In the question of polymorphism it is worth while to discuss at some length the extreme polymorphism exhibited in the case of some of the colonial animals—first, because the matter has recently occasioned an important controversy between Herbert Spencer and Weismann; and, secondly, because the discussion will serve to make still more clear the difference between my views and those of Weismann upon the nature of the process of development.
Among the colonial insects there arise, in addition to males and females, sexless individuals known as neuters. These in certain cases are very different from both males and females in structure and in social instincts.
Among bees there are the queens, sexually mature females; the workers, females whose sexual organs are rudimentary, and parts of whose bodies—the stings, the wings, the hind legs, with their pollen-collecting apparatus—are peculiarly formed; and, lastly, the males, or drones.
In many of the ant and termite colonies still greater differences exist between the different sets of individuals. In addition to males and females, there are sexless workers, and these is many species are of two kinds, known as workers and soldiers. The divergences of structure among the three or four forms are shown, frequently by considerable differences in size, by the presence and absence of wings, by differences in the sense-organs, the brain, and the structure of the head. In the common ant—Solenopsis fugax, for instance, as Weismann quotes from Forel—the males have more than four hundred facets on their eyes, the females about two hundred, and the workers from six to nine. Many soldiers possess enormously large and heavy heads, with massive jaws, and naturally, with the appropriate muscles much enlarged.
But as workers and soldiers, on account of the rudimentary state of their sexual organs, cannot reproduce themselves, all the three or four kinds of ants in the colony must be developed from eggs deposited by the females. In this Weismann finds the most convincing proof of the omnipotence of natural selection, and, I venture to add, for the omnipotence of his doctrine of determinants.
He says (Contemporary Review, vol. lxiv., p. 313): 'It fortunately happens that there are animal forms which do not reproduce themselves, but are always propagated anew by parents which are unlike them. These animals, which thus cannot transmit anything, have nevertheless varied in the past, have suffered the loss of parts that were useless, and have increased and altered others; and the metamorphoses have at times been very important, demanding the variation of many parts of the body, inasmuch as many parts must adjust themselves so as to be in harmony with them.' 'None of these changes' (p. 318) 'can rest on the transmission of functional variations, as the workers do not at all, or only exceptionally, reproduce. They can thus only have arisen by a selection of the parent ants, dependent on the fact that those parents which produced the best workers had always the best prospect of the persistence of their colony. No other explanation is conceivable, and it is just because no other explanation is conceivable that it is necessary for us to accept the principle of natural selection.'
According to Weismann's conception, 'every part of the body of the ant' (loc. cit., p. 326) 'that is differently formed in the males, females, and workers is represented in the germplasm by three (sometimes four) corresponding determinants; but on the development of an egg never more than one of these attains to value—i.e., gives rise to the part of the body that is represented—and the others remain inactive.' This structure of the germplasm Weismann attributes to the operation of selection. 'For in the ant state' (loc. cit., p. 326) 'the barren individuals or organs are metamorphosed only by the selection of the germplasm, from which the whole state proceeds. In respect of selection, the whole state behaves as a single animal. The state is selected, not the single individuals, and the various forms behave exactly like the parts of one individual in the course of ordinary selection.'
Naturally, from the views on the germplasm theory and on the doctrine of determinants that I have expressed in this book, I cannot accept the explanation Weismann thus gives of the facts. It is true that Weismann holds his own explanation to be the only conceivable explanation. 'For there are only two possible a priori explanations of adaptations for the naturalist, namely, the transmission of functional variations and natural selection' (loc. cit., p. 336); 'but as the first of these can be excluded' (on account of the infertility of workers and soldiers), 'only the second remains.'
But are the alternatives really only as Weismann suggests? Is there no choice left for the naturalist?
When I was reading his All-sufficiency of Natural Selection, kindly sent me by the author, it came into my mind that I could not accept his dilemma. For the different individuals in the insect states may be explained in a third way—in a way overlooked by Weismann. This third explanation is nothing more than the subject of all this treatise of mine. It is that, in obedience to different external influences, the same rudiments may give rise to different adult structures.