ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 165 



complexity is due to the necessity for providing space for the sinuses and 

 accessory cavities. He does not regard these cavities as of prime im- 

 portance, and believes that they, like the surrounding hones, owe their 

 form in the first instance to the necessity for providing adequate space 

 for the teeth, the muscles, jaws, eyes, and brain. He illustrates and 

 enforces his view by giving some account of the development of the 

 nasal cavities. 



Development of Pancreas in Amniota.* — Dr. Yolker has studied 

 this in Lacerta agilis, Spermoplrilus citillux, and the pig. In re.ard to 

 the lizard, he corroborates and amplifies the observations of Janosik. 

 The pancreas arises as a simple diverticulum of tbe dorsal enteric wall 

 more distal than the primordium of the liver; it becomes secondarily 

 connected with the rudimentary ductus choledochus; from this pan- 

 creatic duct the proximal pancreas arises, and the connection is never 

 lost; the diverticula which Brachet described in Lacerta muralis as 

 appearing on both sides of the ductus choledochus, between the opening 

 of the ductus hepato-entericus and the ductus cysticus, which he re- 

 garded as pancreatic primordia, unite in Lacerta agilis with the hepatic 

 tissue, and form right and leit ductus choledochohepatici ; these canals 

 take the place of the ductus hepato-entericus, which eventually atrophies. 

 It remains clear that in the lizard the whole pancreas is iormed from 

 the dorsal primordium. 



In Mammals, likewise, the dorsal pancreas is the constant structure. 

 In pig embryos there is added to this another portion which takes its 

 origin from the ductus choledochus, but this fades into the dorsal pan- 

 creas and its duct atrophies, though there may be occasional instances 

 of persistence. In the embryos of SpermopMlus citillus, the only part 

 developed is the dorsal pancreas, which arises directly from the dorsal 

 enteric wall. 



Development of Mid-brain in Bony Fish.f — D. Pedaschenko finds 

 that in the embryos of various Teleosts tiie mid-brain is distinctly seg- 

 mented at an early stage. This segmentation is most distinct in Zoarces 

 viviparv.8, in which he has studied it in detail. The segmentation is 

 only obvious on the inner wall of the roof of the mid-brain, and the 

 author does not discuss its relation to the segmentation of the head. 

 The roof of the mid-brain at an early stage becomes so marked with 

 furrows internally as to produce rows of polygonal projections or 

 columns. The maximum number of segments, as defined by the cross- 

 furrows, appears to be six, and as each row contains 3-4 columns, there 

 are 20-24 of these. Later the columns increase enormously in number, 

 and become smaller in size, for the new columns arise by divisions of 

 the old ones. The subsequent disappearance of the columns is accom- 

 panied by the formation of bundles of fibres connecting the floor and 

 roof of the mid-brain together. This precedes the growing together of 

 floor and roof, and the consequent obliteration of the medullary cavity, 

 except in the extreme posterior region of the mid-brain. The author 

 proposes on another occasion to discuss the theoretical significance of 

 these facts. 



* Arch. Mikr. Anat, lix. (1901) pp. 02-93 (21 figs.). 

 t Tom. cit., pp. 295-314 (3 pis. and 4 figs.). 



