BY R. J. TILLY A RD. 51 



wall. The exceptions to this occur so seldom (I have seen one 

 alveolus with three nuclei upon its wall, a few with two, and 

 perhaps a score or more with none at all, out of thousands ex- 

 amined) that they are scarcely worth considering. But it is 

 also true that the nucleus that does duty for one alveolus is fre- 

 quently the same nucleus that counts for an adjoining alveolus, 

 or even for two other alveoli. This fact seems to have escaped 

 Ris' attention; but it is nevertheless a very important piece of 

 evidence when we come to discuss the true nature of the alveoli. 

 I find that, as a general rule, the nucleus is situated at a point 

 of union between the wall of one alveolus and the walls of one 

 or more adjoining ones. In other words, the walls of separate 

 alveoli are not distinct structures, each the product of a single 

 cell, but are formed by fusion of a number of branching cells 

 having nuclei at their centres. Consequently each alveolus is 

 not, as Ris' supposition would indicate, the hollow interior of a 

 single cell, but merely a small portion of the original htiemocoele, 

 closed off by the ingrowth and fusion of cell-processes around it. 



If my interpretation of the structure of the alveoli be correct, 

 we should expect to find nothing actually inside the alveoli 

 themselves except blood-plasma and, perhaps, an occasional 

 amoibocyte; whereas, if Ris' interpretation be the true one, it 

 would be reasonable to expect some internal structure sui generis 

 within the alveolar cavity. 



Now Ris has figured (28, Plate 6, figs. 17-20) and described at 

 some length, some extraordinary structures from the interior of 

 the alveoli. He says (I translate from the original German) : — 



One part of the above-mentioned alveoli is fuli of very rich 

 and delicate ramifications in moss-like branching tufts of stand- 

 ing fibrils ("reichen und zierlichen Verzweigungen in moosartig 



verastelten Biischeln stehender Fibrillen") In one alveolus 



one single fibrillar tuft of this kind appears to be predominant; 

 but formations are not lacking in which a number of smaller 

 tufts are found in near or distant parts of the same alveolus. 

 Other formations are also to be noticed as stages in the develop- 

 ment of these fibril-tufts : button-shaped and closely packed 

 together, running out from a narrow base in countless threads. 



