15<J ANTS. 



structure of the workers, for these have a pronounced tendency to 

 assume the characters of the queen. Miss Holiday (1904) and 1 have 

 shown that a lar-r percentage of the workers possess ocelli and even 

 resemhle the female in size, in the structure of the thorax, and in pos- 

 sessing a \vell-ile\-elnpc-d receptaculum seminis and ovaries. Hence we 

 may say that the abundant food supply has a tendency to increase the 

 fertility of the workers and to reduce the dimorphism of the female 

 sex. If this tendency, which is the reverse of that observed under 

 similar circumstances among lestobiotic ants like Carebara, should con- 

 tinue, it would probably give rise to one of two conditions in the course 

 of further evolution: either the worker caste would disappear com- 

 pletely, leaving the species to be represented by males and winged 

 females, or the winged female would be suppressed, leaving only males 

 and ergatoid females. As winged females seem to be produced by 

 Lcptothora.v emersoni in rather small numbers, I am inclined to believe 

 that evolution will take the latter course. In the ants to be considered 

 under the head of permanent social parasitism evolution has moved 

 along the other path and led to a complete extinction of the worker caste. 



10. Leptothorax glacialis. This ant, which I have described 

 ( IQOJ/J ) as a subspecies of L. emersoni, occurs in the Rocky Moun- 

 tains of Colorado at elevations of over 2,500 m. It lives with Myrmica 

 alpina, a western variety of M. brevinodis, in relations very similar to 

 those described above for the two eastern forms, except that glacialis 

 seems to feed less on the surface secretions and more on the regur- 

 gitated food of its host. Observations on a colony in an artificial nest 

 also seemed to show that the instinct to feed independently is more 

 blunted or vestigial in glacialis than it is in emersoni. 



The relations of L. emersoni and glacialis to their respective hosts 

 represent the nearest approach to the formation of mixed colonies that 

 has been observed among xenobiotic ants. Indeed, so cordial and inti- 

 mate are the relations between these species and their hosts that even 

 the instinct to rear their broods in separate apartments can be sup- 

 pressed experimentally and the insects induced to form a true mixed 

 colony. If in a state of nature L. emersoni and glacialis should 

 develop a habit of mingling their eggs, larvae and pupae with those of 

 the Myrmica, these inquilines would probably soon cease to excavate 

 nests of their own and become permanent social parasites. 



