166 Chapter IV. 



or at least in neighboring nests. The number of 

 colonies, in which I found Lomechusas (lOo), is more 

 than three times as great as that of nests containing 

 pseudogynes (33) ; these are the centers, from which 

 the Lomechusas gradually spread to the neighboring 

 nests, where by and by they cause the birth of pseu- 

 dogynes. It is scarcely possible, that the presence of 

 these beetles and of their larvae, which are fed by the 

 workers, should have a modifying influence on the 

 ovaries of the queens, but, probably they do so on 

 the nursing instincts of the workers. My recent 

 observations and experiments until 1904 have con- 

 firmed this solution of the interesting problem. 



Some of the above mentioned intermediate ant- 

 forms are useful for the preservation of the colony 

 and the species, whilst others are more or less indif- 

 ferent, and still others positively hurtful, being prob- 

 ably pathological deformities. The rearing of worker- 

 like, wingless queens among the Amazon ants (Polyer- 

 gvis rufescens), for instance, is very appropriate, 

 because their colonies are rather rare and far distant 

 from one another. Therefore, the probability is very 

 slight, that on their nuptial flight the winged sexes 

 will meet with those of other colonies; and besides, 

 this species has to encounter exceptional difficulties in 

 founding new settlements by means of single fertilized 

 females, since the Amazons are entirely dependent on 

 the help of their slaves. The wingless queens, on the 

 other hand, cannot go far from their nests ; and after 

 they have been impregnated by the winged males, 

 some strolling slaves can easily find them in the 

 neighborhood and bring them home again; hence it 



