The Ants of the Baltic Ainb'.r. 21 



myrmex constrictiis (7595/309 [Fig. 42J). Mayr, who discovered this 

 singular specimen, regarded it as a gynandromorph, but I believe 

 that it is an ergatomorphic male of the extreme type, such as is found 

 in a few recent ants, notably in males of Formicoxenus nitidulus and 

 Ponera punctatissima, which have the head much more like that of 

 the worker than in many ergatomorphic males of the genera Cardio- 

 condyla, Symmyrmica and Technomyrwex. 



The larval and pupal stages of the Baltic ants were also in all 

 respects as highly specialized and of the same structure as those of 

 recent species. I have seen larvae and pupse of Iridomynnex geinitzi, 

 I. goejjperti and Lasius schiefferdeckeri The Lasiiis pupae are enclosed 

 in cocoons, while those of I. geinitzi are naked, showing that the 

 cocoon-spinning habit of the larvae had been lost in the DoliclioderincB 

 as far back as the early Tertiary. This is of" considerable interest, 

 because it has been inferred from the occasional occurrence of both 

 naked and enclosed pupae in the same colony of certain species of 

 Formica (F. fusca, etc.^ that the loss of the cocoon is a mutation, or 

 saltatory variation of recent origin. This may, of course, be true in 

 Formica and some other Camponotine genera, but it is quite as pro- 

 bable, in view of the extraordinary persistence of small characters 

 displayed in the preceding paragraphs, that the pupae of F. flori may 

 have shown the same presence or absence of the cocoon in the same 

 colony as is shown by the modern F. fusca. 



There are also unmistakable indications that the habits and in- 

 stincts of the amber ants were nearly if not quite as advanced as those 

 of existing forms. The method of their preservation and the close 

 affinities of most of the species with modern arboreal forms have 

 already been considered. That many of them had learned to attend 

 plant-lice and had therefore become ,,trophobiotic" is shown by a block 

 of amber in the Konigsberg Coll. containing a number of workers of 

 Iridomyrmex gcepperti together with a lot of their Aphid wards. That 

 the amber ants kept myrmecophiles in their nests can scarcely be 

 doubted, for at least three genera of Paiissidcc iCerapierus, Pleuroptcrus 

 and an undescribed genus) are cited by Klebs in his list of amber 

 Coleoptera^j. That these ants also had Acarine parasites is shown by 

 two workers of Lasius schiefferdeckeri in the Konigsberg Coll., each 

 bearing a mite attached to the base of one of the hind tibiae (Fig. 58). 



^) Ueber Bernsteineinpchliisse im allgeraeinen und die Colcoptcrcn nicincr Bcrn- 

 Bteinsammlung. Schrift. Physik.-okouom. Gcsellsch. Konigsberg LI, 11)10. 2, pp. 217 

 bia 242. 



