126 FRESH-WATER RHIZOPODS OF NORTH AMERICA. 



partition could be distinguished, it .appeared to be a crescentoid plate, with 

 a large circular aperture between its upper part and the contiguous portion 

 of the shell. In other instances, the aperture seemed to pass through the 

 upper part of the partition itself. 



In many specimens, the neck is more or less defined from the body of 

 the shell by a narrow constriction in the line of the partition, especially in 

 those in which the latter extends completely across. 



The structure of the shell of Difflugia spiralis is of more variable char- 

 acter than in any other species of the genus. Frequently it is entirely 

 composed of angular, colorless quartz-sand, as in the species of Difflugia 

 generally, and as seen in figs. 1-3, 6, pi. XIX. In these specimens, also, the 

 usual varieties in arrangement occur, some being constructed of particles 

 of some uniformity of size, small or large, while others are composed of 

 scattered coarser particles, with the intervals filled in with smaller ones. 

 Sometimes there is an accumulation, especially of larger stones, centrally 

 on the broader surfaces of the shell, as seen in figs. 4, 5. 



The former figure represents a large specimen, in which the shell was 

 mainly composed of thin, narrow, rectangular plates, mingled with some 

 diatoms, and the central portion of the broader surfaces of the body was 

 occupied by large particles of quartz-sand. 



A singular variety is that in which the shell is composed of short vermic- 

 ular bodies closely laid together, as represented in figs. 7, 9-11. The same 

 form is described by Dr. Wallich as occurring in England. The vermicular 

 bodies are transparent and colorless, but when viewed by reflected light 

 present a silvery white lustre. Dr. Wallich calls them chitinous pellets. 



Occasionally I have found specimens composed of similar bodies sepa- 

 rated by marked intervals, and apparently imbedded in a homogeneous 

 membrane, as represented in fig. 8. 



Other varieties of shells, related with those last indicated, consisted of 

 a net-work of apparently the same nature as the vermicular bodies, as seen 

 in fig. 12, or of minute scattered corpuscles, as seen in fig. 13. With these 

 shells, as represented in the two figures just referred to, there were incor- 

 porated scattered particles of quartz-sand. 



Another interesting variety observed is one in which the shell was 

 mainly constituted in the usual way of a stone-work masonry, but had the 

 neck composed of vermicular bodies, as seen in fig. 5. 



