C*^ special Colleclloftl D© not circu/ale HARVARD UNIVERSITY. LIBRARY OF THE MUSEUM OF COMPARATIVE ZOOLOGY GIFT OF yYltA^uuAV| MAY 2 7 1930 BULLETIN OF THE MUSEUM OF COMPARATIVE ZOOLOCtY AT HARVARD COLLEGE, IN CAMBRIDGE VOL. LXX CAMBRIDGE, MASS., U. S. A. 1930 The Cosmos Press, Inc. Cambridge, Mass., U. S. A. fa N t <»i CONTENTS Page No. 1. — The Fossil Ants of North America. By F. M. Carpenter (11 plates). January, 1930 ...... 1 No. 2. — The Lower Permian Insects of Kansas. Part I. Introduc- tion .\XD THE Order jMecoptera. By F. M. Carpenter. (5 plates). February, 1930 67 No. 3. — The Anoles. I. The Forms Known to Occur on the Neo- tropical Islands. By Thomas Barbour. April, 1930 . 103 No. 4. — Types of Birds Now in the Museum of Comparative Zoology. By Outram Bangs. March, 1930 . . . 145 No. 5. — Reconnaissance of the Waters and Plankton of Mon- terey Bay, July, 1928. By Henry B. Bigelow and Maurine Leslie. May, 1930 . " 427 Jfl/v ; 3 mo Bulletin of the Museum of Comparative Zoology AT HARVARD COLLEGE Vol. LXX. No. 1 THE FOSSIL ANTS OF NORTH AMERICA By F. M. Carpenter With Eleven Plates. CAMBRIDGE, MASS., U. S. A.: PRINTED FOR THE MUSEUM January, 19.30 PUBLICATIONS OF THE MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE. There have been published of the Bulletin, Vols. I to LIV, LVI to LXV, LXVII to LXIX; of the Memoirs, Vols. I to LI. The Bulletin and Memoirs are devoted to the publication of original work by the Officers of the Museum, of investigations carried on by students and others in the different Laboratories of Natural History, and of work by specialists based upon the Museum Collec- tions and Explorations. These publications are issued in numbers at irregular intervals. Each number of the Bulletin and of the Memoirs may be sold sepa- rately A price list of the publications of the Museum will be sent on application to the Director of the Museum of Comparative Zoology, Cambridge, Mass. Bulletin of the Museum of Comparative Zoology AT HARVARD COLLEGE Vol. LXX. No. 1 THE FOSSIL ANTS OF NORTH AMERICA By F. M. Carpenter With Eleven Plates. CAMBRIDGE, MASS., U. S. A.: PRINTED FOR THE MUSEUM January, 1930 No. 1. — The Fossil Ants of North America^ By F. M. Carpenter I. Introduction Sixty-three years ago, when the first fossil insects were discovered in American Tertiary rocks, a new era began in the study of the geo- logical history of the insects. Early expeditions under the stimulating leadership of Dr. S. H. Scudder, with the cooperation of the United States Geological Surveys, secured collections of over 20,000 specimens in the Florissant shales alone. For the most part, the material gathered at this time was described by Scudder between the years 1867 and 1900, when he was finally forced into inactivity by paralysis. For a short period of five years the Tertiary insects of the country were entirely neglected, but in 1905 the work was again taken up by the University of Colorado. The following year Professor W. M. Wheeler, Professor and Mrs. T. D. A. Cockerell, and S. A. Rohwer collected extensively at the Florissant locality, and in 1907 also a large expedition was made to the same beds under the direction of the American Museum of Natural History, Yale University, the University of Colorado, the Brit- ish Museum of Natural History, and the Royal Irish Dublin Museum. The insects found on these two expeditions have chiefly been described by Professor Cockerell, Professor H. F. Wickham (Coleoptera), Pro- fessor C. T. Brues (Parasitic Hymenoptera), and S. A. Rohwer (Ten- thredinoidea). The long illness which Scudder endured after 1900, and which finally caused his death ten years later, prevented him from completing the investigations on the Florissant insects, so that when his collection was donated to the Museum of Comparative Zoology in 1902, fully half of the specimens were unstudied. This unworked material was found by Professor Wheeler to include over 4,000 ants which, together with as many more obtained on the later expeditions, were turned over to him for study. At about that time Professor Wheeler was occupied with the preparation of a monograph of the ants of the Baltic amber and after he had finished this task, he was prevented by other matters from carrying out his intention of describing the Florissant ant fauna. In 1925, at the suggestion of Dr. Wheeler, I undertook the study of these 1 Contribution from the Entomological Laboratory of the Bussey Institution, No. 307. Studies aided by an Anna C. Ames Memorial Fellowship and a Sheldon Traveling Fellowship. 4 bulletin: museum of comparative zoology fossils, and have since been able to increase the collection to about 12,000 specimens, and to include ants from other American insect de- posits. Some of this additional material was collected by the writer at Florissant, Colorado, and Green River, Wyoming, during the summer of 1927, but for the most part it has been received from the following sources: the United States National Museum, Washington, D. C; Princeton University, Princeton, N. J.; the Carnegie Museum, Pitts- burgh, Pa. ; the Museum of Comparative Zoology, Cambridge, Mass. ; Professor T. D. A. Cockerell, University of Colorado, Boulder, Colo- rado; Professor H. F. Wickham, University of Iowa, Iowa City, Iowa; Mr. Earl Douglass, Salt Lake City, Utah; and Professor J. H. Johnson, Colorado School of Mines, Golden, Colorado. To these contributors, as well as those mentioned before, I am greatly indebted for the use of their material. For the privilege of examining and photographing the types of pre- viously described fossil ants, I am especially indebted to Dr. E. M. Kindle, Canadian Geological Survey, Ottawa, Canada; Professor T. D. A. Cockerell, University of Colorado, Boulder, Colorado; Dr. R. S. Bassler, United States National Museum, Washington, D. C; and Professor Nathan Banks, Museum of Comparative Zoology, Cambridge, Mass. In designating the holotypes of the new species an attempt has been made to select the individuals showing most of the necessary characters, and since the Scudder collection includes much better material than the others, the majority of the holotypes are in the Museum of Compara- tive Zoology. Only one of the specimens secured on the 1907 expedi- tion of the several museums mentioned above has been selected as a holotype, and at the suggestion of Professor Cockerell, who was in charge of the expedition, this has been placed in the same institution in order to keep as many of the holotypes together as possible. Pro- fessor Wickham has kindly donated the holotypes in his collection to the Museum of Comparative Zoology. Paratypes of the Florissant species have been allotted to the contributing institutions, so that the museums and universities listed above have as nearly a complete series of these ants as the number of duplicates has allowed. To Professor W. M. Wheeler I am more than grateful for the sug- gestion of an investigation which has proved so interesting ; for the use of his splendid collection of recent ants; and above all for the generous amount of time which he has spent with me discussing the fossils and their affinities. Indeed, without Professor Wheeler's assistance this task would never have been satisfactorily completed. Professor Brues carpenter: fossil ants of north AMERICA 5 has also made many helpful suggestions and I am especially apprecia- tive of his interest and encouragement, which has attended the work throughout. Very little study has previously been made on the fossil ants of American deposits. Scudder described (1877a, 1878) four supposed ants from the Green RiAer formation, and five others (1877b) from the Quesnel beds in British Columbia. Cockerell, more recently (1906, 1927), has described three species from Florissant, two from the Green River shales (1921, 1923b), and one from a small deposit in Brazos County, Texas (1923a). Professor Wheeler has briefly referred to these fossils in his general works on ants (1910, 1926, 1928), and in his study of the mountain ants of western North America (1917) has listed the genera of Florissant which he recognized by a cursory examination of the material at his disposal. As a result of this neglect of the American forms, our knowledge of the geological history of the ants has been based almost exclusively upon the fossils found in the Tertiary formations of Europe, of which the most important is the Baltic amber (Oligocene). The ants con- tained in this resin belong to ninety-two species, referred by Mayr (1867) and Wheeler (1914) to the following genera: Prionomyrmex* Bradoponcra* Edatomma, Electwponcra* PJatythyrca, Euponera, Poncra, Sima, Monomorium, Erehomyrma, VoUenhovia, Stenamma, Aphacnogasier, Eledromyrmex * Agroecomyrmex* Myrmica, Notho- myrmica* Leptoihorax, Stiphromyrmex* Enneanierv^* Protaneuretus* Paraneureius* Dolichodents, Iridomyrmcx, Liometopum, Asymphylo- myrmex* Pitycomyrmex* Plagiolepis, Rhopalomyrmex* Dimorpho- myrmex,^ Gesomyrmex, Prodimorphomyrmcx* Oecophylla, Prenolepis, Lasius, Formica, Glophyromyrmcx* Pscudolasius, Dryomyrmex* and Camponotus. The Formicidae of the much rarer Sicilian amber were studied by Professor Carlo Emery (1891, 1913), who recognized the following genera: Edatomma, Ponera (f), Cataulacvs, Hypomyrmex,* Podomyrmex,* Aeromyrma, Meranophis, Leptothorax, Tapinoma, Tccknomyrmcx, Plagiolepis, Gesomyrmex, and Oecophylla. The two most productive of the Tertiary rock deposits of Europe, as far as insects are concerned, are at Radoboj in Croatia, and Oeningen in Baden. The ants of the Radoboj formation were described by Heer before our present conception of the genera of ants had been reached, so that the species were lumped into Formica, Myrmica, Ponera, and Attopsis. Fortunately, Mayr was able to examine a number of speci- * Extinct. ' Recently shown by Wheeler to be synonymous with Gesomyrmex (Psyche, 36, p. 1-12, 1929). b bulletin: museum of comparative zoology mens determined by Heer, and to correct the generic determinations (1867). He considered that the following genera were represented: Tetramoriwn (f), Prenolepis (f), Aphaenogaster, Myrmica {?), Catau- lacus, Leptothorax (f), DoUchodenis, Liometopum, Lonchomyrmex* Plagiolep'is, Oecophylla, Lasius, Formica, and Camponotus. The Oeningen ants were described by Heer also, at the same time, but since no myrmecologist has revised his determinations, we are obliged to disregard these ants at present. Two deposits in the British Isles have yielded a few members of this family. From the older of these, the Bagshot beds of Bournemouth (Eocene), Cockerell has described (1920) two species, but since only the wings are preserved, the generic determinations are very dubious. The second deposit is at Gurnet Bay, Isle of Wight, and belongs to the Oligocene period. A few ants from there were first described by Cockerell (1915), and these were later revised and added to by H. St. J. K. Donisthorpe, the well-known British myrmecologist, who was able to examine a large series of these fossils (1920). The genera recog- nized by Donisthorpe include Syntaphus* Euponera, Ponera, Emplas- tiis* Dolichoderus, Leucoiaphvs* Oecophylla, and Caviponotus. The ants of the other European deposits have not been sufficiently well treated to warrant their mention in this paper. The stud}' of fossil insects, with the exception of most of those im- bedded in amber, is beset with many difficulties which make progress exceedingly slow, and which at times are responsible for no little dis- couragement. These obstacles are the direct result of the flattened condition of the insects, caused by the pressure of the strata above that containing the specimens. As the weight of these strata increases with the accumulation of sediment, the insects are pressed almost into a single plane. The disadvantage of this is obvious, for when the sys- tematic position of a living insect is to be determined, the specimen must usually be examined in various positions and attitudes in order to reveal all the necessary characters. But since the fossil insect can be seen in just one position, only those characters visible in this posi- tion can be determined. The shape of the head, for example, can be used as a descriptive character only when the fossil shows a dorsal aspect. In the case of the ants this flattening is especially disconcert- ing, because the dorsal aspect of the head and a lateral view of the pedicel are nearly essential for the determination of the affinities of a species. Fortunately, there are a few structures, such as antennae and wings, which are visible in any attitude and are consequently the prin- ciple means of correlating the specimens in various positions. * Extinct. carpenter: fossil ants of north America 7 There are two types of distortions of the ants also resulting from flattening. The most obvious of these is the increase in the width of the specimen which takes place as the latter is pressed flat. The effect of this, of course, is to give the insect a more robust appearance than was characteristic of the ant when alive. Apparently the intensity of this pressure was very great, for the chitin of the head of many specimens is distinctly cracked just in front of the posterior angles, as in the holo- type specimen of Formica cockerelli, sp. nov. (Plate 4, fig. 3). This splitting of the chitin occurs in precisely the same place, if a recent ant is pressed flat, and since it is always followed by a collapse of the entire head, which consequently becomes much broader, the presence or absence of the splitting indicates the degree of flattening which has taken place and provides a means of determining the original shape of the head. The second type of distortion is less evident, but equally noteworthy. It will be observed in most of the photographs of the Florissant species that the eyes are more remote from the lateral mar- gins of the head than they are in the majority of living species. This is not a morphological peculiarity of the extinct species, but is merely due to the flattening of the head, and can be duplicated in recent forms by applying the necessary amount of pressure. The application of a dilute solution of damar to the fossil has been found to improve the visibility of the insects nearly a hundred per cent. jVIany structures, especially the antennal segments and the veins of the wing, which could not otherwise be discerned in some specimens, become very distinct by the use of this medium. The resin hardens in a few hours, thus serving to protect the fossil from dust or scratches, and even from the cracking caused by the changes of atmospheric condi- tions. The hardened damar can easily be removed if desired, by soak- ing the specimen in xylol for a few days and then washing it for an equivalent time in absolute alcohol. The Florissant ants are the only ones from American deposits which are sufficiently well preserved to permit determination of the generic affinities. Even the Green River shales, which have yielded a great number of splendidly preserved insects of other groups, have not pro- duced a single satisfactory ant. My observations of the European fossil ants lead me to the conclusion also that those of Florissant are far better preserved than those of any other known deposit, excepting, of course, the Baltic amber. I have never seen a Radoboj or Oeningen ant with the eyes, antennae, or clypeus preserved, and very few of those described by Heer from those two localities show such details, as do the many of the Florissant specimens. 8 bulletin: museum of comparative zoology Only a small percentage of the ants in the collections at my disposal consist of both obverse and reverse. This is rather unfortunate, since the reverse is never the mirror image of the obverse. If, for example, a specimen is preserved in a dorso-lateral position, one half shows the structures as seen from above (eyes, clypeus, etc.) and the counterpart, only those visible from beneath (maxillae, etc.). This condition is well illustrated by the holotype of Archiponcra wheeleri, sp. nov., of which the obverse is shown on Plate 1 and the reverse on Plate 2, fig. 1. When, however, the ant has been excessively crushed, as has frequently happened, the structures on the dorsal surface of the body may be faintly impressed on the ventral half. Of course, there are no structural differences in the halves of a fossil showing a lateral view of the ant, since the latter is bilaterally symmetrical. Although the classification of living ants is based largely on workers, the poor representation of this caste among the fossils prevents us from following the same procedure in this study. As a consequence, I have substituted wherever possible the female for the worker as the im- portant form of the species. The nature of the preservation of the ants has also required the selection of taxonomic characters somewhat differ- ent from those usually employed in the study of recent ants. The vena- tion of. the forewing is nearly indispensable for classification of the fos- sils, and inasmuch as the commonest castes are males and females, most of the specimens are winged. It is very essential, however, that venational characters be used with considerable caution, for in the ants as a whole the nature of the venation seems to be of little phylo- genic value. The arrangement of the veins in some of the highly special- ized myrmicines, for example, is identical with that of certain primitive ponerines, yet the venation of two species within the same genus may be utterly different. Many ants, as Lasius umhratus (Adolf, 1880), have an exceedingly variable venation, and only a very few species, if any, have the shape of the cubital and discoidal cells exactly con- stant. It is not practical, therefore, to base a species on the micro- measurements of the sides of a cell, as Cockerell has done in his de- scription of three ants from Florissant and a number of others from European deposits. There are some genera, however, which have the arrangement of the veins fairly constant and many of these are made distinctive by certain peculiarities which prove a great help in recog- nizing the genus — e.g., in Myrmica the apical half of the first inter- cubitus is always lacking. As far as the Florissant ants are concerned, the venation affords the best means of distinguishing the dolichoderines from the formicines, for the character ordinarily used to separate these ( arpenter: fossil axts of north America 9 subfamilies, the shape of the cloaca, cannot be seen in the fossils. Vena- tional studies have shown that if a member of one of these groups has two cubital cells, it is a dolichoderine; if it has only one cubital cell, it mav belong to either subfamilv. The shape of the head, although somewhat distorted in most of the fossil ants, can nevertheless be used as a dependable character. In a large series of specimens of one species at least a few individuals are only very slightly flattened or distorted, and, even if all the specimens of a species are somewhat distorted, it is possible, as indicated above, to obtain a fairly accurate conception of the shape of the head. The mandibles are preserved in most of the Florissant specimens, often with sufficient completeness to show the details of the dentition. The cly- peus is occasionally, but not frequently, visible at least to such an ex- tent that the contour of the posterior margin can be determined. The antennae furnish the most reliable characters and this is especially advantageous because the same structure is likewise used in the classi- fication of recent forms. The value of antennal characters in the fossils is also dependent upon the fact that the antennae are only a very little, if at all, distorted by the pressure which flattens the body of the insects. The length of the scape and the relative size of the funicular segments have been determined for nearly all of the Floris- sant ants, the only exceptions being a few aberrant forms which de- serve description because of certain peculiarities. The eyes and even the ocelli have been discerned in most of the species, but it has already been explained above that the position of the eyes with respect to the lateral margins of the head is more or less dependent upon the amount of pressure to which the ant has been subjected. The color of the ants of the Florissant deposit does not usually indicate the original color of the insects and is of little use in identifying the fossil species. Brues has observed (1910) that the metallic colors of the parasitic Hymenop- tera were clearly preserved in the Florissant specimens, but the pig- mental colors of the ants appear to have been affected by the chemical activity which took place during the process of preservation of these insects. Individuals of a species vary from light brown to black, de- pending at least partly upon the rapidity of entombment, for the lighter specimens are usually much better preserved than the darker ones. There are, however, a few species, such as Lasius peritulus (Ckll.) and Formica robusta, n. sp., which are always brown, and since I have found this to be true for the hundreds of individuals of these two species which I have examined, it is very probable that the living ant was this same color. The relative size and qualitative dimensions of the 10 bulletin: museum of comparative zoology various parts of the ants are a necessary addition to the specific de- scriptions, and in the case of some males and a few workers, it has not been possible to give any other specific characters. Unless otherwise noted, the dimension given is the average of the results obtained from measurements of all the individuals of a species; only when a species has been found to be unusually variable are the two extremes indicated. With some exceptions, each description of a new species is accom- panied by a photograph of the holotype and a diagrammatic drawing of the ant. The photographs are essential to show the habitus of the fossils and will be of much assistance in the determination of material, although few details are visible in photographs of the size used. The drawings are not based upon any one specimen, except in the case of uniques, but are composite pictures containing all the characters which have been found in the specimens of the species illustrated. They are not, however, reconstructions in the usual sense of the term. The legs have been omitted from the figures, since they are not ordinarily well enough preserved for taxonomic purposes. The preceding discussion has been made rather detailed in order to explain some of the problems encountered in this study, and the methods by which they have been partly, at least, overcome. This was considered advisable because the average entomologist appears to be skeptical of the results obtained by the study of fossil insects. The specialist who has for many years been determining his species by the distribution of hairs on the insect's head or the structure of the genitals naturally doubts the systematic value of the gross characters which are alone visible in the fossils, and consequently hesitates to accept the conclusions of the palaeoentomologist. Those who hold such an opinion have, I believe, overlooked the very significant fact that the study of fossil insects is essentially a division of palaeontology, rather than entomology. The palaeoentomologist is primarily concerned with the phylogeny of the insects, and whether or not one of the extinct forms has a little more pubescence on the abdomen than another is of very little consequence. If I have included under the name of Formica ro- btista two closely related species, differing only by the intensity of sculpturing on the clypeus, our conception of the geological history of the ants remains unchanged. The important fact is that the genus Formica, or even that a Formica-like genus, existed in Colorado during the Miocene. The reconstruction of prehistoric life is always a slow process, whether we are concerned with the minute insect or the gigantic dino- saur. The picture of the earth's past is necessarily formed by the grad- carpenter: fossil ants of north AMERICA 11 ual accumulation of fragments which, when placed together, make the whole. Just as the extinct reptile, at first known only by a single bone, is finally completely recognized by the addition of further material, so the fossil insect, originally represented by awing or parts of the body, eventually becomes known to us in all details. And although the ac- cumulation of the necessary specimens may be delayed for many years and the important details missing for an equivalent time, the results, on the whole, are dependable. II. North American Ant Deposits Fossil ants have been found in five American localities : ' the Green River formation of Colorado, Wyoming, and Utah; the Florissant shales of Colorado; the Elko oil shales of Nevada; the Quesnel clays of British Columbia; and the Fayette sandstone of Texas. The oldest of these, and in fact the oldest known ant deposit, is the Green River formation. This deposit has been known to be fossiliferous since the middle of the last century when John Evans collected a small fish in the beds near Green River, Wyoming. It was not until 1867, however, when the Hayden Geological Survey began a series of explorations of the Northwest Territories, that the fauna and flora of the shales were systematically studied. At that time Dr. F. V. Hayden, the director of the survey, named and described the deposit as follows (1873) : " A little east of Rock Spring station [Wyoming] a new group commences composed of thinly laminated chalky shales, which I have called the Green River shales because they are best displayed along the Green River. They are evidently of purely fresh water origin and of middle Tertiary age. The layers are nearly horizontal and, as shown in the valley of Green River, present a peculiarly banded appearance. . . . The flora is already extensive, and the fauna consists of Melanias, Corbulas, and vast quantities of fresh water fishes. There are also numerous insects and other small undetermined fossils in the asphalt slates." As these geological explorations continued, it became apparent that the same shales extended into Colorado, Utah, and other parts of Wyo- ming (Emmons, 1877; Endlich, 1878; Peale, 1876; White, 1878). In recent times more detailed studies on the geology of the formation have been made by Winchester (1923) and Bradley (1926). The shales were ' Since this paper was written (1928) a few ants have also been found in the Miocene (Latah) of Washington, and a single specimen has bsen collected in the Eocene (Wilcox) of Tennessee. 12 bulletin: museum of comparative zoology originally supposed to have been deposited by a large lake, some three hundred miles long and one hundred and fifty miles wide, and contain- ing fresh water, as mentioned by Hayden in the description quoted above. Evidence accumulated within the past two years, however, indicates that the beds were formed by a number of small lakes, wdth a sahne content at least part of the time (Bradley, 1926; Henderson, 1926; Cockerell, 1926). Studies on the plants of the formation have determined the geologi- cal age as approximately middle Eocene. Knowlton (1922) in his ex- cellent revision of the flora lists eighty-four species of plants and pre- sents some interesting conclusions on the environment of the biota: "... It appears that an overwhelming preponderance of the living forms in the families represented in the Green River flora are inhabi- tants of tropical or subtropical regions, many of them in both hemi- spheres, yet a considerable number include either genera or species that extend into temperate regions. . . . The physical setting can be pictured somewhat as follows: about the shores of the lake were certain flat, low-lying areas, some of them probably swampy, others sandy, whereon grew the palms, figs, Lomatia, Oreodaphne, hackberries, the several papilionaceous trees and shrubs, the ferns, grass, sedge, etc., and in the water the pickerel weed, Brasenia, algae, etc. On the adja- cent somewhat higher land might have been the willows, waxberries, sweet fern, walnuts, oaks, sumacs, maples (?), hollies, etc. . . . The conditions of temperature and moisture under which the Green River flora flourished are somewhat difficult of interpretation, as there is seemingly more or less conflict between the elements of the flora. The nearest living relatives of certain of the genera that are believed to have inhabited the lowlands . . . are found mainly in tropical and subtropical areas. The palms, at least one species of which existed in abundance, could hardly have lived where the temperature fell below 42° F. and probably not even where it was considerably higher than this. . . . The upland flora . . . could well have withstood some degree of frost, but on the other hand all these genera contain species that could find a congenial habitat in a warm temperate region. It is doubtful if any of them had to withstand cutting frosts." The insect fauna of the formation contains nearly 300 described species and is not very different from that of the region at the present time. The abundance of the Fulgoridae, however, is rather striking, and Cockerell (1920) believes that these insects have a certain tropical appearance and resemble tropical genera. On the other hand, Alex- ander (1920) considers that the tipulid fauna is typical of that of the carpenter: fossil ants of north AMERICA 13 north temperate region, so that the insects, as well as the plants, ap- pear to show both tropical and temperate affinities. The only other American Eocene deposit to yield fossil ants is at Mossy Creek, about three miles southwest of Wellborn, Brazos County, Texas. The beds, which belong to the Jackson series, consist at this outcrop of kaolinite lenses in sandstone, and contain twenty-four species of plants, including a common Comhretum and the littoral palm, Nipaditcs. Professor E. W. Berry (1924) has concluded more or less tentatively, from his study of the flora, that the latter is indicative of a subtropical climate and a strictly coastal location. The insect fauna is very little known, only two species having been described. The deposit at Quesnel, British Columbia, consists chiefly of fine grayish and greenish-white clan's. The age of the formation is still somewhat uncertain, but the latest researches point to the Miocene (Reinecke, 1920). The flora is badly in need of revision, and since the insect fauna is a small one, nothing definite can be said of the climatic conditions under which the biota existed. The two remaining ant deposits also belong to the Miocene and appear to be very similar as to fauna and flora. The smaller of these is the oil shale of Elko and its vicinity, in Nevada. No fossil insects have previously been described from this outcrop or, in fact, from any other rocks in the state, although the presence of insects has been recognized since Emmons's explorations during 1867-73. In his report on the geol- ogy of the region (1877) he states, "Adjoining the coal beds are fine bituminous shales, which closely resemble the brown paper shales of the Green River series at Green River City, Wyoming. In these are found the same plentiful remains of fishes, and also occasional insects." The geology of this bed was more carefully investigated by Winchester (1923), who states that " the shales . . . are in part clean clay shales but are mainly sandy. They usually lack sharp and distinct lamination and are generally interbedded with thin layers of muddy sandstone. In color they are commonly light gray, bluish gray, or brown. . . . Very thinly laminated paper shales are common at certain horizons." These strata have yielded a few fossil plants, which have been referred to the following genera (Knowlton, 1919): Coiiiptonia, Carpinus, Fagvs, Ficus, Lycopodium, Myrica, Planera, Popidus, Scdix, Sapoiadtes, Sequoia, and Thuja. Lesquereux, who first studied the flora, believed (1878) that the beds were the same age as those at Florissant, and Cope came to this conclusion from his studies on the fishes. At that time the Florissant shales were placed in the late Eocene or Oligocene, but further researches by Cockerell, Henderson, and Knowlton have 14 bulletin: museum of comparative zoology shown that they belong to the Miocene. In 1919 Knowlton definitely referred the Elko shales to this latter horizon, and this decision was later substantiated by the discovery of a Miocene mammal in the deposit (Winchester, 1923). The bed covers only a small area, not over thirty square miles, and appears to have been laid down by a fresh water lake under climatic conditions not unlike those which existed at Florissant during the Miocene. The only insects from this deposit which I have been able to locate are in the Museum of Comparative Zoology, and were collected by S. W. Garman in the thinly laminated paper shales about twenty miles northeast of the Elko station. The other Miocene ant beds are the Florissant shales, which are lo- cated about thirty miles west of Colorado Springs, Colorado. Scud- der's description of the location of the deposit is so admirable that I quote his own words : " By climbing a neighboring peak, thrice bap- tized as Crystal Mountain, Topaz Butte, and Cheops Pyramid, and known to the old miners as Slim Jim, we obtain an admirable view of the ancient lake and the surrounding region. To the southeast is Pike's Peak; to the west, South Park and the canon of the South Platte, shown by a depression; to the extreme south the Grand Caiion of the Arkan- sas; while to the north a few sharp, ragged, granite peaks surmount the low wooded hills and ravines characteristic of the nearer region. Among these hills and ravines, and only a little broader than the rest of the latter, lies to the south, the ancient Florissant Lake basin, marked by an irregular L-shaped grassy meadow, the southern half broader and more rolling than the northwestern, the latter more broken and with deeper inlets." This deposit, which has produced more insects than any other known locality, was found to be fossiliferous by A. C. Peale in 1876. The geology and stratigraphy have been discussed in detail by a number of investigators, so only a brief survey of that aspect wnll be presented here. The upper part of the formation alone is fossiliferous, and this is composed of strata which vary much in thickness and com- position, although for the most part volcanic ash, sand, and mud are the constituents. The shales apparently had their origin at the bottom of a lake, in the vicinity of which were a number of active volcanoes. The dust and ashes from the frequent eruptions of these volcanoes fell to the surface of the lake, carrying along the insects which happened to be flying or blown over the water, and quickly entombed them in a matrix of ash, sand and mud. Leaves of trees and shrubs, torn from their branches by violent winds and falling cinders, are exceedingly common in these shales. The flora of the deposit, which has been studied mainly by Lesquereux (1878, 1883), Kirchner (1878), Cockerell carpenter: fossil ants of north AMERICA 15 (1908), and Knowlton (1917), includes such genera as Acacia, Acer, Alnus, Amclanckicr, Aster, Bctula, Carpiims, Comptonia, Ficus, Fraxi- nus, Hicoria, Ilex, Juglans, Magnolia, Myrica, Pinus, Populvs, Quer- cus, Rhamnus, Rhus, Rosa, Salix, Sequoia, Smilax, and Ulmv^. The insect fauna is exceptionally large, over a thousand species hav- ing been described, and seems to be modern in most respects. Both the insects and the plants suggest that the climate at the time of the depo- sition of the shales was similar to that of our southern states. Scudder has frequently observed that some of the insects have subtropical and even tropical affinities, and Cockerell has also called attention (1907) to a few genera which are now restricted to the old world (e.g. Glossina). The ant fauna shows this same geographical relationship. III. The Eocene Ant Fauna 1. The Green River formation, belonging to the Middle Eocene, contains the oldest ants know^n.' The only other ant deposits of Eocene age are the Bagshot beds, England, and the Fayette sandstone, Texas, both of which are somewhat younger than the Green River. The shales of this latter formation also have the distinction of being the first American rocks to produce Tertiary insects, one of the first specimens collected being an ant. In 1865, Professor William Denton, of Boston, discovered a series of Tertiary beds at the Junction of the Green and WTiite Rivers, near the Colorado-Utah border (Fossil Canon and Chagrin ^'alley). During the course of his examination of the petroleum shales which formed a part of the deposit, he found numbers of " Dip- terous insects, especially mosquitos, and their larvae" (Denton, 1866). The insects were examined by Scudder who reported that the collection consisted of ninety specimens, representing sixty-five species, one of which belonged to Myrmica. This ant Scudder later concluded to be a dolichoderine, and described it as Liometopum. pingue. Three years later. Dr. F. V. Hayden, who conducted many geological explorations into the Northwest Territories, found a few insects in a bed of these same petroleum shales which were exposed along a section (" Petrified Fish Cut") of the then recently built Union Pacific Railroad, at Green River City, Wyoming. Scudder studied these insects also, and stated that they belonged to " three species, one being an ant, the others flies. The ant is rather poorly preserved, and must be examined with great care before its precise characters can be determined." This species was ' An earlier ant, Euponera berryi Carp., has recently been found in tbe Lower Eocene of Tennessee. See Journ. Wash. Acad. Sci., 19, p. 300-301, 1929. 16 bulletin: museum of comparative zoology eventually described as Lasius terreus. During 1870 Scudder himself collected in these shales, both at the Wyoming and Utah exposures. In more recent times, ants haAe also been taken in this formation by Mr. Earl Douglass (1908, 1923), Mr. J. L. Kay (1923), Mr. Dean Win- chester (1916), and Professor and Mrs. T. D. A. Cockerell (1922). Although five supposed ants have been described from the Green River shales, only the one following is well enough preserved so that it can be placed in a subfamily with any degree of certainty. I am unable to add anything definite to the original description of the genus or species. Myrmicinae Archimyrmex Cockerell CockereU, T. D. A., 1923. Entomologist, 56, p. 51-52. " Rather large, elongated ants, with a general resemblance to Myrme- cia, but with the eyes (as in Prionomyrmex) high up on the side of the head; the epinotum with a distinct elevation (presumably pair of ele- vations), placed as in Edatomma tuberculatum, but large and obtuse; mandibles less elongate, but still long, the cutting edge with coarse, obtuse teeth, between which are smaller ones; femora apparently shorter than in Myrmecia; first joint of pedicel elongated, with a dorsal elevation beyond the middle, the joint less massive than in the other two genera, but similar in principle to that of Myrmecia vindex Smith; second joint large and robust, quite like that of Myrmecia, as also the gaster." Genotype. — Archimyrmex rostratus Ckll. Archimyrmex rostratus Ckll. (Plate 2, fig. 5) Cockerell, T. D. A., 1923. Entomologist, 56, p. 51-52. Wheeler, W. M., 1928. Soc. ins., p. 117. "Worker: Color as preserved brown, the upper part of head and the gaster blackened, the coloration perhaps originally similar to that of Myrmecia vindex var. nigraceps Mayr. Length nearly 16 mm.; head with mandibles about 4 mm.; thorax about 5.3 mm. ; middle femurabout 3.7 mm." Loccdity. — Roan Mountain, Colorado (Ute Trail). Holotypc. — Obverse, no. 15174, University of Colorado; reverse, no. 69617, U. S. N. M. carpenter: fossil ants of north AMERICA 17 In his original description Cockerell regarded this ant as a ponerine, because of the apparent constriction at about the middle of the gaster. Wheeler, however, concluded (1928) from Cockerell's figure that "the specimen is more probably a Myrmicine. This is suggested by the shape of the petiole, the blunt or broken (?) spine on the epinotum and the shape of the head, which is unlike that of the existing Ponerinae." Through the kindness of Dr. R. S. Bassler I was able to study and photograph the reverse of the type at the National Museum. The obverse, at the University of Colorado, I was also permitted to ex- amine, but was not able to secure a satisfactory photograph because of the lack of the necessary apparatus. My observations on these fossils did not aid materially in determining the affinities of the ant, although I believe that what appears to be an epinotal spine is merely the only part of the epinotum which is exposed to its dorsal surface, the adjoin- ing parts of the epinotum being covered by the matrix of the rock. In as much as the specimen was a holotype, however, no attempt was made to test this conclusion by exposing the hidden part of the thorax. The habitus of the insect is certainly more suggestive of a myrmicine than a ponerine. The two following Green River species, although unquestionably ants, are not sufficiently well known to permit even subfamily classifi- cation. EoFORMiCA PINGUE (Scudder) (Plate 2, fig. 6) Ldometopum pingue, Scudder, S. H., 1877. Bull. U. S. Geol. Geogr. Surv. Terr., 3, p. 742-743. Liotnetopum pingue, Scudder, S. H., 1890. Bull. U. S. Geol. Surv., 13, p. 617. Eoformica eocenica, Cockerell, T. D. A., 1921. Proc. U. S. Xat. Mus., 59, p. 38. Eoformica eocenica, 'VVTieeler, W. M., 1928. Soc. Ins., p. 117. Male. — Length, 7.5 mm.; head small, nearly round, but somewhat broader behind; thorax broad, about twice as wide, and more than twice as long as the head; gaster rounded, about twice as long as the head and somewhat broader. Length of head, 1.2 mm.; thorax, 3.0 mm. ; gaster, 2.7 mm. ^Yidth of head, 1 .2 mm. ; thorax, 2.2 mm. ; gaster, 2.7 mm. Locality.— Green River City, ^yyoming; Rio Blanco County, Colo- rado; Uinta County, Utah. Holotype.— No. 2937, M. C. Z. This species was originally referred by Scudder to the genus Liomc- 18 bulletin: museuai of comparative zoology topum, but none of the known specimens are sufficiently well preserved to warrant this conclusion. In 1921 Cockerell described an ant (no. 66932, U. S. N. M.) from the Green River shales as Eoformica eocenica, for which he established a new genus. At the end of his description he suggested that this species might be identical withScudder's Liomctopum pinguc, and my comparison of the two types shows that this really is the case. Cockerell attempted a restoration of the frontal view of the head of this species, although the only specimen which he saw pre- sented a lateral aspect of the insect. Through the courtesy of the National Museum I was able to make a careful study of the fossil which Cockerell used as the basis of this restoration, but could not discern any definite indications of the eyes or mandibles shown in his figures. If either of Cockerell's drawings (which do not agree in certain details) represent the true characteristics of the ant, the species can have only the remotest affinities with Formica or Liometopum. Regardless of the vague relationships of this insect, there are several noteworthy features associated with it. Although none of the other Green River ants is known from more than a single specimen, I have seen twenty-six individuals of E. pinguc, all of which are males. The wide distribution over the various outcrops of the Green River forma- tion is also remarkable. Scudder's specimens were collected at Fossil Canon, White River, Utah (Denton), and Green River, Wyoming (Packard) ; and the one described by Cockerell as Eoformica eocenica, at Cathedral Bluffs, Colorado. The additional fossils which I have examined represent the following localities: Wagon Hound Caiion, Uinta County, Utah (Douglass); White River Caiion, Uinta County, Utah (Kay); White River, Uinta County, Utah (Douglass); Roan Mountains, Rio Blanco County, Colorado (Winchester and Cockerell) ; Dripping Rock Caiion, Rio Blanco County, Colorado (Douglass); and Green River, Wyoming (Winchester). If the species were not so common at the localities mentioned, which are spread over an area of about 34,000 miles, one might easily assume that the relative abun- dance of individuals was due to the drowning of a number of specimens of a single nuptial flight, which happened to be directed over the lake. But the regular occurrence of the species over so large an area is con- clusive evidence that this ant was in reality the commonest in the vicinity of the Green River lakes. That the species also existed in the region for a long period is evinced by the presence of specimens at various levels of the shales, which at some exposures exceed a thousand feet in thickness. carpenter: fossil ants of north AMERICA 19 (Formicidae) terreus (Scudder) Lasius terreus, Scudder, S. H., 1878. Bull. U. S. Geol. Geogr. Surv. Terr., 4, p. 747-748. Lasius terreus, Scudder, S. H., 1890. Bull. U. S. Geol. Surv. Terr., 13, p. 618; pi. 10, fig. 23. Worker (f). — Length, 7.5 mm.; "head small, rounded, with an- tennae shaped as in Lasius, but of which the number and relative length of the joints cannot be determined from their obscurity; the long basal joint, however, appears to be comparatively short and of uniform size, being not quite so long as the width of the head, while the rest of the antenna is more than half as long as the basal joint and thickens very slightly near the apex. The thorax, preserved so as to show more of a dorsal than a lateral view, is compact, oval, less than twice as long as broad, with no deep separation between the meso- and metathorax, tapering a little posteriorly. The peduncle, as preserved, is a minute circular joint, but from its discoloration appears to have had a regular, rounded, posterior eminence. The abdomen consists of live joints, is very short, oval, compact and regular, and of about the size of the thorax, although rounder." Length of head, L4 mm. ; thorax, 3.3 mm. ; gaster, 3.0 mm. Width of head, LO mm.; thorax, L9 mm.; gaster, 2.2 mm. Localify. — Green River City, Wyoming. Holotype — No. 69618 U. S. N. M. The single specimen 'of this species is very poorly preserved, and since I have not been able to discern in the type many of the characters given by Scudder in the description quoted above, it is very probable that the fossil has deteriorated since Scudder's examination of it. The assignment of this species to Lasius or any other existing genus will not be justified until additional specimens have been found. The two follomng Green River insects described by Scudder as ants do not really belong to the family, but apparently to some other groups of aculeate Hymenoptera: "Myrmica sp." Scudder Scudder, S. H., 1878. Bull. U. S. Geol. Geogr. Surv. Terr., 4, p. 748. "Camponotus vetus" Scudder Scudder, S. H., 1877. Bull. U. S. Geol. Geogr. Surv. Terr., 3, p. 742. In addition to the preceding fossils I have seen four other ants from the Green River formation, collected by Mr. Earl Douglass in Uinta 20 bulletin: museum of comparative zoology County, Utah. One of these, an isolated specimen, is possibly a poner- ine queen ; the remaining three, which are close together on a small slab of the shale, seem to belong to different genera, but they are so poorly preserved that their generic affinities cannot definitely be ascertained. At any rate, these four species, together with the three discussed above, are sufficient to show that the ant fauna of the Middle Eocene was es- sentially a modern one, at least to the extent that several of the living families were already established and the castes differentiated. 2. The kaolinite of the Jackson formation (Upper Eocene) has yielded a single, splendidly preserved forewing of an ant. (Formicidae) eoptera (Ckll.) Formica eoptera, Cockerell, T. D. A., 1923. Amer. Journ. Sci.,5 (29), p. 399-400. "Anterior wing, 11 mm. long and 4 wide; hyaline, faintly reddish, with pale but stout veins; stigma lanceolate, slender; costal cell very slender; basal nervure wnth upper section only slightly out of straight line with lower, the lower distinctly but not much longer; nervulus about 1.6 mm. basad of basal nervure; discoidal cell large, subquad- rate, but narrower above than below, and apically broader than basally, the upper basal corner obtuse; marginal cell long and broad, its inner corner acute; marginal and cubital nervures forming a cross as in Camponotus." Locality. — Mossy Creek, Brazos County, Texas. The holotype of this insect appears to be lost; it is not recorded at the National Museum, where the rest of Professor Berry's types are lo- cated. As Wheeler has already pointed out (1928), a generic determination of this ant is impossible. IV. The Miocene Ant Fauna 1. The ant fauna of the Quesnel clays (Fraser Formation) is even more fragmentarily known than that of Green River; only four speci- mens have been found and these are so poorly preserved that very little can be said of their affinities. The species represented by these fossils were described by Scudder, three of them as ants and the fourth as a brachonid. carpenter: fossil ants of north AMERICA 21 (Dolichoderinae) obliterata (Scudder) Hypoclinea obliterata, Scudder, S. H., 1877. Rep. Progr.Geol. Surv.Can., 1875- 76, p. 267. Hypoclinea obliterata, Scudder, S. H., 1890. Bull. U. S. Geol. Surv., 13, p. 616, pi. 3, fig. 25, 26. The specimen from which this ant was described consists only of the gaster, the posterior half of the thorax, and a forewing. Since the pedicel is apparently single jointed and the wing has two cubital cells, the species is probably a dolichoderine, but there is no evidence what- ever that it belongs to Dolichoderus. Holoiypc. — Obverse, no. 6179, Canadian Geological Survey; re- verse, no. 2938, M. C. Z. (Myrmicinae) longaeva (Scudder) Aphaenogaster longaeva, Scudder, S. H., 1877. Rep. Progr. Geol. Surv. Can., 1875-76, p. 267. Aphaenogaster longaeva, Scudder, S. H., 1890. Bull. U. S. Geol. Surv., 13, p. 61, pi. 13, fig. 24. The forewing and some parts of the body are preserved; the pedicel is typically myrmicine, but no generic determination can be made. Holotypc. — Obverse, no. 6178, Canadian Geological Survey; reverse, no. 2939, M. C. Z. (Formicidae) arcana (Scudder) Formica arcana, Scudder, S. H., 1877. Rep. Progr. Geol. Surv. Can., 1875-76, p. 266-67. Formica arcana, Scudder, S. H., 1890. Bull. U. S. Geol. Surv., 13, p. 618, pi. 13, fig. 24. This ant is represented by a forewing, possessing a cubital and dis- coidal cell, and consequently might belong to any one of a number of genera. Holotypc. — No. 6180, Canadian Geological Survey. (Formicidae) antediluvianum (Scudder) Calyptites antediluvianum, Scudder, S. H., 1878. Rep. Progr. Geol. Surv. Can., 1876-77, p. 270. Calyptites antediluvianum, "Wheeler, W. M., 1908. Journ. f. Psych, u. Neurol., 13, p. 417, 22 bulletin: museum of comparative zoology This species, based on one specimen consisting of a forewing, was originally described by Scudder as a braconid, but it is really a formicid with uncertain generic affinities. Holotype. — Canadian Geological Survey. 2. The oil shales at Elko, Nevada, have contributed one fossil ant, which, although poorly preserved and much distorted, is nevertheless described below because it represents a new locality for the family, and even for the insects as a whole. FORMICINAE PsEUDOCAMPONOTUs, gen. nov. Similar to Camponotus but with the eyes and antennal insertions farther forward on the head; antennae consisting of twelve segments in the female. Genotype. — Ps. elkoanus, sp. nov. PsEUDOCAMPONOTUS ELKOANUS, Sp. nOV. Plate 2, fig. 2 Female. — Length, 7.0 mm. Head quadrate, a little longer than broad ; mandibles massive, triangular; clypeus large, the anterior border with a small tooth on each side of a median notch, the posterior margin with a lobe extending back nearly' to the middle of the head; eyes small, situated at about the middle of the sides of the head; scape just reach- ing the posterior margin of the head, slender; funicular segments sub- equal, about as long as broad; thorax about as long and as broad as the head; petiole apparently rather wide; gaster small, only a little longer than the head, rounded. Length of head, L8 mm.; scape, 122 mm.; funiculus, L8 mm. ; thorax, 2.0 mm. ; gaster, 2.4 mm. ; forewing, 6.0 mm. Width of head, L3 mm.; thorax, L2 mm.; gaster, 2.0 mm. Locality. — Near Elko, Nevada, "20 miles or more northeast from the station, from a shaft sunk by the Central Pacific Railroad Com- pany." (S. W. Garman). Holotype.— No. 2940, M. C. Z. The obscurity of the petiole and venation of this species prevents the generic affinities from being accurately determined. The habitus is nearest to that of the Camponotini, with the exception of the position of the eyes and the antennal insertions, so that until additional material carpenter: fossil ants of north AMERICA 23 has been found the species had probably best be assigned to a new genus within this tribe. Professor Wheeler has suggested to me that the pro- longation of the clypeus gives somewhat the appearance of the trun- cated head of the subgenus Colohopsis, and this character would agree with the structure of the mandibles as indicating that the ant was a wood-inhabiting species. The Florissant shales have produced more fossil ants than any of the other deposits, excepting, of course, the Baltic amber. Scudder remarks in his volume on the Tertiary insects of North America that " the ants are the most numerous of all the insects at Florissant, com- prising, perhaps, a fourth of all the specimens; they form more than three-fourths, perhaps four-fifths, of all the Hymenoptera; I have al- ready about four thousand specimens of perhaps fifty species (very likely many more)." Some four thousand additional specimens were obtained by the expeditions conducted after 1900, and Mr. S. A. Roh- wer tells me that a great number of poorly preserved ants were dis- carded at the locality. Still further evidence of the abundance of the ants at the time of the existence of the Florissant biota is afforded by the presence of many specimens of fossil fish excrement, apparently consisting of the "hard, indigestible heads of ants" (Wheeler, 1910). About half of the 12,000 specimens which I have examined are well enough preserved to permit specific determination, and nearly a half of the remainder show details sufficient for generic diagnosis. By far the majority of the specimens are males and females, which are nearly equally represented; only about two per cent, are workers. This scarc- ity of neuters is obviously due to their inability to fly over the lake, for since only a relatively few specimens were blown from the trees or shrubs into the water, they were rarely preserved as fossils. The same deficiency of workers was observed by Heer in his study of the ants in the Oeningen beds, in which "mit einigen wenigen Ausnehmen finden sich nur geflugelte Individuen vor, well die ungeflugelten Thiere, hier also die geschechtslosen Individuen, viel seltener im Wasser verungluchten, als die ersteren." The opposite tendency is naturally found in the Baltic amber fauna, most of which " are workers and be- long to more or less arboreal species, but there are also quite a number of males and females. As nearly all of the latter have wings, they must have been caught in the liquid resin just before or after the nuptial flight." (Wheeler, 1910.) The correlation of the castes of the Florissant ants is very difficult. This is especially so because the females and males of a species do not usually occur even in an approximately equal abundance. The com- 24 bulletin: museum of comparative zoology monest female, for example, is that of Protazteca elongata, sp. nov., while the most abundant male is that of Lasius peritulus (Ckll.). An even better illustration is afforded by Miomyrmc.x impadus (Ckll.) and M. striatus, sp. nov., the former of which is represented by thirty-seven females and one male, and the latter by two females and fifteen males. The determination of the affinities of the males has been exceedingly troublesome, partly because this caste has been so little used in the taxonomy of existing ants, and partly because of the slight generic differentiation of this sex. The number of species of Florissant ants was estimated by Scudder to be close to fifty, but this is considerably more than the actual amount. Thirty-two species are described below and although a few others may turn up in later collections, the total number will probably not exceed forty. These species are distributed among five of the seven recognized subfamilies. The following list shows the abundance of the species, and the table summarizes the ant faunas of the Florissant shales and the Baltic amber, so that they may be easily compared. Name Archiponera wheeleri, sp. nov. Pseudomyrma extincla, sp. nov. Aphaenogaster mayri, sp. nov. Aphaenogaster donisthorpei, sp. nov. Pheidole tertiaria, sp. nov. Messor sculpturatus, sp. nov. Pogonomyrmex fossilis, sp. nov. Lithomyrmex rugosus, sp. nov. Lithomyrmex striatus, sp. nov. Cephalomyrmex rotundatus, sp. nov. Mianeuretus mirabilis, sp. nov. Dolichoderus antiquus, sp. nov. DoUchoderus rohweri, sp. nov. Protazteca elongata, sp. nov. Protazteca qvxidrata, sp. nov. Protazteca capitata, .sp. nov. Liomeiopum miocenicum, sp. nov. Liometopum scudderi, sp. nov. Elaeomyrmex gracilis, sp. nov. Elaeomyrmex coloradensis, sp. nov. Iridomyrmex jlorissantius, sp. nov. Iridomyrmex obscurans, sp. nov. Miomyrmex impactus (Ckll.) Miomyrmex striatus, sp. nov. No. of Sexes known Specimens ^ c? 2 9 2 f^ c^ 9 200 9 1 9 2 9 20 ^ 1 c^ 9 40 ^ 9 2 9 1 9 1 y? 9 10 ^ 9 7 y d^ 9 1,500 S 9 50 y? 9 30 ^ d^ 9 1,500 S 9 200 e 9 50 5 9 29 9 34 9 26 y c? 9 38 c^ 9 17 carpenter: fossil ants of north AMERICA 25 Name Petraeomyrmex minimus, sp. nov. Formica robusta, sp. nov. Formica cockerelU, sp. nov. Formica grandis, sp. nov. Lasius peritulus (Ckll.) Camponotus fuscipennis, sp. nov. Camponotus microcephalus, sp. nov. Camponotus peirifactus, sp. nov. No. of Sexes k nown Specimens ? 11 d' 9 400 & 9 5 9 1