CRANGON VULGARIS. 31 



cavity. It becomes divided at an early date into a dorsal 

 genital cavity and into a ventral series of paired nephridial 

 cavities, and throughout life thenephridia have no connec- 

 tion with the general body cavity, but are covered with a 

 thin membrane, the dorsal ccelomic spaces never extending 

 beyond the expanded inner ends of the segmental organs. 

 If a similar view be adopted with regard to all arthro- 

 pods it will readily be seen that both antennal and shell 

 glands must fall into full accord with the nephridia of Per- 

 ipatus and were they to communicate with the so-called 

 body-cavity (blood-vascular space) their claims to the posi- 

 tion here assigned them would be weakened to a consider- 

 able extent. 



In connection with these studies of the development of 

 Crangon several questions have suggested themselves, a 

 few of which may be briefly mentioned here, though a full 

 discussion of them would require volumes as well as a spe- 

 cial knowledge of the details of the morphology of seg- 

 mented animals which few possess. 



As has already been suggested in this series ('866, p. 

 147), the Nauplius cannot be regarded as the adult condi- 

 tion of any crustacean and thus have an ancestral value. 

 It is rather to be regarded as an introduced feature in 

 the development of the ancestors which, though frequently 

 masked, is more or less clearly distinguishable in all of the 

 class. Its introduction into the series and its later tendency 

 towards obsolescence are, in my opinion, to be attributed 

 to paucity and abundance of the much-abused food yolk. 



A careful consideration of the distribution of proto- 

 plasm and deutoplasm in the crustacean egg will, I think, 

 show that the latter is an element which has been intro- 

 duced at a comparatively recent date. In those eggs where 

 the developmental history shows us that food yolk has long 

 been present, we find it either uniformly distributed through- 



