262 THE APODIDtE part ii 



uneven development is not much to ask, and if it 

 occurred as described, it could hardly fail to lead to a 

 bending of the head-shield along the dorsal middle 

 line, every time, for instance, a larva sought to put 

 in practice its inherited tendency of contracting its 

 muscles for the purpose of rolling up. The failure 

 to develop a head-shield stiff enough to counteract the 

 pulls of muscles lying in the transverse plane, may 

 have thus led to the conversion of the head-shield into 

 the bivalve shells, which have, in the long run, proved 

 a better defence than rolling up. 



We thus explain the rudimentary state of the 

 abdomen and trunk. It was only in comparatively 

 young animals in which but few trunk segments 

 had been developed, that the bending was likely to 

 take place, and, when once acquired, it would be 

 clearly an advantage to keep the abdomen in a larval 

 stage, in order that it might be quite enclosed within 

 the halves of the head-shield. 



We therefore suggest that the Ostracoda have had 

 more than one root, and may in fact be derived from 

 the larvae of any of the primitive Crustacea with 

 large head-shields, whether Trilobites or Phyllopods. 

 There seems to be some evidence for both these 



COPEPODA. 



The origin of this very rich group of Crustacea 

 is very obscure. The general opinion is that they 

 must be ranked as perhaps the lowest of all the class. 

 We have now to try to suggest a possible origin for 



