[77] EMBEYOGRAPHY OF OSSEOUS FISHES. 531 



hitherto been generally ascribed to them. They are, in fact, the primal 

 representatives of nutritive spaces — lymph cavities j perhaps even of 

 the food and water vesicles of Protozoa. 



On the twenty-seventh to the thirtieth day of development, as shown 

 in Figs. 49 and 45, resi)ectively, the regions of the intestine for the first 

 time begin to be clearly mapped out. In Fig. 49 the depressed oesopha- 

 geal portion of the alimentary canal ends just over the lower lobe of the 

 liver Iv and just in front of what appears to belts upper portion y. In 

 i'ront of ?/ there lies a body, covered with large stellate i^igment cells, 

 which 1 have identified with the air or swim bladder. Its mode of 

 origin I have not made out in the young cod, but in Gamhusia and Alosa 

 it is a distinct dorsal diverticulum of the intestine, which arises a little 

 to one side of the median line. Its hinder end is prolonged backwards 

 with the advance of development, and is at fli"st a small and inconspicu- 

 ous structure, with a thick wall, which, on its ventral face, may be lined 

 by what appears to be glandular epithelium, as in Gamhusia. The con- 

 nection of the pneumatic diverticulum with the intestine is by a nar- 

 row open canal, which may remain open in the adult, as in physosto- 

 mous forms, for example, where it forms a pneumatic duct, or it may be 

 aborted during a post-larval stage, as in the physoclistous species. 



Behind the liver and air-bladder the intestine becomes suddenly 

 widened, as shown in Fig. 49, and has its internal surface elevated into 

 low folds or i)aiiillae, which are the rudiments of the gastric and intes- 

 tinal follicles of a later stage. This widened portion of the intestine is 

 continued backwards until a constriction is encountered at ic. From 

 the liver back to the constriction alluded to, the middle portion of the 

 intestine later becomes the stomach. The constriction is apparently 

 the pylorus and i)yloric valve, while the section of intestine from the 

 constriction to the vent v becomes the hind-gut of the adult, with an 

 almost uniform caliber throughout. 



Peristaltic action of the intestinal wall shows itself very early in fish 

 embryos, or about the time that the three regions are distinctly marked 

 out as described above. I have frequently witnessed its manifestation 

 in Bewly hatched shad, and also when they were a few days old and 

 had begun to take small crustaceans as food. The peristaltic contrac- 

 tions of the intestinal wall would jjush back the food to about the point 

 where the oesophagus ended and where the liver began, and where the 

 intestine was considerably widened. This widened portion was then 

 continued back to a similar pyloric constriction, beyond which I but 

 rarely saw the food carried. 



The histological features of the intestinal walls of embryo fishes are 

 interesting in that it is the mucous or epithelial layer which is princi- 

 pally developed. The muscular layers, both the longitudinal and an- 

 nular, are thin. The latter is pretty thick in embryos of Salmo. In the 

 neighborhood of the commencement of the stomach in embryos of Alosa 

 the mucous pits and folds of the enteric epithelium are most pronounced 



