500 PROCEEDINGS OF THE AMERICAN ACADEMY. 



But the morphological is not the only point of view from which to 

 regard this question. As has already been said (p. 429), polymorphism 

 has been commonly attributed to a division of labor, and if this is so, 

 it presupposes a difference of instinct among otherwise similar workers 

 before structural differences arose. Such a physiological difference 

 seems to be illustrated by the experiments of Lubbock on monomorphic 

 forms, which have already been cited; Wheeler (:10, p. 370), indeed, 

 states that such a division of labor probably exists in various forms 

 and degrees among all ants with monomorphic workers. Also, Figure 

 19 (p. 4G4), based on the conditions in a graded series, seems to indi- 

 cate something of the same kind, for here there seems to be a difference 

 in activities not yet entirely correlated with size. 



lleichenbach (:08) maintains that the fundaments of all the instincts 

 must be present in each ant egg ; if the offspring is a male, the female 

 instincts do not mature. Wheeler (:06'') states that most, if not 

 all, of the characters of the worker are quantitatively, but not quali- 

 tatively, different from those of the queen ; i. e. the worker does not 

 differ from the queen as a mutant, but as a fluctuating variant, 

 which has been produced by imperfect feeding in the larval stages. 

 The same idea is again suggested by him (Wheeler, :06'') when he says 

 that the queen is the epitome of the instincts of the ant colony, etc. 

 According to the view of Weismann, that there must be " determinants " 

 to cause the different castes of workers, it would be difficult to account 

 for the fact that form and function are not more closely correlated, 

 even in species like Pheidole pllifera, where morphologically the classes 

 are very distinct. If, on the other hand, we substitute for " determin- 

 ants "a different potentiality of the same germ-plasm to develop into any 

 one of several forms, according to more or less favorable conditions, we 

 should expect, what we have seen to be the case, that while there is a 

 greater tendency for each class to perform certain duties, there is an 

 overlapping of the functions. Each individual seems to have poten- 

 tially an instinct to perform the duties even of those classes in which, 

 owing to structural differences, we should not expect it to take part. 

 As it is with the queen (when the workers which normally perform 

 certain duties are present, she does not perform those duties), so it is, 

 to a less extent, with the workers ; for though they do not take an equal 

 part in the work to which another class is especially adapted, they do 

 share in it to a certain extent, unless physically incapacitated. We 

 may say, then, that while there are some differences in the duties of 

 the different classes of workers and of the queen, all female ants have 

 an instinct — which may become more or less latent — to perform all 

 the duties of the colony j while the male, coming from an egg which is 



