30 CHILDHOOD OF ANIMALS 



a long spine, and the abdomen is very long, almost devoid of appen- 

 dages along the greater part of its length, but with a large pair on the 

 second last segment. After several moults, with further slight 

 changes, a larva appears which is called the mysis stage or schizopod 

 stage (Fig. 14), from its resemblance with the adult form of a lower 

 kind of crustacean. In this stage the projecting spine of the carapace 

 is very long, the abdomen has a complete set of swimming limbs, 

 those of the last pair being large and forming with the last segment 

 itself a swimming tail-fan like that of an adult lobster or prawn. 

 In a further set of moults the complete shape of the adult is 

 acquired by the body and limbs. 



In most of the higher Crustacea, the number of moults is smaller, 

 and there are bigger jumps between the successive types of larvse. 

 The earliest larva of crabs is a fully formed zoea, which is dis- 

 tinguished from the zoea of other Crustacea by a very long spine 

 on the carapace, but, almost immediately after hatching, a thin 

 cuticle is cast off, and this differs from the zoea itself and appears to 

 be the last remnant of one of the suppressed larval stages. Next 

 come a set of larvse called the megalopa stages; which quickly acquire 

 the appendages and general form of the adult crab, but which have 

 a long extended abdomen. After the moult from which an animal 

 that can first be called a crab appears, the abdomen is tucked up 

 under the body as a rudimentary triangular flap. 



Study of the larval development of a very large number of marine 

 crustaceans, of which I have chosen only a few examples, would 

 seem to give a clear picture of the general course of events. Because 

 they have a hard, sheU-like skin, young crustaceans cannot grow 

 larger in the usual way of soft -skinned animals. They must grow 

 in size by a succession of moults. This makes it impossible for the 

 youthful period to be a time of slow and continuous change, from 

 the first larva to the adult. The changes must take place by jumps. 

 Where there are a great many different successive larvse, each a 

 little more complicated than its predecessor, we seem to see the 

 simplest method of arriving at the result, and the greatest probability 

 that the larval history is at least partly a repetition of the ancestral 

 history. And the facts that many of these larvse are closely alike, 

 although they belong to different groups of Crustacea, and that the 

 larvae of the higher groups not infrequently resemble the adults of 

 the lower groups, greatly increase the probability of this ancestral 

 interpretation being correct. 



Insects, like Crustacea, are Arthropods with a hard external 



