THE MESODERMAL STRUCTURES. 307 



According to the view of VIALLETON, which has already been mentioned, 

 a delamination of the mesodermal part from the outer layer takes place even 

 at the time when, according to the account here given, the germ-layers have 

 long been differentiated. At a much later stage also, when the rudiments 

 of the organs have appeared and have to some extent developed, the ecto- 

 derm continues to yield mesodermal cell-material. VIALLETON compares this 

 process with the development of mesoderm in the Annelida as described by 

 K i, KIN KN BERG (Vol. i., p. 286), and refuses to acknowledge this mesoderm as 

 the equivalent of a distinct gerro-layer. Such a view can best be understood 

 in connection with the Cephalopoda in which the formation of the germ- 

 layers is evidently greatly influenced by the large amount of yolk in the 

 egg, the distinction between the layers appearing almost obliterated, but in 

 this case, as in that of the alimentary canal and the nervous system, we 

 must postpone a decision until further light is thrown upon the matter. 



In the Cephalopoda, especially in Nautilus and the Decapoda, the 

 secondary body-cavity (the coelom). is very well developed and sur- 

 rounds the heart with its afferent and efferent vessels, the branchial 

 hearts, the pericardial glands, the genital organs and the stomach, 

 and is connected with the cavities of the nephridia (GROBBEN, No. 

 15). The last two organs lie in the posterior section of the body- 

 cavity which is incompletely separated from an anterior section by 

 a septum. 



In the Octopoda the body-cavity is much reduced, being represented 

 merely by a system of narrow canals formerly claimed as a water-vascular 

 system. In consequence of its slight development, the coelom no longer 

 surrounds the heart, the branchial hearts and the stomach (GROBBEN). 



As the condition of the coelom in the Cephalopoda is so primitive, we 

 are led to expect the coelomic sacs to appear distinctly in their ontogeny. 

 But the conditions of the formation of the mesoderm above described seem 

 in this respect to be unfavourable. A splitting of the mesoderm into a 

 somatic and a splanchnic layer has indeed been asserted by Ussow, but 

 is not noticed by other observers and, in any case, does not lead to the 

 formation of coelomic and pericardial sacs such as were met with in the 

 Gastropoda or such as we might expect in connection with the well-developed 

 coelom of the adult Cephalopod. We may observe in passing that there 

 are reasons for believing that structures of this sort may yet be found in 

 the Cephalopoda. 



The kidneys, in the Cephalopoda, show the same primitive condition as in 

 other Molluscs. In the Octopoda, they are represented by two sacs lying 

 - yi n metrically ; in the Decapoda these have united to form a single sac. 

 The renal sacs open out through ureters on either side of the anus. 



A comparison of the Cephalopodan kidney with the nephridia of the 

 segmented animals seems specially suitable on account of two pairs of renal 

 sacs being found in Nautilus, each pair opening out through a distinct 

 aperture. One of these pairs, however, is without an aperture into the 

 coelom, and its significance as a true nephridium is therefore doubtful ; we 



