the ancestral line to meet certain special requirements of the situation. It can never return to that line. The mere fact that an ancestor once existed, with certain characteristics, has in itself no controlling force upon the development of the embryo.
The secondary adaptations of larval forms have the same bearing upon development as have peculiar ancestral conditions. They become characteristic steps in the line of development to maturity. The sexually mature animal has passed through them all in its growth from the germ, and conditions of the same character are implanted in its own germs, and must unfold in their development. There is no longer an exact phylogenetic line. Many of the ancestral stages have become greatly modified. To the new developing animals those modified stages of growth are ancestral stages so far as it individually is concerned. Development follows this new line, although it may have become a strangely warped and irregular one, and though at certain stages of growth it may yield peculiar organs or tissues which are discarded as useless, or consumed as nutriment, at later stages. The true line of growth in such cases is restricted to the more deep-lying and important parts of the organism, and though, at certain stages of growth, forces appear which produce a special growth of secondary tissue, this is reabsorbed or discarded when the development is resumed. Marked instances of such discarded tissue are seen in the pupal development of certain insects, and in the case of the star-fish development above referred to.
We have paid some little attention to the characteristics of larval growth for two reasons. Their true bearing on the mystery of evolution has been little attended to, and the above-given hypothesis of explanation has not heretofore been offered, so far as the writer is aware. The second reason is that they bear a much closer relation to the phenomenon of neuter insects than might at first sight appear. The neuter insect has not as yet been looked upon as a resting-stage in the line of full development, and as analogous to the lower stages of larval growth. It has, indeed, a peculiarity of its own, that it fails to attain full development. And as its secondary characteristics are not participated in by the sexually mature form, but have arisen by adaptation which is still operative, the fact of their transmission becomes difficult to understand. Yet we think it may be shown to be but an extension of the principle above considered.
It is a significant fact that a neuter worker class is found only in those animal tribes in which the social principle has reached its highest development, such as the bees, ants, and termites among insects, and the hydroid polyps in the other sub-kingdoms of life. In each of these communal types of life there has been a division of duties, the work of reproduction being confined to one or a few members of the community, at least so far as maternity is concerned, while the other members have gained special adaptations to other duties. In bee com-
