 80 or 90%. It is hardly necessary to point out that the relative humidity at the time of flight, generally well below 60%, is closely correlated with the high tem- peratures during afternoon and evening (fig. 4). Evaporation and Saturation Deficit. — Since evaporation was measured 2.5 miles from the study area, the following figures are included only as a general indication of the evaporative power of the atmosphere in a region occupied by Paraneotermes. Figure 5 shows that flights occur during the season with the highest evaporation rate of the entire year. The mean daily evaporation rate for 29 flight days in 1962 and 11 in 1963 was 0.36" (range, 0.10 — 0.5 1) in 1962, and 0.30" (range, 0.10 — 0.56) in 1963 (U. S. Department cf Commerce, ’62; ’63). The overriding influence of high tempera- ture at this time of year insures a high level of evaporation in spite of the moderating effect of the rains which are, at best, sporadic and only infrequently heavy. The saturation deficit at the study site provides another measure of the dryness of the atmosphere which the alates encounter during their brief period of actual flight. Also closely dependent on tem- perature, this factor varies about a very high level during the flight season. The weekly mean saturation deficit for the approximate time of flight (sunset plus 15 minutes) is plotted in figure 5. In 1962 the 29 daily flights began at a mean saturation deficit of 12.8 mm. Hg (range, 2.1 — 20.7 mm. Hg). The typically wide daily variation in saturation deficit is shown in figure 4. Although this information may be of little value at present, laboratory experiments might be designed to determine whether the alates respond to differ- ences in relative humidity or in evaporation (or saturation deficit), and thus provide a means for evaluating the effect of these factors on flight behavior. For example, perhaps the contrast between a relatively dry atmosphere and the higher moisture levels in micro- climates at the level of the soil provides gradients which are used by the alates, first in leaving the colony and later in seeking out suitable nesting sites. Light The rather consistent appearance of alates at lights in the early evening for several seasons suggested that light intensity might be a cue for the initiation of daily flight periods. The times of “first sightings” of alates during 1962 are shown in relation to time of sunset for Tucson in figure 7. Starting in 1963, measurements of light intensity at the zenith have been made with a photographic light 142 Psyche [June meter on a number of evenings during each flight season. Para- neotermes has never been observed in flight until the light intensity has thus registered approximately 0.5 lumen per square foot (foot candles) or less. As with Neotermes tectonae Damm. (Kalshoven, ’30) and probably other evening fliers, a few observations have shown that flight time is earlier on cloudy evenings. Considering the handi- cap of unknown nesting sites, the correlation between the beginning of the flight period and the time of sunset, or light intensity, is close. Whatever other advantages there may be, it is certain that many predators must be avoided by swarming at dusk. These few notes indicate the desirability of continuous and more sensitive measurements of light intensity in the vicinity of the emer- gence holes, whenever they can be found. The subject of rhythms and other aspects of behavior in the alates, nymphs and soldiers should also be studied in laboratory colonies under controlled illumination. Indeed, a most encouraging find has been made in this connection with certain species of ants. McCluskey (’65) has shown that en- dogenous activity rhythms of the males correspond to the timing of mating flight in the held. Flight behavior. — Considering the relative scarcity of evidence for Paraneoterm.es in the vicinity of Tucson and the numbers of alates appearing at lights, it is probable that there were no more than one or two colonies in or near the study area. On any basis, the number of alates involved in each daily flight would be very small. Assuming that most of the alates in 71 flights were from a single colony and visited the lights (front and back of house), the average Figure 5. Relationship between flight season of Faraneotermes simplici- cornis and evaporation rate and saturation deficit in Tucson, Arizona, 1962. Saturation Deficit, mm Hg at sunset + 15 min. (weekly mean) 1966] Nutting — Paraneotermes 143 number per flight was approximately five. The usual number seen during a flight was between one and 12, but on three occasions in seven years, the numbers were 23, 48 and 57. Both sexes have been taken in single flights, although no record of the distribution was made. Concurrent flights have been recorded five times: At 1.8, 2.6, 2.8, 12 and 13 miles from, the study area. Paraneotermes was observed in natural flight only as it flew in rapidly, between two and six feet above the ground, to lighted areas. After alighting on walls or pavement, individuals frequently leaped and flew about very erratically. Light (’37) and Light and Weesner C48) briefly described flights under what were probably abnormal conditions, and Light termed the species a strong and rapid flier. Although in a dry container alates will die in less than a day after capture, presumably of desiccation, they have been held for as long as two weeks in petri dishes containing 4% agar. Alates have been noted to remain in the vicinity of lights up to 80 minutes after their arrival. Post-flight behavior. - — Alates of this species were observed to lose their wings much more readily than many other members of Figure 6. Combinations of temperature and relative humidity at approxi- mate flight time (sunset plus 15 min.) for Paraneotermes simplicicornis at Tucson, Arizona. The limits of these conditions for 49 flights are shown by the closed curve. Conditions for the 91 evenings in the 1962 flight season are plotted as follows: x, at the start of 29 flights; solid circles, at flight time on 15 evenings when flights may have been prevented by rain or wind ; open circles, at flight time on 47 evenings when no flights were ob- served. In the latter case these conditions could have been limiting on 13 evenings where they fell outside the curve. 144 Psyche [June the Family Kalotermitidae. However, it is strongly suspected that the environment under which the following observations were made (residential area with irrigated plantings) is far more suitable for the rapid entrainment of normal post-flight behavior than for that cf other local, but strictly dry-wood species such as Incisitermes (—Procryptotermes) hubbardi. The de-alation of many adults around lights usually occurs within a few minutes to a half hour of their first appearance. This is accomplished through the usual bodily contortions and, frequently, by lunging forward and back- ward and rubbing the wings against the edges of cracks or protru- sions. Similar behavior was also observed by Light (’37). Using several groups of alates of both sexes taken at lights, it was found that alates would eventually die in possession of their wings when confined in a dry environment. In contrast, and even after such confinement for several hours, alates rid themselves of their wings within one to five minutes upon being released in a petri dish con- taining agar and decaying wood. Further, both sexes were observed to lose their wings within as little as 30 seconds after being placed in a dish containing de-alated individuals. (In similar manipulations with Amitermes emersoni \T ermitidae] , alates lost their wings on an agar surface which had recently been occupied by de-alated in- dividuals.) The readiness with which these alates divested themselves of their wings in the presence of others, albeit in a suitable environ- ment, suggests that a pheromone may be released by the de-alated forms which provokes, or at least facilitates, de-alation. Since a strict accounting of the sexes was not kept, it is possible that the female odor (either in calling or shortly before) may also stimulate this behavior in the males, or even in both sexes. Female alates frequently assumed the calling attitude soon after alighting on vertical or horizontal surfaces. The abdomen is typically raised at an angle of about 25 0 from the horizontal with the tip slightly downturned and thus roughly parallel with the surface. Under the artificial conditions provided in a petri dish containing agar and wood, a single, de-alated female assumed the calling attitude many times during three days’ confinement. On the third evening a winged male responded to her presence by dropping his wings and following her within one minute. These disconnected observations on Paraneotermes are strongly suggestive of the post-flight behavior of Neotermes tectonae described by Kalshoven (’30) : either sex may locate a suitable nesting site where it then survives, apparently capable of attracting a mate, for some weeks or months. In species such as these where small numbers fly rather frequently over a long season, Hours MST, Tucson, Arizona 1966] Nutting — Paraneotermes 145 Figure 7. Relationship between observed time of flight of Paraneotermes simplicicornis and time of sunset in Tucson, Arizona, in 1962. Inset scale gives approximate light intensity (in lumens per square foot) in the study area on a cloudless evening early in the flight season. No flights have ever been observed earlier than 15 min. after sunset or at a light intensity greater than approximately 0.5 lumen per square foot. the ability to survive and to attract a mate from succeeding flight periods is an obvious asset. Alates often formed tandem pairs on walls and on pavement shortly after their arrival at lights. In situations involving several members of both sexes, de-alated pairs were soon formed and these generally left the lighted area within a half hour. This movement appeared to result from the unsuitability of the featureless pavement in the vicinity of the lights, and perhaps from a preference for more humid areas of soil nearby, for they seemed to be completely indifferent to light. Light (’37, pp. 431-4) showed that colonizing pairs were able to dig into fine, damp soil, and did so in preference to wood, although they frequently dug in close to pieces of wood. In the present study, pairs isolated in petri dishes (containing 4% agar and a piece of decayed wood) readily dug into the agar and formed small cells partly in the agar and partly in the wood. These attempts to set colonies from primary pairs have been generally unsuccessful. All pairs started in dishes of agar and wood died within two months 146 Psyche [June without producing any eggs. On the other hand, pairs of replace- ment reproductives have appeared in large laboratory subcolonies of 1000-2000 individuals, and these have produced dozens of eggs which hatched successfully over the course of a year or more. Coloniz- ing pairs have been found a few hours after flights under pieces of dead wood in the study area. In the desert where the soil is hard and dry during much of the flight season, pairs may find points suitable for digging in next to, or beneath, dead wood. The sequence of post-flight behavior thus appears to bear certain similarities to that of many subterranean termites. Predators. — It has been assumed that most of the diurnal reptiles would find no opportunity for feeding on Paraneotermes. It was, therefore, a surprise when Asplund (’64) found 24 alates in one, and 10 in another, stomach of the tree lizard, Urosaurus ornatus. Such numbers suggest that the alates may have been taken at their emergence holes. Observations in the study area and surrounding desert have estab- lished that most of the common birds have ceased their feeding ac- tivities and final vocal chorus shortly before the daily flight; periods of Paraneotermes begin. Both Nighthawks and bats have been seen patrolling the area as early as five minutes before sunset and until no longer visible, approximately 25 minutes after sunset. Although no stomachs from any birds in the area have been examined, many stomachs of several species of bats have been carefully studied. Un- determined fragments of termite wings were found in the stomachs of three specimens of the Hoary Bat, Lasiurus cinereus , taken either in southern Arizona or New Mexico (Ross, ’64). Discussion. — In the preceding presentation a variety of prob- lems have arisen and possible approaches to their solutions have been suggested by the evidence at hand. A few additional questions of a more general nature are included here. During the long flight season of Paraneotermes there are many evenings when no flights are staged. In 1962 flights were recorded on only 29 out of 91 evenings between the inclusive dates July 9 — Sept. 7. On at least 15 of these evenings rain or moderate to strong winds occurred which are probably valid deterrents to flight. In this connection it should be worth determin- ing the temperature limits within which the alates are capable of sustained flight. For example, on 13 evenings, the combinations of temperature and relative humidity (at sunset plus 15 minutes) were outside the range determined for a total of 49 flights in the study area (fig. 6). Allowing for a few flights which may have been missed, this still leaves 34 evenings, or over one-third of the total, 1966] Nutting — Paraneotermes 147 when no flights took place. There are no other environmental factors which might obviously have prevented flights on these days. The most reasonable explanation which can be advanced at present — and which would also explain the long flight season characteristic of many termites — is a progressive production of alates. Evidence from the two subcolonies taken in 1961 does not support this idea, for large numbers of alates were produced within a short time, just prior to the flight season. It is still possible, however, that the pre- alates develop deep within the colony and migrate to the more super- ficial out posts to complete their development as the season progresses. Herfs (’51) found that Reticulitennes Lucifugus (Rossi) (maintained in 13 large groups under constant conditions) produced alates over a period varying from 0.25 to 3.5 months. During their flight season, and not always at the time of flight, one or more castes of the following genera in Arizona have been noticed at emergence (“observation”?) holes: Incisitermes , Zooter- mopsisj, Heterotermes , Amitermes and Gnathamitermes. This habit, which has probably been noted by many other observers, has led to the suspicion that termites may not be so much the “dwellers in darkness” as generally believed. It further suggests that many ter- mites may thus have considerable, and perhaps rather continuous, information on the photoperiod and other external environmental factors, either through actual openings to the outside or through the walls of superficial galleries in wood or soil. The behavior of alates, or of an entire colony, might then be readily imagined to be adjusted to the approximate frequency of one or more points in the daily photoperiod. Data on the flight behavior of Paraneotermes strongly suggest that the alates are initially stimulated by high temperatures, but that the daily flights may be finally triggered in response to a definite level of diminishing light intensity on otherwise favorable evenings. The flights of many termites and of several other insects have been shown to be closely correlated with dusk or to decreased light intensity ( e.g ., Myers, ’52; Bates, ’49). Field studies in progress on a variety of other termites in Arizona and Mexico are aimed at providing a basis for the comparative study of termite flight behavior. Thus far, limited experience has shown that it may be possible to resolve certain behavioral problems by the manipulation of captive colonies in the laboratory. It is anticipated that a balanced combination of field and laboratory studies may eventually permit the forecasting of both flight seasons and daily flight periods of the species within limited geographic areas. Summary. — Weather data have been collected and studied in 148 Psyche [June conjunction with field observations on 78 flights of the desert damp- wood termite, Paraneotermes simplicicornis , in southern Arizona. Its flight behavior is characterized by a large number of generally small flights occurring over a prolonged season and over a rather wide range of environmental conditions. Alates appear in the colonies about three weeks in advance of the flight season. As many as 31 flights have been observed in one season which may extend from late May into early September. Flights begin when weekly mean tem- peratures remain consistently above 23 °C. They are staged during evening twilight and begin at a mean temperature of 27.6°C. (range, 23.4 — 32) and a mean relative humidity of 60.8% (range, 36 — 100). Mean saturation deficit at the start of 29 flights in one season measured 12.8 mm. Hg and ranged between 2.1 and 20.7 mm. Rec- ords clearly indicate that this termite does not begin swarming, nor stage its daily flights, in response to rainfall. Since the summer rains occur in the form of sporadic thunderstorms, the mean daily evaporation rate remains high: 0.36 inches (0.10 to 0.51) for the flight days in 1962. The sequence of post-flight behavior is described and bears certain similarities to that of subterranean termites. References Cited Asplund, Kenneth K. 1964. Seasonal variation in the diet of Urosaurus ornatus in a riparian community. Herpetologica, 20:91-94. Bates, M. 1949. The Natural History of Mosquitoes. Macmillan Co., New York, pp. 54-55. CUTRIGHT, C. R. 1927. Notes on the computing of mean temperatures for biological use. Ann. Ent. Soc. Amer., 20:255-261. Harvey, P. A. 1934- The distribution and biology of the common dry-wood termite, Kalotermes minor. II. Life history of Kalotermes minor, p. 221. In Kofoid, C. A., et al. [eds.] Termites and Termite Control, 2nd ed., Univ. Calif. Press, Berkeley. Herfs, A. 1951. Der Schwarmflug von Reticulitermes lucifugus Rossi. Zeitschr. Angew. Ent., 3 3:69-77. Kalshoven, L. G. E. 1930. (Bionomics of Kalotermes tectonae Damm. as a base for its con- trol.) Mededeel. Inst, plantenziekt., Buitenzorg, 76:1-154. Kannowski, Paul B. 1959. The flight activities and colony-founding behavior of bog ants in southeastern Michigan. Insectes Sociaux, 6: 115-162. 1966] Nutting — Paraneotermes 149 Light, S. F. 1934. The southern and mountain dry-wood termites, Kalotermes hub- bardi and Kalotermes marginipennis, p. 268. In Kofoid, C. A., et al. [eds.] Termites and Termite Control, 2nd ed., Univ. Calif. Press, Berkeley. 1937. Contributions to the biology and taxonomy of Kalotermes (Para- neotermes) simplicicornis Banks (Isoptera). Univ. Calif. Publ. Ent., 6: 423-464. Light, S. F. and Frances M. Weesner. 1948. Biology of Arizona termites with emphasis on swarming. Pan- Pacific Ent., 24: p. 56. McCluskey, E. S. 1965. Circadian rhythms in male ants of five diverse species. Science, 150:1037-1039. Myers, K. 1952. Oviposition and mating behavior of the Queensland fruit-fly, Dacus (Strumeta) tryoni (Frogg.) and the Solanum fruit-fly, Dacus ( Strumeta ) cacuminatus (Hering). Australian J. Sci. Res., B, 5:264-281. Ross, Anthony. 1964. Ecological aspects of the food habits of some insectivorous bats. Ph.D. Thesis, Univ. Arizona, pp. 33, 35. Sellers, W- D., Ed. 1960. Arizona Climate. Univ. Ariz. Press, Tucson, v + 60 pp. + climatological summaries. Snyder, T. E. 1954. Order Isoptera, the termites of the United States and Canada. Nat. Pest. Control Assoc., New York, Tech. Bull., pp. 38, 40. 1961. Supplement to the annotated, subject-heading bibliography of termites, 1955 to 1960. Smithsonian Misc. Coll., 143: p. 27. 1964. Personal communication. Talbot, Mary. 1956. Flight activities of the ant Dolichoderus (Hypoclinea) mariae Forel. Psyche, 63:134-139. 1964. Nest structure and flights of the ant Formica obscurwentris Mayr. Animal Behaviour, 12:154-158. U. S. Department of Commerce. Weather Bureau. 1962. Climatological Data, Arizona, v. 66. 1963. Climatological Data, Arizona, v. 67. Weesner, Frances M. 1960. Evolution and biology of termites. Ann. Rev. Ent., 5: pp. 160-162. CHEMICAL RELEASERS OF SOCIAL BEHAVIOR. X. AN ATTINE TRAIL SUBSTANCE IN THE VENOM OF A NON-TRAIL LAYING MYRMICINE, DACETON ARMIGERUM (LATREILLE)1. By M. S. Blum2 and C. A. PORTOCARRERO2’ 3 Although the poison gland secretion may provide species in several myrmicine genera with a readily available source of odor trail phero- mone, this venom does not represent a highly species-specific secretion. It has been demonstrated recently that odor trails prepared from extracts of the poison gland are non-specific even among unrelated genera, although specificity may be absolute when these extracts are tested on species in the same genus (Blum and Ross, 1965). The lack of specificity of the poison gland extracts appears to be due to the fact that trace constituents, which are common to different venoms, are employed by different genera as odor trail releasers. Thus, al- though members of two genera may be employing different trail sub- stances, they will follow trails prepared from each other’s poison glands because their odor trail compounds are present in both venoms. Since some of the trace compounds in the venoms synthesized by un- related trail-laying myrmicines appear to be similar, it would not be surprizing if the venom of a non-trail laying species contained a trace constituent which was the same as the odor trail pheromone employed by trail-laying members of this subfamily. The purpose of this present paper is to report on the occurrence of just such a case. Daceton armigerum (Latreille) a primitive member of the tribe Dacetini, contains in its venom a powerful releaser of trail following for membeis of three attine genera as well as for an inqui- line cockroach which is associated with one of these genera. The biology of Daceton has been studied in detail by Wilson (1962). Daceton workers hunt singly and no evidence of trail laying or recruitment was observed either in the field or in the laboratory. The large eyes of Daceton workers appear to endow them with This work was supported in part from two grants from the University Council on Research of Louisiana State University department of Entomology, Louisiana State University, Baton Rouge, Louisiana. 3Present address: Department of Communications, Michigan State Uni- versity, East Lansing, Michigan. Manuscript received by the editor June 29, 1966. 150 1966] Blum and Portocarrero — Chemical Releasers 15 1 Table 1. Response of attine species and a species of Attaphila to artificial trails prepared from extracts of the poison gland of Daceton armigerum. Test species No. of workers tested No. of workers responding Daceton armigerum 100 O T rachymyrmex septentrionalis 100 90 A cromyrmex coronatus 100 87 Acromyrmex nr. coronatus 100 94 Sericomyrmex urichi 60 0 A tta texana 100 73 Atta cephalotes 100 82 A ttaphila sp. 30 28 exceptional vision and they are able to pursue their prey for con- siderable distances (Wilson, 1962). The two colonies of Daceton that Wilson studied were strictly aboreal and in no instance did he see workers move from the trees to the ground. However, in the single Daceton colony which we observed we noted three workers on the leaf litter adjacent to the base of their nest tree, making no apparent effort to forage on the leaves, and two other workers some distance from their colonial tree. Under the nest tree there were two small shrubs separated from the overhanging tree by at least three feet. A Daceton worker was rest- ing on each shrub. These workers only could have reached these shrubs by walking across the leaf litter under their nest tree or by falling from the tree onto the shrubs. These observations suggest that under certain conditions Daceton workers may not be totally arboreal. METHODS AND MATERIALS A nest of Daceton was located in a large tree in a nursery at Buenos Aires, 25 kilometers south of Pucallpa, Peru, near the Rio Ucayalli. The colony consisted of three fragments in two branches of the tree. Approximately 600 workers were collected from the nest and were transported to Pucallpa for odor trail studies. Their poison glands were extracted in absolute ethanol. These ethanolic extracts were employed for preparing circular trails (Moser and Blum, 1963) and Daceton and Atta cephalotes (L.)4 workers were tested on these trails. The living Daceton workers were transferred “Collected at Pucallpa, Peru. 152 Psyche [June to Tingo Maria, where further artificial trail testing was performed using chloroform solutions of the Daceton poison glands. The fol- lowing species were also tested on the trails: Acromyrmex coronatus (F.)5, Acromyrmex nr. coronatus (F. )5, and a cockroach, Attaphila sp.5, which was found in the fungus garden of A. coronatus. The Daceton workers then were frozen and packed in dry ice for air transport to the United States. Workers of Trachymyrmex septentrionalis (McCook)6, Sericomyrmex urichi Forel7, and Atta texana (Buckley)8 were tested on artificial trails prepared from methylene chloride solutions of poison glands dissected from the frozen Daceton workers. RESULTS Although Daceton workers do not follow artificial trails prepared from their own poison gland secretion, these trails release strong trail following in attine members of the genera Trachymyrmex , Acromyrmex , and Atta (Table i). On the other hand, workers of Sericomyrmex urichi did not follow trails prepared from the venom of Daceton. The concentration of the attine odor trail pheromone in the venom of Daceton appears to be the same as it is in the venoms of the attines. Extracts which were prepared from Daceton and Atta poison glands of equivalent size exhibited about the same odor trail potencies after serial dilution. It is quite likely that the Attaphila are responding to the same compound in the venom of Daceton as are the attines. Moser ( 1965) has shown that Attaphila fungicola Wheeler will follow artificial trails prepared from the venom of both A. texana and T. septentrio- nalis. Although Moser reported that A. texana workers were more sensitive to the odor trail pheromone than the cockroaches, we noted that the Attaphila held to the artificial trails much more tenaciously than the ant workers in any of the genera. Since Sericomyrmex urichi did not follow artificial trails prepared from the venom of Daceton, we wished to determine whether or not this attine would respond to artificial trails prepared from the poison glands of other attines and vice versa. Assuming that the venom of Daceton contains a compound which is similar to those employed by several of the attine genera as trail substances, then Collected at Tingo Maria, Peru. Collected at Baton Rouge, Louisiana. ’Collected at Trinidad by Prof. Neal A. Weber, Department of Biology, Swarthmore College, Swarthmore, Pa. Collected at Pineville, Louisiana. 1966] Blum and Portocarrero — Chemical Releasers 153 Table 2. Numbers of Sericomyrmex , Trachymyrmex , and Atta workers responding to the poison gland secretion in the artificial test. Number of replications in parentheses9. Test species Source species Sericomyrmex urichi Trachymyrmex septentrionalis A tta texana Sericomyrmex urichi Trachymyrmex 78(8) 0 (8) 40(10) septentrionalis 10(8) 90(8) 91 ( 10) Atta texana 0(8) 88(8) 96(10) the failure of Sericomyrmex to follow Daceton trails would indicate that its trail pheromone was different from those of the other attines. Circular trails were prepared from extracts of the poison glands of S. urichi, T. septentrionalis and A. t ex ana, and the response of workers of each species to the trails was determined. The results, presented in Table 2, demonstrate that S. urichi does not follow the odor trails of A. texana or T. septentrionalis. Similarly, the trail substance in the poison gland secretion of Sericomyrmex releases virtually no trail following in workers of T. septentrionalis and is only slightly active when tested with workers of A. texana (Table 2). DISCUSSION Although Daceton armigerum does not lay odor trails, its venom contains a substance which is either similar or identical to the trail pheromones employed by attines in the genera Trachymyrmex, Acromyrmex, and Atta. In all probability this substance is a trace constituent of the poison gland secretion produced by Daceton. The venom of Daceton is rich in proteins which solidify when the poison vesicle is ruptured in the air. An attine odor trail pheromone can be readily extracted from the solidified venom without a measur- able weight loss occurring and it is probable that a trace component is being removed during the extraction process. Similarly, the odor trail pheromones can be extracted from the solidified attine venoms without causing any detectable weight loss in the venomous residues. The attine venoms, like that of Daceton , are rich in proteins but in addition, the poison gland secretions of the attines contain large series of free amino acids which cannot be detected in the venom of Daceton. More evidence that the venom of Daceton contains an attine odor trad nheromone is derived from the fact that the Attaphila follow 9Ten workers per replicate. 154 Psyche [June the artificial Daceton trails. Moser (1965) has demonstrated that another cockroach species, Attaphila fungicola, follows artificial at- tine trails, and Echolls (1965) has seen this cockroach on trails of A. texana in the early morning (1 a.m.) Obviously these blattids can maintain a close association with their A tta hosts because of their ability to follow the odor trails of the ants. Thus, since members of the genus Attaphila appear to be following the same trail substance as the attines employ, these cockroaches would be expected to follow artificial Daceton trails if their venom contained an attine trail pheromone. It should be added that the Attaphila also follow artificial trails of both Acromyrmex species, as they would be expected to do. The evidence concerning Sericomyrmex is also circumstantial but is equally persuasive. Previously it had been shown that there was no specificity in the trail substances produced by four attine genera which encompassed the broad phylogenetic spectrum of the tribe Attini (Blum et al., 1964). Since these attines follow each other’s odor trails, then they would be expected to follow the artificial Daceton trails if these trails contained a substance which was chemi- cally similar to their apparently common odor trail pheromones. On the other hand, if an attine produced a trail substance which was different from the one being employed by these several attine genera, then this attine would respond neither to the artificial trails of these other attine genera nor to the Daceton trail. Sericomyrmex urichi is the first member of an attine genus which has been shown not to respond to the artificial trails prepared from the poison gland secre- tions of other attine genera. Weber (1966) considers Sericomyrmex a somewhat aberrant member of the higest genera and it is certainly distinguished from at least two of the other higher attine genera by this apparent employment of a different trace constituent as an odor trail pheromone. The other gland associated with the sting, Dufour’s gland, also has been shown to be an unexpected source of a myrmicine trail phero- mone. Wilson and Pavan (1959) reported that the Dufour’s gland secretion of the dolichoderine Monads hispinosa (Olivier) contained a powerful trail substance for the fire ant Solenopsis saevissima (Fr. Smith) which produces its trail substance in Dufour’s gland. How- ever, M. hispinosa synthesizes its own trail pheromone in Pavan’s gland, a special organ which lies on the sixth abdominal sternite. Thus the function of the Dufour’s gland secretion in M. hispinosa is completely unknown but it is obviously not employed for laying odor trails. It seems eveident that M . hispinosa , like DacetonJ pro- 1966] Blum and Portocarrero — Chemical Releasers 155 duces some natural products in its sting-associated glands which are also produced by members of the Myrmicinae. It is not unlikely that the natural products’ chemistry of the glands associated with the ant sting will continue to be characterized by the presence of common products, some of which are employed as trail substances. Under these circumstances we may anticipate that the Formicidae will continue to be a rich source of unsuspected trail substances whose specificities will be at best, unpredictable. SUMMARY Artificial trails prepared from the poison gland of Daceton armi- gerum (Latreille), a myrmicine which does not lay odor trails, cause trail following in attine species in the genera Trachymyrmex , Acro- myrmex , and Atta. An inquiline cockroach, Attaphila sp., also fol- lows trails prepared from the venom of Daceton. The poison gland secretion of Daceton apparently contains a trace constituent which is similar or identical to the odor trail pheromone used by these attine genera. An attine species in the genus Sericomyrmex responds neither to artificial Daceton trails nor the odor trail pheromones of Atta or Trachymyrmex. References Blum, M. S., J. C. Moser, and A. D. Cordero. 1964. Chemical releasers of social behavior. II. Source and specificity of the odor trail substances in four attine genera. (Hymenoptera : Formicidae). Psyche 71: 1-7. Blum, M. S. and G. N. Ross. 1965. Chemical releasers of social behaviour. V. Source, specificity and properties of the odour trail pheromone of T etramorium guineense (F.). Formicidae: Myrmicinae). J. Insect Physiol. 11: 857-868. Echolls, W. H. 1965. Private communication. Moser, J. C. 1964. Inquiline roach responds to trail-marking substance of leaf-cut- ting ants. Science 143: 1048-1049. Moser, J. C. and M. S. Blum. 1963. Source and potency of the trail marking substance of the Texas leaf-cutting ant. Science 140: 1228. Weber, N. A. 1966. Private communication. Wilson, E. O. 1962. Behavior of Daceton armigerum (Latreille) with a classification of self-grooming movements in ants. Bull. Mus. Comp. Zool. 127: 403-421. Wilson, E. O. and M. Pavan. 1959. Glandular sources and specificity of some chemical releasers of social behavior in dolichoderine ants. Psyche 66: 70-76. CAMBRIDGE ENTOMOLOGICAL CLUB A regular meeting of the Club is held on the second Tuesday of each month October through May at 7:30 p m. in Room B-455, Biological Laboratories, Divinity Ave., Cambridge. Entomologists visiting the vicinity are cordially invited to attend. The illustration on the front cover of this issue of Psyche is a reproduction of a drawing of a female bethylid wasp, Pseudiso- brachium terresi Mann, from Haiti (Psyche, vol. 22, p. 165, 1915). BACK VOLUMES OF PSYCHE The Johnson Reprint Corporation, in Fifth Avenue, New York 3, N. Y., has been designated the exclusive agents for Psyche, volumes 1 through 62. Requests for information and orders for such volumes should be sent directly to the Johnson Reprint Corporation. Copies of issues in volumes 63-72 are obtainable from the editorial offices of Ps)^che. Volumes 63-72 are $5.00 each. F. M. Carpenter Editorial Office, Psyche, 16 Divinity Avenue, Cambridge, Mass., 02138. FOR SALE Classification of Insects, by C. T. Brues, A. L. Melander and F. M. Carpenter. Published in March, 1954, as volume 108 of the Bulletin of the Museum of Comparative Zoology, with 917 pages and 1219 figures. It consists of keys to the living and extinct families of insects, and to the living families of other terrestrial arthropods; and includes 270 pages of bibliographic references and an index of 76 pages. Price $9.00 (cloth bound and postpaid). Send orders to Museum of Comparative Zoology, Harvard College, Cambridge, Mass. 02138. PSYCHE A JOURNAL OF ENTOMOLOGY Vol. 73 September, 1966 No. 3 CONTENTS Three New Species of Accola (Araneae, Dipluridae) from Costa Rica and Trinidad, W. I. Arthur M. Chickering 157 Distribution and Biology of the Primitive Dry-Wood Termite Ptero- termes occidentis (Walker) (Kalotermitidae) . W. L. Nutting 165 Postembryological Development of Spiderlings from Two Peruvian Latrodectus Populations. John D. McCrone and Herbert W. Levi .... 180 The Rediscovery of Camponotus (Myrmaphaenus) yogi Wheeler (Hymenoptera : Formicidae). William S. Creighton and Roy R. Snelling 187 Two New Genera of South American Cockroaches Superficially Re- sembling Loboptera , with Notes on Bionomics (Dictyoptera, Blattaria, Blattellidae) . Ashley B. Gurney and Louis M. Roth 196 New Species of Palpimanidae (Araneae) from the West Indies. A rt/iur M. Chiekering 208 A Review of Halocoryza Alluaud, with Notes on its Relationship to Schizogenius Putzeys (Coleoptera: Carabidae). Donald R. W hitehead 217 CAMBRIDGE ENTOMOLOGICAL CLUB Officers for 1966-67 President J. Reiskind, Harvard University Vice-President C. C. Porter, Harvard University Secretary F. Coyle, Harvard University Treasurer ,.F. M. Carpenter, Harvard University Executive Committee W. L. Brown, Jr.} Cornell University K. W. Cooper, Dartmouth College EDITORIAL BOARD OF PSYCHE F. M. Carpenter (Editor), Professor of Entomology , and Alexander Agassiz Professor of Zoology , Harvard University P. J. Darlington, Jr., Alexander Agassiz Professor of Zoology , Harvard University W. L. Brown, Jr., Associate Professor of Entomology , Cornell University; Associate in Entomology , Museum of Comparative Zoology E. 0. Wilson, Professor of Zoology , Harvard University H. W. Levi, Associate Curator of Arachnology , Museum of Com- parative Zoology H. E. Evans, Curator of Insects , Museum of Comparative Zoology PSYCHE is published quarterly by the Cambridge Entomological Club, the issues appearing in March, June, September and December. Subscription price, per year, payable in advance: $4.50 to Club members, $5.00 to all other subscribers. Single copies, $1.25. Checks and remittances should be addressed to Treasurer, Cambridge Ento- mological Club, 16 Divinity Avenue, Cambridge, Mass. Orders for back volumes, missing numbers, notices of change of address, etc., should be sent to the Editorial Office of Psyche, Biological Laboratories, Har- vard University, Cambridge, Mass. IMPORTANT NOTICE TO CONTRIBUTORS Manuscripts intended for publication should be addressed to Professor F. M. Carpenter, Biological Laboratories, Harvard University, Cambridge, Mass. Authors contributing articles over 4 printed pages in length may be required to bear a part of the extra expense, for additional pages. This expense will be that of typesetting only, which is about $10.00 per page. The actual cost of preparing cuts for all illustrations must be borne by contributors: the cost for full page plates from line drawings is ordinarily $12.00 each, and the full page half-tones, $18.00 each; smaller sizes in proportion. AUTHORS SEPARATES Reprints of articles may be secured by authors, if they are ordered at. the time proofs are received for corrections. A statement of their cost will be furnished by the Editor on application. The June, 1966 Psyche (Vol. 73, no. 2) was mailed October 21, 1966. The Lexington Press. Inc., Lexington, Massachusetts PSYCHE Vol. 73 September, 1966 No. 3 THREE NEW SPECIES OF ACCOLA (ARANEAE, DIPLURIDAE) FROM COSTA RICA AND TRINIDAD, W. I * By Arthur M. Chickering Museum of Comparative Zoology The genus Accola Simon was established in 1889 on the basis of an immature specimen of A. lucifuga Simon from Venezuela. During the next thirty-six years eight additional species were recognized. Three of these were from Luzon and New Guinea and all were described from females alone. The remaining five were from St. Vincent, W. I., Panama, Chile and Venezuela. Two of these, one from Venezuela and one from Panama, were described from imma- ture individuals. The remaining three species were all described from females alone. In 1945 I was able to describe the male of A. spinosa Petrunkevitch together with mature females. In 1964 I was able to describe both sexes of A. petrunkevitchi from Puerto Rico and both sexes of A. lewisi from Jamaica, W. I. During the past few months I have been able to recognize a new species of this genus taken in a recent collection in Costa Rica and two new species in a collection made in Trinidad, W. I. in April, 1964. These last- mentioned three species are described in the following pages of this brief paper. The present status of the Genus Accola Simon, 1889 may, therefore, be stated as follows: fourteen species are now recog- nized; six of these are known from both sexes; two are known only from immature specimens; the remaining six species are known only from females. I think it safe to predict that numerous additional species will become known as soon as careful collecting is carried out in South America and the West Indies. The greater portion of my collection of this genus has come from weed and hay debris with leaf debris considerably less productive. All specimens relating to the new species described in this paper will be deposited in the Museum of Comparative Zoology at Harvard University. In my study of the genus Accola I have noted irregu- *Manuscript received by the editor February 24, 1966. 157 158 Psyche [September larities in the outline of certain of the eyes; there seem to be differ- ences in the relative positions of certain of the eyes in some individuals and, apparently, noticeable differences in size of corresponding eyes in certain individuals. When eyes are oval in outline, as they usually appear, the long axis is used in measurements. The collections from Costa Rica and Trinidad, W. I. from which the three new species of A ccola were obtained were made possible by Grant GB-1801 from the National Science Foundation. My gratitude and appreciation are also again expressed for the privilege of working in the Museum of Comparative Zoology at repeated intervals over a period of many years. Publication and library privileges together with continued encouragement from directors and staff members have been indispensable for the continuation of my studies. Genus A ccola Simon, 1889 Accola downeyi sp. nov. Figures 1-5 The species is named after Dr. John C. Downey, Professor of Zoology, Southern Illinois University, Carbondale, Illinois. Male holotype. Total length from clypeus to posterior end of abdomen 3.9 mm; from anterior border of porrect chelicerae to pos- terior end of abdomen 4.29 mm. Carapace 1.67 mm long; 1.32 mm wide opposite second coxae where it is widest; .44 mm tall; dorsal striae moderately well developed ; dorsal median fovea a well defined pit opposite third coxae; stiff bristles along broad posterior border as usual in the genus; with a recurved row of stiff bristles shortly behind the median fovea; with numerous hairs and stiff bristles else- where. Eyes: six only; essentially as in A. lewisi Chickering, 1964 from Jamaica, W. I.; all compactly grouped on a low tubercle (Fig. 1); viewed from above, posterior row strongly recurved as usual; all eyes white; ratio of eyes ALE : PME : PLE = 11 : 5 : 11. ALE separated from one another by a little less than one fourth of their diameter; PLE separated from one another behind by slightly less than their diameter; PME separated from one another by about 3/10 of their diameter and from PLE by about one fifth of their diameter. Chitinized clypeus almost non-existent ; membranous region below chitinized portion quite conspicuous. Chelicerae: paraxial, parallel, porrect as usual in the genus; fairly robust; with a well developed coat of stiff bristles; fang long, slender, evenly curved; promargin of fang groove with a row of ten teeth; a cluster of very minute teeth, irregularly arranged along the obscure retromargin opposite the last four promarginal teeth (teeth observed on a para- 1966] Chickering — Accola 159 Figs. 1-5. Accola downeyi sp. nov. Fig. 1. Eyes of male as seen from above. Figs. 2-3. Spines and associated parts at articulation of first tibia and metatarsus of male holotype ; prolateral and ventral views, respectively. Figs. 4-5. Left palpal tibia and tarsus of male holotype; prolateral and retrolateral views, respectively. type to avoid injury to fragile holotype) ; retromargin with a well developed scopula of long, slender hairs or bristles. Lip and sternum : essentially typical of the genus. Legs: 4123; tibial index of first leg 10, of fourth leg 12. Spines: all legs bear numerous spines essen- tially as described for A . petrunkevitchi Chickering, 1964 from Puerto Rico; special spines or cusps at articulation of first tibiae and metatarsi shown in Figures 2-3 ; numerous trichobothria observed on tibiae, metatarsi and tarsi. Palp : essential features shown in Figures 4-5 ; coxa with a rudimentary maxillary lobe. Abdomen and color in alcohol: both essentially as described for A. spinosa Petrunke- vitch (1945) ; anterior spinnerets very fragile. Female paratype. Total length from clypeus to posterior end of abdomen 5.33 mm; from distal end of porrect chelicerae to posterior end of abdomen 5.92 mm. Carapace 2.05 mm long; 1.52 mm wide opposite second coxae where it is widest; .61 mm tall; otherwise i6o Psyche [September essentially as in male. Eyes: six; essentially as in male. Chelicerae: in general as in male holotype; promargin of fang groove with eleven small but clearly defined teeth ; the cluster of very minute teeth along obscure retromargin has about 19 or 20 in the group. Lip and sternum: essentially as in male holotype. Legs: 4123; tibial index of first and fourth legs 12; spines essentially as in male; trichobothria observed on tarsi, metatarsi, tibiae and patellae as well as on palpal segments; palpal claw finely dentate. Abdomen: typical of females of the genus; second pair of lungs clearly delineated. Epigynum : with a well defined plate ; posterior margin procurved ; typical of females of the genus. Type locality. The male holotype and the described female para- type are from Turrialba, Costa Rica, July 25 to August 15, 1965. Four male paratypes and about three dozen females and immature specimens were taken in the same general locality and during the same period. In 1945 when I described the mature male and mature female of A. spinosa Petrunkevitch I also had a few mature females from Boquete, Panama with only six eyes. These six-eyed specimens were then regarded as deviates of A. spinosa Petrunkevitch as ex- plained in a footnote, page 1. Since that time numerous females and three mature males from Boquete and El Volcan, Panama have been collected. These are all six-eyed and agree well with my specimens of Accola downeyi sp. nov. For this reason they are, therefore, transferred to this species. Accola simla sp. nov. Figures 6-8 The name of the species is a noun used in apposition after the type locality. Male holotype. Total length from clypeus to posterior end of abdomen 3.38 mm; from anterior border of porrect chelicerae to posterior end of abdomen 3.64 mm. Carapace 1.52 mm long; 1.1 mm wide opposite second coxae where it is widest ; 0.4 mm tall ; other features typical of males of the genus. Eyes: eight essentially as in A. petrunkevitchi Chickering from Puerto Rico; all arranged on a low tubercle (Fig. 6) ; viewed from above, posterior row strongly recurved; AME dark and difficult to distinguish; all others white; ratio of eyes AME : ALE : PME : PLE = 2 : 1 1 : 6.5 : 10.5 ; AME separated from one another by about their diameter; only slightly separated from ALE ; ALE separated from one another in front by about one sixth of their diameter; PME separated from one another by about one sixth of their diameter; PME separated 1966] C bickering — A ccola 161 Figs. 6-8. A ccola simla sp. nov. Fig. 6. Eyes of male holotype as seen from above. Fig. 7. Spines and associated parts at articulation of first left tibia and metatarsus of male holotype; prolateral view. Fig. 8. Left palpal tibia and tarsus of male holotype; retrolateral view. Figs. 9-11. A ccola barona sp. nov. Fig. 9. Eyes of male holotype as seen from above. Figs. 10-11. Left palpal tibia and tarsus of male holotype; prolateral and retrolateral views, respectively. Figs. 12-13. A ccola spinosa Pet. Fig. 12. Spines and associated parts at articulation of first left tibia and metatarsus of male ; prolateral view. Fig. 13. Left palpal tibia and tarsus; retrolateral view. Psyche [September 162 from PLE by about one fourth of their diameter; PLE separated from one another behind by about four fifths of their diameter. Chitinized height of clypeus very low and irregular; membranous portion much higher. Chelicerae paraxial, parallel, porrect as usual in the genus; fairly robust; with a well developed supply of stiff bristles ; fang long, slender, evenly curved ; promargin of fang groove with a row of ten small teeth; along the obscure retromargin, as seen on a paratj^pe, is a row of about 7-8 very minute denticles; retromargin with a well developed scopula of long, slender hair or bristles. Lip and sternum : essentially typical of the genus. Legs : 412 = 3; tibial index of first and fourth legs 11. Spines: all legs bear numerous spines essentially typical of the genus; special spines at articulation of first tibiae and metatarsi shown in Figure 7 ; the most ventral of the tibial spines is probably divided but the two parts are so closely apposed that they appear as one. Palp : essential features shown in Figure 8 ; bulb terminates in two spines. Abdo- men: typical of the genus; with a heavy coat of dark hair. Color in alcohol : as usual in the genus ; a unicolorous yellowish. Female paratype. Total length of female paratype, selected for description, from clypeus to posterior end of abdomen 4.03 mm; from anterior border of porrect chelicerae to posterior end of abdomen 4.42 mm. Carapace 1.78 mm long; 1.32 mm wide; .45 mm tall; otherwise essentially as in male. Eyes: essentially as in male. Cheli- cerae: promargin of fang groove with nine small teeth; retromargin with about eleven minute denticles somewhat irregularly arranged ; otherwise essentially as in male. Lip and sternum : essentially typical of the genus. Legs: 4132; tibial index of first and fourth legs 12; spines on legs essentially as usual in the genus. Abdomen : typical of females of the genus ; with the usual heavy coat of hairs. Epigynum : with the usual swollen plate, typical of females of the genus. Color in alcohol : as in male. Type locality. Male holotype and female paratype selected for description taken at Simla, Arima Valley, Trinidad, W. I., April 15th, and nth, 1964, respectively. Fourteen male paratypes and several dozens of what are believed to be females and immature specimens belonging to this species were taken in same general locality from March 31st to April 26th, 1964. Accola barona sp. nov. Figures 9-1 1 The name of the species is an arbitrary combination of letters. Male holotype. Total length from anterior border of ALE to 1966] C bickering — A ccola 163 posterior end of abdomen 3.71 ; from anterior border of porrect cheli- cerea to posterior end of abdomen 3.9 mm. Carapace: 1.69 mm long; 1 .2 1 mm wide opposite second coxae where it is widest; .34 mm tall; slightly the highest just behind median thoracic pit; other features typical of the genus. Eyes: six only; no traces of the AME observed; viewed from above, posterior row quite strongly recurved; all six eyes white; ratio of eyes ALE : PME : PLE = 10.5 : 5 : 9; ALE separated from one another by about one fifth of their long axis; PME separated from one another by about 3/10 of their long axis and separated from PLE by somewhat less than this; PLE separated from one another behind by a little less than their long axis. Height of clypeus equal to about 1/7 of the long axis of ALE. Chelicerae: essentially as described for A. simla sp. nov. ; teeth and denticles along fang groove seem to be much as in A. simla sp. nov. but they have not been closely observed because of the fragility of the holotype and absence of male paratypes. Lip and sternum : essentially typical of the genus. Legs: 4i?23; tibial index of fourth leg 12. Spines: ordinary spines on legs appear typical of the genus; first legs missing and, hence, no record possible in respect to the special spines at articulation of first tibiae and metatarsi. Palp : essential features shown in Figures 10-11. Abdomen: typical of males of the genus. Color in alcohol : as usual, unicolorous yellowish throughout. Female paratype. Total length of female paratype selected for description 4.23 mm from clypeus to posterior end of abdomen; from anterior border of porrect chelicerae to posterior end of abdomen 4.68 mm. Carapace 1.63 mm long; 1.17 mm wide opposite second coxae where it is widest; about .46 mm tall; other features as usual in females of the genus. Eyes: six as in male holotype; ratio of eyes ALE : PME : PLE = 8:4: 6.5 ; other features essentially as in male holotype. Chelicerae: general features as usual in the genus; the second female shows promargin of fang groove with about twelve teeth, those in nth and 12th places the largest; the retromargin has a well developed scopula and a row of a dozen or more very minute denticles irregularly arranged. Legs: 412 = 3; tibial index of first leg 11, of fourth leg 12; with many spines as usual; tarsal claws finely and multidentate. Abdomen: as usual in the genus. Epigynum: with the usual somewhat thickened plate and the procurved posterior border. Color in alcohol : unicolorous yellowish as usual. Type locality. Male holotype and described female paratype to- gether with one additional somewhat damaged female all taken from 164 Psyche [September grass and weed debris in the immediate vicinity of the William Beebe Tropical Research Station, Simla, Arima Valley, Trinidad, W. I., April 23, 1964. Accola spinosa Petrunkevitch Figures 12-13 As pointed out in the last paragraph of my description of A. downeyi sp. nov., all six-eyed specimens now in my collection of the genus Accola from Boquete and El Volcan, Panama are transferred to A. downeyi sp. nov. Since the publication of my paper on this species in 1945 I have collected a large number of both sexes in the Panama Canal Zone and, especially, on Barro Colorado Island. Females as expected, far out- number males. In order to make the distinction clearer between this species and other eight-eyed forms such as A. petrunkevitchi Chicker- ing from Puerto Rico and A. simla sp. nov. from Trinidad I am including Figures 12-13. Special attention is directed to Figure 12 which shows the division of the most ventral of the spines at the distal end of the first tibia. This division was noted in my 1945 description. There is some variation in the appearance of these spines among males of this species; apparently in some specimens the two spines overlap to such an extent as to obscure one of them. Disarticu- lation of the first tibia and metatarsus reveals the division more clearly. Selected Bibliography Bonnet, P. 1955. Bibliographia Araneorum. Toulouse, Vol. 2 (2). Chickering, Arthur M. 1945. Hypotypes of Accola spinosa Petrunkevitch (Dipluridae) from Panama. Trans. Connecticut Acad. Arts and Sciences, 36: 159-167. 1964. Two New Species of the Genus Accola (Araneae, Dipluridae). Psyche, 71 (4): 174-180. Petrunkevitch, A. 1925. Arachnida from Panama. Trans. Connecticut Acad. Arts and Sciences, 27: 51-248. Roewer, C. Fr. 1942. Katalog der Araneae. 1 : 1-1040. Simon, Eugene 1892- Histoire naturelle des Araignees. Deuxieme Edition. 1903. 2. Librarie Encyclopedique de Roret, Paris. DISTRIBUTION AND BIOLOGY OF THE PRIMITIVE DRY-WOOD TERMITE PTEROTERMES OCCIDENT IS (WALKER) ( KALOTERMITIDAE ) 1 By W. L. Nutting Department of Entomology, University of Arizona, Tucson This summary of recent studies on Pterotennes is part of an irreg- ular series of contributions to the biology and ecology of the relatively rich termite fauna of southwestern North America. The fact that such a large and primitive termite has been known from about a dozen rather poorly documented collections is indicative of the need for such knowledge in this rapidly developing region. Indeed, very little additional information on this species has beeen published since the important summary by Banks and Snyder in 1920. Most of the 25 new records have been accumulated by members, students and friends of the Department of Entomology at the University of Arizona, largely within the last ten years. The most extensive addi- tions were made during a trip through Baja California in the late summer of 1959 by Floyd G. Werner and Keith W. Radford. Their almost nightly collections of this termite in a light trap indicate that their schedule must have coincided closely with the peak of the flight season. Although most of the records consist of alates taken in light traps, in seven instances one or more colonies have been found which have added much new biological information. Distribution. This species has long been included in the large, cosmopolitan genus, Kalotermes. On the basis of careful morpho- logical and taxonomic considerations, Krishna (1961) removed it to a previous place in the genus Pterotermes. Because of its probable key position in arising directly from ancestral kalotermitids and its apparent rarity, all readily available records have been brought to- gether in Table 1 and plotted in Figure 1. Those in Mexico, particularly from the generally unfamiliar geography of Baja Cali- fornia, have been numbered to facilitate their location. The distribution is thus found to conform closely to the Sonoran Desert as it has been delimited on the basis of the vegetation. A few marginal localities in southeastern Arizona and southern Baja Cali- fornia might be considered as local extensions of the region. A brief characterization of this, the richest of the four areas comprising the Arizona Agricultural Experiment Station Journal Article No. 1149. Manuscript received by the editor June 22, 1966 66 Psyche [September North American Desert, is pertinent to the discussions which follow. The physical features, vegetation and flora of this desert have been well reviewed by Shreve (1951). The Sonoran Desert lies in a region of plains, hills and mountains, and extends in elevation from approximately 3500 feet in eastern Arizona and northern Sonora to sea level on the coasts of Sonora and Baja California, and to below sea level in the vicinity of the Salton Sea in southern California. The Colorado River is the only permanently flowing stream in the entire region. A fairly uniform, continental type of climate prevails, in which the temperature varies from some of the highest records for North America to occasionally heavy frosts in the northern and eastern sections. Annual rainfall generally increases in amount from west to east, with less than 5 inches falling in the lower Colorado Valley and much of Baja California to slightly more than 15 inches in eastern Arizona and Sonora. The seasonal distribution of rainfall also varies importantly, with the extreme western areas receiving virtually all of their moisture during the winter from December to March. Going eastward the pattern shifts so that the eastern borders of the desert receive their major rainfall during a well-defined summer season from July to September. In Arizona where more precise information is available, Ptero- termes has been found from approximately 4000 feet on the lower slopes of a few mountain ranges, across the bajadas and onto the adjacent valley floors. It is apparently restricted to the more open foothills and bajadas rather than to the recesses of deep canyons. Here it is found in the larger woody plants of the desert scrub or spinose desert plant communities (Figs. 2 and 3). It is hardly necessary to point out that nearly all the collections from Baja Cali- fornia have been made in the lower, less rugged areas which are more readily accessible, either by sea or by the few passable roads over the peninsula. Emerson (1955) has mentioned that temperature and moisture are the major physical factors limiting termite dispersal, largely by their effects in determining vegetation types. Pterotermes is a primi- tive, monotypic genus, apparently endemic to the hot, dry Sonoran Desert. Although almost pure speculation, it may be of interest to suggest that Pterotermes is so restricted not so much by high tempera- ture and low moisture as by the higher rainfall of the surrounding regions. Approximately 12 vigorous and healthy colonies have been personally examined in extremely desiccated wood. A single large colony was completely removed from a dead palo verde tree in February, 1966, after a series of unusually heavy rains. This 1966] Nutting — Dry-wood termite 167 colony contained a high percentage of individuals which had ap- parently succumbed in large terminal chambers to drowning and attack by fungi, bacteria, or both. There is also limited evidence that Pterotermes may occupy a very narrow niche in this region where it only rarely comes into contact with other dry-wood termites. Two colonies have been found in wood previously or concurrently attacked by Marginitermes hubbardi (Banks). Marginitermes is itself a rather primitive, mono- typic genus which is endemic to the southwestern United States and western Mexico. However, from the above evidence and wider field experience with the latter, it is my impression that Pterotermes is able to occupy even hotter and dryer situations than Marginitermes. Comparative studies on water loss and cuticular structure (Collins and Richards, 1966) should provide a basis for explaining the adapta- tions of various species of termites to their particular environments. Indeed, Collins (1966, in litt.) has found that older and larger nymphs of Pterotermes have a comparatively low rate of water loss. She feels that the cuticular cement layer, which is particularly well developed in these forms, may be largely responsible for the ability of this species to conserve water and, hence, to occupy the severely dry environment of the Sonoran Desert. As a result of very inadequate collecting there are a few broad discontinuities in the distribution. Further field work should be conducted to determine whether or not this termite has an essentially uninterrupted distribution generally within the Sonoran Desert. It may actually be absent over many of the intermont plains which are dominated by smaller and lower plants such as Larrea and Franseria. However, it is a strong flier and may well be able to cross narrowed valleys between adjacent mountain ranges — if not in one generation, then in several by way of relatively isolated trees and cacti, essential- ly as in island-hopping. After all, it does occupy a region where hosts are of necessity very widely spaced. Host trees and nesting site. According to the meager records, Pterotermes has thus far been recorded nesting in the dead wood of only three plants: in the ribbed, woody skeletons of the giant or saguaro cactus, Cereus giganteus Engelm. (Fig. 2) ; one species of the green-stemmed palo verde tree, Cercidium floridum Benth.; and in the dead, flowering stalk of the non-arborescent Spanish bayonet, Yucca Whipplei Torr. In the foothill areas of the Santa Rita Range Reserve and south of Oracle Jet., several colonies were discovered in dead palo verdes of this species from which the bark had long since slipped off. A TABLE 1. Distributional and flight data for Pterotermcs occidcntis. Localities in Mexico are numbered for reference on map in figure 1. 68 Psyche [September fa O h u w X fa o u * o hC fa « £ FQ fa> o be CM 3 be ~ 3 >i 3 ■Sz o he bC cm c c 3 w> bj ^ m 3 ^ . FQ fa fa fa *T! 3 3 -O Z Z >k C C/D £ £ o «i C CM « ca - £ s n- £ o « c/d c fa -M he _ _ - C TD -3 £ *5 3 3 .52 £2 03 03 3 3 o*- tw cn J*! 33 3 3 ,• k n fl 0 t> FQ FQ fa fa FQ 'y (U g K £ £fq> O \r. CO O O U X >, 3 3 ° Ow O O U U o o U U cb £ a a 3 3 U U CO 1 T — ( ON • ■ 2 T T A -+■ ig t S - ci 7 *7 . , ^ ;N^^NH+NNN^ ^ 2 o g> So n o o O o CM co a> 3h 3 C/D £ 5 a* C-s o O o o o o O o O o O NO O o NO OO OO O OO OO co CO CM to CO CM CO to £ * 3 X (10 m talina * e ® « „ n N Os + to t-h i-h CM 3 CT1 O 3D 03 FQ W c a, fa 3 3 he a ±: o O .|£ i 1 1 £ «r 3 r-5 in lii Si o' a*- o S o o o .. £ >-. 3D 3D O 3 O o ca cs - ••-> U fa fa < < < 1966] Nutting — Dry-wood termite 169 z h w x X O < x o z Z ^ ►r a X > S c o O O 'o U U T3 j-i O T3 OS £ ^ 1_'X o c * £ <■ £ fa, fa c O CM o U fa fa fa c c o oo O -a -a -a h ^ ^2 -.2 T3 T) T3 rt rt os Pi Pi Pi X X X fa fa i-T to aj cu , >* C G o o o o U U -G ^ bl) fa fa -a o "O 03 id fa X"~ . vcT l-i VO 03 - G 1-1 03 fcH fa C/3 00 b VO O ^ o >> G o u VO to b 7 CM Ov 1 vo O 1 CM vo N J CM CM > > Ov Ov Ov 10 Vil VO VO CM 1 1 1 I VO N 00 H H CO CM CM CM CM CM >>>>>> X Ov O O H CM CO 03 vo to HH ^ b| <7 ^ > > x a > o o O co o E 03 O (h G ,_ g -r W ^ 03 • _ (U fa ^ S r" M c fa C c C CO u £ CM ^ X '§ S § El fa CCS ^ « C sc E co *+■ . G o' S ^ -s PP cr S-4 03 «» Ctf Z ^ z •5? G CO 0 mi. Bahi ate s’ VOJ ™ _ or'E 0 ^ CO G G ~ | o E ^ CM o' - 6n 03 $ ^ 5 « G U X CO 'g '■!> 03 b 03 x u fa fa fa G 03 CO N 03 cn fa ° ^ -T3 03 O fa h vo 03 03 bJ3 w 03 ^ .— 03 G G 03 03 CO CO vo vo 00 21 San Jose del Cabo, 10 mi. SW ca. 100' IX-1-59 2 F. Werner, K. Radford 23 Cabo San Lucas Banks and Snyder, ’20 Psyche [September 170 few colonies were also noted in dead branches, down to one inch in diameter, on living trees in the same areas. An incipient colony was found by George Hofer in the sapwood of a dead palo verde ( Cerci - dium sp. ?) in Sabino Canyon (Banks and Snyder, 1920). The fairly large colony from Guaymas, Sonora, was from a short branch, one inch in diameter, on a living palo verde (species?). Burger took a very large colony in a fragment of palo verde (species?) trunk, about six feet long and 10-12 inches in diameter, partly buried in the sand of Alamo Wash. A colony in a “palo verde stump” was sampled by C. C. Lamb at La Paz, Baja California del Sur (Light, 1929). There are three or four species of Cercidium , one or more of which are very abundant and characteristic trees in various parts of the Sonoran Desert. In the Arizona Upland, and probably elsewhere in their ranges, C. floridum is more abundant on upper bajadas and along drainageways, while C. microphyllum prefers hills, outwash slopes and plains. There seems to be no reason why Pterotermes should not utilize the wood of all these species wherever they occur. In the Tucson Mts. Werner collected specimens from a colony in a disintegrating saguaro skeleton. The only other association with this wood was the small colony in a single rib, about one inch in diameter and a few feet long, found on the open desert floor in the Ajo Mts., also by Werner. Cereus giganteus is prominent over most of the Arizona Upland and the Plains of Sonora, but it is absent from the Foothills of Sonora and Baja California. Extensive collect- ing has been done in southern Arizona for termites and other xylo- phagous insects for many years. Since Pterotermes has not been found in any other wood (mesquite, Prosopis juliflora , and ironwod, Olneya tesota for example) it is not improbable that the palo verdes and saguaro provide the most suitable nesting sites in this part of the desert. In the remaining subdivisions of the Sonoran Desert, however, there is a different and even wider selection of woody plants which should be searched for Pterotermes ; for example, the dead skeletons of other large columnar cacti in the genera Cereus and Pachycereus. The only additional host data accompany the collection made by J. A. Powell from Canyon del Tajo, Sierra Juarez, in northern Baja California. “These specimens were taken in dry (1952) flower- ing stalks of Yucca whipplei on the trail down into the canyon probably at an elevation of about 3000 feet.” Powell also added that “I have subsequently examined stalks of this host over a wide range in California from perhaps 40 or 50 localities in connection 1966] Nutting — Dry-wood termite 171 Figure 1. Map of the Sonoran Desert showing distribution of Pterotermes occidentis: open circles, previous records; solid circles, new records. See Table 1 for key to numbered localities in Mexico. 172 Psyche [September with studies of the moths associated with it and have never seen termites in the stalks elsewhere” (Powell, 1966, in litt.) . A col- lection of nymphs was taken “in Yucca” by C. C. Lamb at Santo Domingo Landing, Baja California del Sur (Light, 1929). The fibrous or pithy tissues of yuccas would seem to be a rather unusual situation for the nest of a dry-wood termite. The haphazard system of tunnels and chambers made by a colony of approximately 3000 individuals is shown in the longitudinal section of a dead palo verde in figure 4 B. In this and in one other standing palo verde, the colony had penetrated wood a few inches below ground level. The rough-surfaced galleries of Pterotermes could not be mistaken for those of any other termite in the desert. Although in a quite different environment, the workings of a large colony are nearly as large and extensive as those of Zootermopsis (Nutting, 1965). Galleries are generally driven in sound wood, although many palo verdes have been found where 50 to 75 percent of the wood has been previously tunneled by wood-boring beetle larvae. In this case the termites work through the tightly packed sawdust and even use the cleared tunnels for short distances. Since the soldier head is approximately 4 mm. wide and the abdomen of a large nymph may be nearly 5 mm. in diameter, the tunnels and access holes between chambers seldom measure less than 3X5 mm. Active galleries are nearly free of fecal spottings and pellets, but large dumps of loose pellets are found in abandoned chambers. In a caged palo verde colony, large quantities of fecal pellets were dumped from an old borer exit a few inches from the ground. Considerable use is made of semi-liquid fecal material in walling off old galleries and in plugging lateral tunnels made by borers to the outside. Colonizing flights. Previous flight records from Arizona had suggested that Pterotermes flies on a relatively few nights during late July and August. Alates from Baja California were taken in light traps in late August and early September by Werner and Rad- ford (Table 1). Apparently the only observations on a flight in nature were made by the author on the night of August 7, 1963, in the Tucson Mts. A lantern was operated from dusk until midnight, and hetween 2235 and 2330 hrs. four males and four females flew rapidly in to the light on the ground. It was impossible to tell from what altitude or direction they had come. They were very active under the light but did not leave. The evening was quiet and nearly cloudless with the temperature between 22 and 23°C. The moon had risen about an hour before the flight and was just past full. As part of a long-term study, flight and meteorological data were 1966] Nutting — Dry-wood ter?nite 173 collected from June through September, 1965, in Tucson, Arizona. A six-foot section of dead palo verde, containing what was probably a single colony of Pterotermes , was set in a field cage (6x12x6 feet) of 20-mesh Saran screen. Observations were made at various times nearly every night as well as for the entire period from dusk to dawn on a few other nights. A small black light trap was run each night in a corner of the cage so that there was reasonable assurance that practically all alates were captured. The colony staged 40 separate flights during the 43 days beginning July 22 and ending Sept. 2. The smallest flight consisted of one alate, the largest 199, and roughly equal numbers of males and females were trapped. Twenty-five alates were later found to have escaped the trap and established themselves in cells within the same tree from which they flew. These plus those from the trap gave a total of 1688 alates produced by the colony. Nearly six months later the entire colony was removed from the tree and found to contain 4055 nymphs (Table 2, No. 3). Assuming that there were approximately 5600 nymphs in the colony during the spring of 1965 (number of nymphs produced since then not counted, but probably less than 200), this means that an impressive 30 percent of them developed into alates. It should be stated that it was impossible to determine whether this was actually a single colony or perhaps two, since a queen and a pair of replacement reproductives were found widely separated in the tree. Although accompanying data have not yet been analyzed, a few generalizations may be made relating weather conditions to the flight season and the daily flight periods. The flight season began about three weeks after the highest weekly mean temperature of the year (30.5°C) was reached and continued for approximately six weeks with weekly means between 26.8 and 29.5 °C. Individual flight periods took place with nighttime temperatures ranging from ap- proximately 19 to 29°C. The flight season occurred while some of the highest weekly mean relative humidities of the year were re- corded: 48-74%. Individual flight periods took place with night- time relative humidities ranging between 39 and 100%. Flight periods began at temperature-relative humidity combinations between 29°C/39% RH and 24°C/iOO% RH. Nearly two inches of rain had fallen during the summer rainy season ( June-September) prior to the flight season, and four significant rains (0.3 to 1.5 inches) occurred during the season itself. No flights took place on three nights during the 43-day season, apparently because of rain or con- siderably cooler temperatures. All flights were staged during dark- Psyche, 1966 Vol. 73, Plate 9 Nutting — Pterotermes 1966] Nutting — Dry-wood termite 175 ness, from two hours or more after sunset until three-quarters of an hour or more before sunrise. A complete account of these flight studies on Pterotermes, including behavior of the alates, will appear as part of a separate series on the colonizing flights and associated activities of termites. Table 2. Composition of five colonies of Pterotermes occidentis from Arizona. It was impossible to determine whether colonies No. 4 and 5 were fragments of a single colony in the same log, or whether No. 3 was actually made up of two colonies in the same tree. K and Q represent primary reproductives ; Rep., replacement reproductives. No. Nymphs Soldiers Oracle Jet., 8 mi. S. Reproductives Eggs 1 721 37 K + Q 79 2 2867 42 K + Q 3 4055 82 Alamo Wash Q + 2 Rep. ca. 30 4 59i 21 2 Rep. 5 1082 28 1 Rep. Founding of the colony. About 30 incipient colonies have been collected in dead palo verdes in the vicinity of Tucson. Several were found 6 to 10 feet above the ground. Observations on these and on post-flight behavior of caged alates indicate that the alates use cracks and borer exits as primary points of attack. The initial chamber, or copularium, (Fig. 4 A) is usually hollowed out of sound wood less than one inch from the surface. A few captive pairs have been maintained in petri dishes containing a slice of wood on a layer of plaster of Paris, with a few drops of water occasionally added to the wood or plaster. Shrunken reproductives and nymphs become excited in the presence of moisture and will readily drink to repletion from free droplets or a cotton wick. However, laboratory colonies may die within two weeks where free moisture or even moderate humidities prevail. The pairs were kept in a study from September through May where temperatures varied from 13 to 2J°C. Explanation of Plate 9 Figure 2. Scene in the foothills of the Santa Catalina Mts. north of Tucson, Arizona. To the left is the ribbed, woody skeleton of a saguaro cactus, Cereus giganteus ; to the right, a palo verde tree, Cercidium micro- phyllum. Pterotermes has been found in these skeletons and in the dead wood of Cercidium. Figure 3. Cercidium — Cereus association on the western slopes of the Tucson Mts. west of Tucson, Arizona. Pterotermes has been taken in this area which is typical of the Sonoran Desert in southern Arizona. Photo- graph by Peter J. Mehringer, Jr. 76 Psyche [September Perhaps because they were already enclosed, many of the pairs worked intermittently for as long as a month to construct the copu- larium: irregular grooves in wood beneath the cover of the dish or chambers within the wood, measuring roughly 5X 10 X6 mm. Semi- liquid fecal material was used freely between wood and cover to encircle the grooves or to plug the chamber entrance. Some pairs pro- duced up to six eggs within the first two or three weeks, while others produced no eggs during ten months of observation. The eggs are opaque, pinkish white and ultimately measure approximately 0.7 X 1.7 mm. Both sexes occasionally manipulated or cleaned the eggs with their mouthparts, but the eggs were not kept together in any particular place. No detailed observations were made on the progress of these incipient colonies, although most of the eggs had not hatched at the end of six months. Replacement reproductives of both sexes have been found in fragments of colonies in the laboratory and in natural colonies. In the latter case it has not always been possible to determine whether they were heading distinct colonies or were associated with remote parts of colonies headed by primary reproduc- tives. They are pale yellowish-brown and possess very short wing pads. As would be expected, these preliminary observations show that captive colonies of Pterotermes provide good material for many types of biological studies. Colony size and other biological notes. Incipient colonies, probably within their first year, have been found consisting of barren pairs and pairs with as many as eight nymphs and five eggs. One colony, presumed to be in its second year, consisted of 13 nymphs, one small soldier and one egg. Another pair, perhaps in their second or third year, had produced 29 nymphs, 2 soldiers and 5 eggs. Data on the size of five older colonies are presented in Table 2. Assuming that all were single colonies, the soldier-nymph ratio varied from 1 to 20 in one of the smaller colonies to 1 to 68 in one containing nearly 3000 nymphs. It should be of interest to mention the extremely high temperatures which Pterotermes must commonly meet in its exposed nesting sites. Temperature records were kept in connection with the flight studies made on the caged colony in Tucson from July through September, 1965. Temperatures were noted several times each day from a remote-reading thermometer, accurate to approximately db i°C. The bulb was fixed within the central galleries of the palo verde trunk, three feet above the ground. Brief excursions above 38° were common in the late afternoon and on one day reached 41. 8°; the maximum air temperature (shade) on that day was 38°C. The 1966] Nutting — Dry-wood termite 177 Figure 4. A. Longitudinal section through a copularium of Pterotermes occidentis in outer wood of a dead palo verde tree. This chamber contained a primary pair and four small nymphs. B. Longitudinal section showing the galleries of Pterotermes in a dead palo verde from the vicinity of Oracle Jet., Arizona. Cross sections 1-9 represent tracings of the galleries in ad- jacent cuts. temperature rarely fell below 24°C in the early morning shortly after sunrise during the entire period. This tree received somewhat more than 50 percent shade for most of the day and complete shade early and late each day. Since colonies in nature are rarely in such shaded situations, these figures are certainly conservative. Further, for most colonies in branches and trunks from one to ten inches in diameter, there is no apparent escape from such temperatures. Association with other termites and insects. In southern Arizona, Marginitermes hubbardi (Banks) has been found in palo 178 Psyche [September verde logs, and on two occasions in the same wood with Pterotermes. Paraneotermes simplicicornis (Banks) is the only other kalotermitid which is known to attack the palo verde. This unusual termite, which is essentially subterranean, generally infests the roots and stumps of dead trees in the desert. Both of these termites probably begin their attacks before wood has dried to the point where it is most favorable for Pterotermes. They are much more common in the skeletons of the saguaro cactus, where they appear shortly after the flesh has rotted and fallen off. They have never been found in saguaros inhabited by Pterotermes. Heterotermes aureus (Snyder) is one of the most common sub- terranean termites in the Sonoran Desert region of southern Arizona. It commonly works in dead saguaros, particularly during periods of higher soil moisture following the winter and summer rains. On the north slopes of the Tucson Mts. its abandoned galleries were obvious in the saguaro stump containing Pterotermes. The only other insects commonly found in wood attacked by Pterotermes are buprestid and cerambycid larvae which are often very numerous in palo verde. REFERENCES CITED Banks, N., and T. E. Snyder. 1920. A revision of the nearctic termites (Banks) with notes on bi- ology and geographic distribution (Snyder). U. S. Nat. Mus. Bull. No. 108, 228 pp. Collins, Margaret S. 1966. Personal Communication. Collins, Margaret, and A. Glenn Richards. 1966. Studies on water relations in North American termites. II. Water loss and cuticular structure in eastern species of Family Kalotermitidae (Isoptera). Ecology, in press. Emerson, Alfred E. 1955. Geographical origins and dispersions of termite genera. Fieldi- ana: Zoology, 37: pp. 473-475. Hagen, H. A. 1858. Monographic der Termiten. Part 2. Linn. Entomol., 12: pp. 77- 78. Krishna, Kumar. 1961. A generic revision and phylogenetic study of the family Kalo- termitidae (Isoptera). Bull. Amer. Mus. Nat. Hist., 122: 303- 408. Light, S. F. 1929. New termite records for Lower California. Pan-Pacific Ent., 6: pp. 67, 69-70. 1933. Termites of western Mexico. Univ. Calif. Publ. Ent., 6: pp. 85- 86. 1966] Nutting — Dry-wood termite 179 Nutting, W. L. 1965. Observations on the nesting site and biology of the Arizona damp-wood termite Zootermopsis laticeps (Banks) (Hodotermi- tidae). Psyche, 72: pp. 119-121. Powell, Jerry A. 1966. Personal communication. Shreve, Forrest. 1951. Vegetation of the Sonoran Desert. Carnegie Inst. Washington Pub. 591, xii + 192 pp. + 37 pis. Snyder, T. E. 1966. Personal communication. POSTEMBRYOLOGICAL DEVELOPMENT OF SPIDERLINGS FROM TWO PERUVIAN LATRODECTUS POPULATIONS By John D. McCrone, Florida Presbyterian College AND Herbert W. Levi, Harvard University It is becoming increasingly clear that the solution of the complex taxonomic problems presented by the black widow genus Latrodectus will require more basic biological data than are now available. Of particular importance in this regard is information on the postembryo- logical development of color pattern. Many Latrodectus populations show marked interpopulation differences in color pattern particularly of the immature stages. It is difficult to assess the taxonomic sig- nificance of these differences without reliable information on the de- velopment of these patterns in various populations. Little such information has been published. The most extensive study was that done by Smithers (1944) with Latrodectus mactans indistinctus . He raised spiderlings from three different areas in South Africa and recorded the rate of development and the changes in color pattern after each molt. Unfortunately he confined his ob- servations on color pattern to the females. Recently we were able to obtain egg sacs from two different areas in Peru, Lima and Cuzco. We have reared 54 adults from these sacs and have recorded the changes in their color pattern after each molt. Morphologically these adults all appear to be Latrodectus mactans , probably the same as Latrodectus species “no. 3” of Abalos’ cultures from Santiago del Estero Province in Argen- tina. The spiders were raised in the laboratory in St. Petersburg, using a technique previously described (McCrone and Levi, 1964). A record was kept for each developing spider of the number of molts to maturity, the duration of the stadia, and the color pattern of each post-emergent instar. These data are presented in Table 1 and Figs. 1 -1 9. The color patterns figured are the basic patterns found in each population. There were slight individual differences in pattern; however^ these did not obscure the basic color patterns. From the illustrations it can be seen that some instars are more polymorphic than others (Fig. 2, 15, 16) even though the adults end up with more or less similar patterns in females (Fig. 9, 18, 19), * Manuscript received by the editor June 28, 1966. 180 TABLE I — RATE OF DEVELOPMENT OF SPIDERLINGS FROM TWO PERUVIAN LATRODECTUS POPULATIONS 1966] McCr'One and Levi — Spiderlings n Cl a r£3 s o T ^ o£ 3 Z.ST3 o ~ Z 6 >s i-H 5 3 O OO + 'O vA 'O u~, O co £ ^ co ^ Cxi ^ Cvt O Os Os us «2