1914] Digestion and Digestive Epithelium in Insects 313 
cells in size take their places. Thus we have a regular progres- 
sion of cells from the very small and scarcely distinguishable 
ones in the nidi to the full-grown secretive cells. The most 
interesting point here in connection with the study of cell 
structure and behavior is the existence of the nidi. What 
these are and how they originate is a question. They have 
been variously called “‘cryptes,’’ ‘“drusenkrypten,”’ and “epi- 
thelial buds.’’ None of these terms is very specific, and the 
idea of these bodies being glandular can hardly be retained, fora 
gland which secretes nuclei or cells is inconceivable. The nidus 
appears as a group of nuclei, exactly like those in the fully- 
developed cells except smaller, crowded together, and with 
very little protoplasm about them. (Figs. 1 and 2, n). In 
some instances this group of nuclei is enclosed in a sort of sac 
protruding out into the muscles surrounding the alimentary 
canal (Faussek, Frenzel, Rengel), and to this type in particular 
the term ‘‘drusenkrypt’’ has been applied. More often, how- 
ever, the nidus is an integral part of the wall of the canal, and 
there seems to be no special limiting membrane. Are these 
cell “‘anlagen’’ of which the nidus is composed split off from a 
mother cell? If so, where is this mother cell? Each nucleus 
of the nidus looks exactly like each other nucleus. Or is the 
nidus as a whole a cell which produces these nuclei, perhaps by 
division of its own nucleus? But this approaches the gland 
idea. As I look at it, it is simply impossible to apply the 
ordinary theories of cell constitution and cell existence to this 
structure, this nidus. We must look elsewhere. Mobusz, 
quoting Adlerz, mentions the presence of a network of proto- 
plasm between the basement membrane and the cell bases, from 
which new cells arise. We may have to advance a theory of 
something similar to this to account for the origin of the nuclei 
in the nidi. If they are not formed by division from others, 
can they by any possible means be formed from a net of 
protoplasm? A further and more careful study of these nidi 
is essential, and will undoubtedly throw light on the general 
question of cell origin. 
Let us turn to the merocrine type of digestive cells, that is, 
the type where the cell contents is only partly dicsharged as a 
digestive fluid. This type is to be seen in the alimentary canal 
of an insect that feeds continuously, thus demanding a con- 
