active life and abundant nutrition. Insect larvæ, for instance, simply grow during their active feeding-stage. New development only begins during the inactive pupal stage, in which the tissue formed during the larval stage is modified and transformed. After the insect becomes again active, as the imago, no further development of special importance takes place; and it would appear that, if the larval stage is not allowed its full period or its complete course of nutrition, the pupal development is checked at an imperfect stage, and the imago remains immature.
Such is evidently the case in bee communities. The division of the community into males, queen, and workers seems less an operation of natural selection than of intelligent selection. It is a matter of choice among the workers whether any female larva shall develop into a worker or a queen. By giving more room for growth, and more and better food, they can produce a queen from any female larva chosen at will. By contracting the growth-space and diminishing the food, the power of development is checked, and the insect, in its pupal stage, becomes incapable of developing sexual organs and powers.
Thus in every female larva it seems evident that innate powers to become either queen or worker exist. The queen is the higher phase of development, but in attaining this stage the worker stage must be passed through. Why does it not become apparent? This is not difficult to understand, since a similar phenomenon is of very common occurrence. It is simply slurred over in a rapid course of development. The sexual organs begin to unfold, and in so doing exhaust the nutriment and the life-energy which would be needed for the full unfoldment of the worker organs. Thus the superior force checks the inferior, and the innate tendency to develop into a worker is overcome by the activity of a more energetic innate tendency. Where the latter remains aborted the worker tissues fully develop, and with them the worker instincts, since every stage of structural development seems accompanied by its peculiar instincts, as if tissue dominated instinct.
In the case of the ant we have closely similar phenomena. Here there is no satisfactory evidence of intelligent selection, though many observers believe that it exists. So far as we know, however, chance decides whether the larva shall have food enough to carry it to one or other of the worker stages, or to the queen stage. Thus numerous individuals of each stage appear. But the two or more worker castes are not completely separated, since intermediate forms exist, sufficient to make a line of insensible gradation from one form to the other or others. Here, then, we have a complete line of development, reaching from the germ to the queen, but checked at various stages, in which nutrition becomes active and secondary adaptations appear. These secondary adaptive features have undoubtedly become part of the direct line of structural unfoldment. But, as soon as a higher phase of structure begins to unfold, these lower conditions of tissue
