by later hands, and at times so cunningly as to defy detection.

The chief disturbing cause arises from the necessity of supplying the embryo with nutriment. This acts in two ways. If the amount of nutritive material in the egg is small, then the young animal must hatch early, and in a condition in which it is able to obtain food for itself. In such cases there is of necessity a long period of larval life during which natural selection may act so as to introduce modifications of the ancestral history, or spurious additions to the text. If, on the other hand, the egg contains a considerable quantity of nutrient matter, then the period of hatching can be postponed until this has been used up. The consequence is that the embyro hatches at a much later stage in its development, and if the amount of food material, or food yolk as it is called, is enough, may even leave the egg in the parent form.

This varying condition, as regards amount of food-yolk, affects recapitulation-i.e., the tendency of the embryo to pass through the ancestral stages-in two principal directions. If there is very much foodyolk, there is a tendency for the embryo to shorten its development by the omission of certain of the ancestral stages, and especially by the suppression of characters which, though functional in the ancestors, are of no use in the adult state of the animal itself. Thus tadpoles, after hatching, breathe for a time by gills: this gill-breathing condition being an ancestral one for all Vertebrates. In the West Indies there is a little frog (Hylodes) which lays its eggs, not in water, but on the leaves of plants. These eggs are larger than those of the common frog-i.e., contain more food-yolk-and the young embryo is thereby enabled, just like the lobster or crayfish, to develop to a later stage before hatching : it passes through the tadpole stage within the egg, and hatches, like the crayfish, in the form of its parent. Although it passes through a gill-cleft stage no gills are developed; being of no use to the embryo, it would be a sheer waste of time to form them, and so they have dropped out of the ontogeny or individual development. (Figs. 15, 16.)

Exactly the same thing has happened in reptiles, birds, and mammals, in which gill-clefts are found, but gills are not developed. A similar tendency to the omission or blotting out of useless characters is seen in the development of all forms which have sufficient food-yolk to carry them over the stages at which these characters would be of functional value before the time of hatching.

In the case of embryos developed from small eggs there is a tendency to distortion of the ancestral history of a very different nature. Such embryos, owing to the small supply of food-yolk in the egg, have to hatch very quickly-i.e., not merely of small size, but in a condition representing a very remote ancestral stage. The intervening stages between the early condition and the adult one have to be repeated while the larva is enjoying a free existence: this process will necessarily be slow, for the larva has not merely to develop, but has to obtain for itself food, at the expense of which the further development

Stages in the development of the common Frog (Rana temporaria). A.-The egg.

×3

B. The Tadpole at the time of hatching. The mouth is not yet formed, the Tadpole being still dependent on the food-yolk present in its body. Two pairs of external gills are present as branched fingerlike processes at the sides of the neck. Below and in front of these is the horse-shoe shaped sucker by which the Tadpole fixes itself, and at the sides of the front of the head the rudiments of the nose and eye

×3

are seen.

C.-A Tadpole shortly after the time of appearance of the limbs: the hind limbs are seen at the junction of the body and tail: the fore limbs are present, but are concealed by the opercular folds covering the gills. At this stage the Tadpole breathes by both gills and lungs.

XI

D.-A young Frog with the tail only partially absorbed.

XI

Stages in the development of the West Indian Frog (Hylodes) : illustrating the effect of increased amount of food-yolk in causing the omission of ancestral stages. The free living Tadpole stage of the common Frog is entirely suppressed, and no gills are ever formed. The entire development from the laying of the eggs to the hatching of the Frogs occupies from a fortnight to three weeks.

A.--The larva at the end of the first week. The head, eyes, stumps of the limbs, and the long tail are well shown. The food-yolk is contained within the large yolk-sac in the middle of the figure.

×3

B.-The young Hylodes shortly before hatching. There is still a very large tail present, which is believed to be used as a respiratory organ.

×3

C.-Young Hylodes at the moment of emerging from the egg: a short

stump of a tail is still present.

×3

D.-Young Hylodes at the end of the first day. The tail is completely

absorbed, and the Frog has already the form of an adult.

×3

may be effected. Hence such larvæ may take weeks, months, or even years in recapitulating these later stages, which their larger-egged allies get through in a few days. Moreover, during the whole of this time they are exposed to competition, both amongst themselves and with other animals; they have to obtain food for themselves, and they are liable to be themselves devoured as food by other animals.

Owing to this competition, and to the length of time during which it lasts, these larva are liable to acquire, through natural selection, characters which are connected with their existence as larvæ, but which form no part of the ancestral history; characters which will aid them in obtaining food, or in escaping from their enemies, but which were not found in any of the ancestors of the species. Of such secondary larval characters the long spines with which the Pluteus larva of sea-urchins are provided are good examples; so also are the enormous spines on the young larvæ of crabs and other crustacea. Other excellent illustrations are afforded by the developmental history of many fresh-water forms, in which, from the danger of their being swept down by the currents of the rivers or streams in which they dwell, or to obtain protection from the cold of winter, special characters are often acquired. The glochidium larva of the fresh-water mussel is a good instance of the former, and the specially-protected statoblasts or winter buds of Polyzoa and sponges, of the latter specially-acquired character.

It is not easy to distinguish between these later