206 DR. W. N. F. WOODLAND ON TUE 



occasional intra-cellular ducts and ca])illaries. I have not observtd 

 a pancreas in the region of tlie I'ete. The cell-outlines of the 

 gas gland are faint and a small amount of giunular material is 

 present in the bladder lumen in my preparation. 



1 stated above that the massive type of gas gland is quite 

 distinct from the folded type in certain genera, and, indeed, these 

 two types of bladder ej^ithelium are almost always easily distin- 

 guishable fi'om each other in the genera which I have studied : 

 genera, with a few exceptions to be mentioned, either possess 

 the folded or the massive type of gland and not glands inter- 

 mediate in structure. In young specimens of Atlterina about 

 15 mm. in length, that portion of the bbulder epithelium which 

 is about to produce the gas gland is at fiist composed of a single 

 layer of cubica,l or short cokunnar cells, and it is worthy of notice 

 that these cells, in order to build up the fully-formed massive 

 gland, all divide at right angles to the plane of the bladder wall 

 (PI, VII. fig. 55), i. e. the successive planes of cell-cleavage are 

 parallel with the plane of the bladder wall — the initial unilaminar 

 epithelium never becoming folded at any stage of development. 

 As the gas gland cells multiply connective tissue cells may be 

 observed to extend in between them in order to furnish the 

 connective and vascular tissues which penetrate tlie mass of the 

 gland when fully formed. Also if figures 54 and 55 be compared, 

 it will be seen that the cells (especially the nuclei) of the gas 

 gland of the adult Atherina are noticeably smaller as compaied 

 with those in the young tish. The fact that there is no indication 

 of folding of the bladder epithelium during development of the 

 massive type of gland is important, since Reis (64) supposes, on 

 the ground that there exist forms of bladder epithelium more 

 or less intermediate in structure between the folded and massive 

 types, that the latter has been derived from the former, and if 

 such be the case we might expect that the massive type of gland in 

 its development would pass through a folded developmental stage. 

 As we have seen, this is not the case in Atherina and presumably 

 not in other genera possessing typically massive glands. I assume 

 this with the more confidence since in some gas glands which, 

 owing to the exceedingly close apposition of the folds, at fiist 

 sight appear to be massive, it is always possible to refer them 

 at oncG to the folded type by observing in each such case the 

 margins of the gland where it merges into the squamous epithe- 

 lium lining the greater part of the bladder cavity, since here the 

 folds are quite simple : such marginal folds arc never present in 

 true massive glands. 



Reis, in contending that the massive type of gland is but a 

 derivative of the folded type, instances as transitional forms 

 between the two types the gas gland epithelia of Syvgnathus and 

 Girardinus. I have not examined GirardimLS, but the gas gland 

 epithelium of my specimens of Syngnathus can certainly not be 

 said to be transitional, since, as my description above of this epithe- 

 lium clearly shows, the massiveness of the gland merely results from 

 the extreme leneth of the folds ^nd their occasional anastom^osis : 



