consideration as to many others, or his remarkable power of analysis could not have failed to perceive other important bearings of the subject. A full review of the phenomena of larval development certainly seems to remove the mystery of the neuter ant and bee from the position of an anomaly to that of an ordinary method of structural unfoldment. If the appearance of sexual organs and powers is the final step to maturity, then all neuters are larval forms, although in every other respect their development may be complete; and they are subject to the same modifying influences as are all larvæ. It is one of the most common conditions of invertebrate life-development for the unfolding offspring to stop at certain stages of growth, and devote itself for a while to nutrition, ere resuming its course of structural development. Such "resting-stages" are those in which there exist specially favorable conditions of nutrition, or of adaptation of the larval form to the conditions of the food-supply. The most notable instances are those seen in the extraordinary larval forms of some of the Echinodermata, and the little less remarkable larval structure of some of the insects and crustaceans. In certain cases several successive larval forms, each deviating considerably from the normal type of the animal, appear.
Yet these peculiarities of structure have never yet been advanced as stumbling-blocks in the way of natural selection. The caterpillar, for instance, while resembling the moth or butterfly in its more deep lying peculiarities, displays remarkable external deviations, and assumes organs and instincts still more anomalous than those shown by neuter ants. The larval star-fish presents an instance of still stranger anomaly. Only the stomachal region and its immediate surroundings pertain to the type, and all the rest of the structure is accessory. When the development of the star-fish is resumed the new form grows out of this internal region of the body of the bipinnaria, or larval form, whose external parts are discarded as useless, or absorbed as food by the new creature. This is the most aberrant instance of such temporary development known. No trace of the star-fish type can be perceived in its larva. It doubtless exists, but is quite masked by secondary formations. Or it may be that this larva represents an ancestral form of the star-fish, as divergent in character as is the crustacean larva of the barnacle from the mature form.
Yet this explanation of atavism, or temporary check to development at an ancestral form stage, only partly meets the difficulties of the case. There is an unquestionable new adaptation to new circumstances to be explained. Natural selection acts upon all forms which give it sufficient opportunity, without regard to whether they are larval or mature. Let us take for an instance the case of the butterfly. Here the development does not proceed continuously, from the germ to the mature form, as in some insects, but is checked for a considerable period at the caterpillar stage. The active nutrition at this stage seems to act as a check upon development, so that the caterpillar is a form
