June 18, 1915] 



SCIENCE 



907 



time of Wheeler's studies, 24 genera and 52 

 species were known from Baltic amber ; lie 

 now adds 21 genera and 40 species, in addition 

 to revising those already known. 



No ants are known from the Mesozoic; one 

 or two recorded as such prove to belong to 

 quite a different group of Hymenoptera. The 

 oldest fossil ants, by far, are those described 

 by Scudder from the Eocene of Green River 

 and White River in Wyoming and Colorado. 

 These, unfortunately, are poorly preserved, and 

 afford very little information. The beds 

 along White River near the Colorado-Utah 

 boundary are certainly Eocene and not Oligo- 

 cen-e, as has repeatedly been stated, apparently 

 from confusion with the White River group 

 of Oligocene beds from which mammals are 

 obtained. Further collections from the insect- 

 bearing Eocene rocks of the west are much to 

 be desired, as well as a more complete exami- 

 nation of those already obtained, for there is 

 a chance to discover very important entomo- 

 logical facts. In Europe, the Lower Oligocene 

 contains the earliest ants, but includes the 

 Baltic amber, as well as the beds at Aix in 

 Erance, and probably the Gurnet Bay deposit 

 in the Isle of Wight. This last, from which I 

 have recently described a number of ants 

 (Doliclioderus, Leptothorax, (Ecophylla and 

 Ponera) is perhaps later than the amber. The 

 Middle and Upper Oligocene and all three di- 

 visions of the Miocene (the latter including 

 the important localities CEningen in Baden, 

 Radoboj in Croatia and Sicilian amber) have 

 afforded fossil ants in Europe, and there is a 

 single Italian locality assigned to the Lower 

 Pliocene. Between the last and the Pleisto- 

 cene is a blank. The supposed Miocene record 

 from Spitzbergen is to be deleted, the speci- 

 men being the abdomen of some insect, and 

 wholly unrecognizable. 



Thus it appears that our first real knowl- 

 edge of fossil ants begins with the amber, 

 probably at least two million years ago. What 

 development has the group shown in all this 

 long time? To what extent are the remark- 

 able habits and structures of modern ants 

 products of recent evolution? From Dr. 

 Wheeler's researches we gather these facts : 



1. Of the amber genera, over 55 per cent, 

 are still living; that is to say, 24 genera, of 

 which four are at present cosmopolitan; four 

 universal in the tropics, but invading semi- 

 tropical or temperate regions with some of 

 their species; four essentially paleotropieal; 

 five belonging to an Indomalayan and Aus- 

 tralian series; six circumpolar; and one 

 (Erehomyrma) known to-day by two species, 

 one in Texas, the other in Peru. 



2. The extinct genera are mostly allied to 

 paleotropieal forms. There is, however, little 

 afiinity with the African fauna. 



3. It is by no means certain that the amber 

 fauna all belongs strictly to the same time or 

 general locality; yet ten cases are recorded in 

 which two species of ants exist in the same 

 block of amber, proving their strict contem- 

 poraneity. 



4. Since the amber, " the family has not 

 only failed to exhibit any considerable taxo- 

 nomic or ethological progress, but has instead 

 suffered a great decline in the number of spe- 

 cies and therefore also in the variety of its 

 instincts, at least in Europe." Already, in 

 the Lower Oligocene, the subfamilies and mod- 

 ern genera were established ; even some of the 

 species were almost identical with those of 

 to-day. Formica flori of amber is almost ex- 

 actly the modern F. fusca; other species of 

 Formica represent different subdivisions of 

 the genus, quite as we have them to-day, 

 though' there is no representative of F. san- 

 gvinea. Other amber ants show similar re- 

 semblances. Not only was polymorphism fully 

 established, but the larval and pupal stages 

 show such peculiarities as we see to-day; thus 

 the larvffi of Prenolepis had already lost the 

 cocoon-spinning instinct. The amber Irido- 

 myrmex pupae were likewise naked, just as 

 they are now. Specimens of Lasius carry 

 gamasid mites, showing that these arachnids 

 had already developed their specialized myr- 

 mecophilous characters. So also, aphids were 

 kept by ants in those days. 



5. Perhaps it would hardly be going too far 

 to say that if the ants of to-day were likewise 

 preserved in amber, and were submitted to a 

 future entomologist along with those of the 



