REVIEWS. 115 



the others may each have had a more limited distribution in space and 

 time. 



Wheeler is on the whole of the opinion that the tropical and boreal 

 components of the amber ant fauna belonged to different periods of the 

 Oligocene. 



A certain number of amber ants resemble very closely species living 

 to-day ; and this resemblance, if it does not actually amount to identity, 

 implies almost a lineal descent of the latter from the former. These 

 are : — ■ 



Extinct. Extant. 



Ponera ataria Mayr, and P. coarctata Latr. 

 DolichoderiiH tertiariua Mayr, and D. 4:-punctatu.'i L. 

 Prenolepis hejischei Mayr, and P. nitens Mayr. 

 Lasius schieff'erdeckeri Mayr, and L. niiier L. 

 L. nemorivag us Wheeler, and L. uinbratiix Nyl. 

 Formica flori Mayr, and F. fitsca, L. 

 F'. horrida Wheeler, and F. cinerea Mayr. 

 F. phaethusa Wheeler, and F. truncicola Nyl. 



Some amber ants are more generalised and primitive in their struc- 

 ture than their nearest modern allies. Pvionoimjrmex may be men- 

 tioned as being more primitive than the allied Mj/niiecia, the Australian 

 "Bull-dog ants," which are the most primitive of all modern species. 

 The amber species of the genus (Jecophylla are also somewhat more 

 primitive than the modern O. iimarcujdina of the Old World tropics. 

 Wheeler records 51 specimens (of two species of Oecophylla) from the 

 amber, all of which are workers, with the exception of one male. I 

 may mention in passing that a very large number of the Isle of Wight 

 fossil ants belong to species of Oecophylla, the majority of which, how- 

 ever, are males and females. 



Wheeler refers to other cases of archaic types, but as a rule the 

 amber ants are as highly specialised as existing forms. It is not un- 

 likely that living species of any of the extinct genera may turn up in 

 little explored portions of the Old World Tropics. Indeed, such was 

 the case when a living species of Gesoviyniie.c was found in Borneo 

 years after this genus had been discovered in the amber. 



The various castes or phases are also as sharply differentiated, and 

 in the same manner as in the recent forms, and polymorphism has been 

 developed not only in the worker caste, but also in the male and female 

 forms. Major and minor workers occur in several amber species, and 

 though it is true no "soldiers" have yet been detected, it is neverthe- 

 less probable that they did exist. 



Females of r^latythyrea primaeva and Bradoponera meieri belong to 

 the ergatoid or apterous type. Emery has described a pseudogynic 

 female of Caiiiponotus )nengei, and Wheeler discovered two pseudogynes 

 of Prenolepis hemchei. Mayr regarded a specimen of Iridomyrme.c 

 constrictiis as a gynandromorph, but Wheeler, who figures this insect, 

 is satisfied that it is an ergatomorphic male of the extreme type, with 

 the head more like the worker, such as exist to-day in our British 

 Ponera pitnctatissima and F^orwiGo.cemis mtididus. 



The larvf© and pupae of the Baltic amber ants were also as highly 

 specialised, and similar in form to those of the modern species. The 

 pupffi of Lasius we find in cocoons, and those of Iridouiyriuex are naked, 

 the latter fact showing that the DoUchoderinae had lost the cocooW' 

 spinning habit as far back as the early Tertiary, 



