28 INVERTEBRATA CHAP. 



or two groups of animals the immature eggs destined to destruction 

 are supplied to the zygote or fertilized ovum to be used as food. 

 This leads to the strangest modifications of the early stages of 

 development. The cells which result from the first divisions of the 

 zygote may actually separate from one another and come together 

 again in such a way as to surround the follicle cells, and this has 

 led to the statement that, in certain cases, the egg dies and the 

 embryo is developed out of follicle cells, but there is apparently no 

 justification for this statement (cf. p. 635). 



But an embryonic stage may be, so to speak, intercalated between 

 two larval stages. In the history of the race the change of habits 

 which is recapitulated in the life-history, must have been con- 

 tinuous, for no animal ever suddenly changed from one mode of 

 life to another. Now the dangers incident to larval life and the 

 opportunities of obtaining food, may vary very much, and will be 

 much greater in some stages than others. If in one stage a large 

 store of nourishment can be accumulated, it will be an advantage to 

 the animal to pass quickly over the next stage, which is probably less 

 favourable, and so we may get these intercalated embryonic, or, as 

 they are usually termed, pupal stages. During these the animal is 

 sometimes as quiescent as a true embryo, as in most insects ; in others, 

 such as Cirripedia and Holothuroidea, it is active but takes no food. 



But there is one outstanding feature about most larvae which 

 strikes the observer, and that is their extremely small size compared 

 with that of the adult into which they eventually develop. This 

 reduction in size is in all probability a secondary modification, but it 

 has led to other modifications. An alteration of size produces an 

 alteration in the physiological relations of the organism, and we 

 find that where, from other evidence, we have reason to suspect that 

 the ancestor had a long series of similar organs, the larva may only 

 show one or two ; for all these organs, if reduced to the same scale 

 as that to which the whole body of the animal has been diminished, 

 would become physiologically ineffective. Take, for instance, the 

 gill slits in the larva of Amphioxus. These must have a certain 

 minimum size if they are to work, on account of the viscosity of water, 

 and therefore whilst they remain larger in proportion than the other 

 organs of the body, their number becomes diminished, and so where 

 the ancestor had almost certainly two rows of such slits, we find 

 them represented in the larva by one row of slits which occupy the 

 whole ventral surface. 



Lastly, it may be remarked that whereas it is true, generally 

 speaking, that the more primitive features an adult exhibits, the 

 more primitive features are found in the larva, yet the change from 

 the larval to the embryonic method of development seems to take 

 place quite independently of the status of the adult, and some 

 animals preserving very primitive features have a development 

 almost completely embryonic, whilst others higher in the scale retain 

 a long larval history. 



