1899] PRINCIPLES OF ANLMAL DEVELOPMENT 133 



The number is so constant in the whelk, that one may suspect 

 that it is governed by some law. Probably, too, if one counted them, 

 the number destined to be devoured would be found to be equally 

 constant. The whole process must take place under the workings of 

 a law or laws, and it is part of our task as embryologists to try to find 

 out what these are and what determines them. 



If, after the examination of the whelk-cases, a fully-formed egg- 

 purse be taken from the oviduct of a dog-fish or skate, on opening this 

 we shall almost certainly find but a single egg. 



The eggs of different species of dog-fish or skate also differ con- 

 siderably in size, both as regards the egg-cases and the yolks. These 

 differences are not always in relation to the specific differences in size. 

 The smooth skate (B. batis) is much larger than the starry ray (B. radiata), 

 but no proportion can be detected between the sizes of the two skate 

 and those of their eggs. Indeed, if we study the ripe eggs of two 

 closely allied species of shark, Mustelus vulgaris and M. laevis, we find 

 that, although the two fishes are almost of identical size and so alike 

 that no fisherman could ever be expected to distinguish between them, 

 their eggs exhibit great differences in size along with almost inappreci- 

 able differences in the texture of the egg-capsule. 



What, then, are the essential differences between the eggs of these 

 two species ? The egg-shell of M. vulgaris appears to be rather 

 thicker than that of M. laevis, and — a very important point — the 

 actual egg, the yolk, of the former is the equivalent in size and weight 

 of about four of the latter. 



This difference cannot be ascribed to the relative sizes of the two 

 species, and, as a matter of fact, it is in association with a very funda- 

 mental difference in the mode of development. 



Both species are viviparous, and it is possible that the young are 

 born in both cases in the same condition of development, though this 

 has yet to be determined. In M. vulgaris the yolk suffices for the 

 whole of the uterine development, whereas in M. laevis it is used up 

 long before uterine life is over, and for what is probably a long portion 

 of its uterine existence the young M. laevis is nourished by a sort 

 of placental attachment of its yolk-sac to the uterine wall, as recorded 

 by Aristotle, and as rediscovered during the present century by that 

 great embryologist, Johannes Muller. 



The instances above mentioned may serve as the text for further 

 study. Modern, embryology has never yet seen any difficulties or 

 even any problems in the matter at all. We have our classifications 

 of various kinds of eggs. We recognise eggs with little or no 

 food-yolk, and those with much. Various forms of segmentation or 

 cleavage of the egg are distinguished, such as equal or adequal, un- 

 equal but complete, and meroblastic or partial. Our leading text- 

 books either say, or lead the reader to infer, that the differences 

 between the various forms of cleavage are dependent upon the amount 



