320 PROCEEDINGS OF THE ACADEMY OF [May, 



U()t destroyed by the fixation, the space should correspond to the 

 neighboring alveoles in size — i. e. , in such cells as that represented 

 in fig. 6, where the alveoles are so evident, the granule would be 

 plainly seen encased by the alveolar wall. Such an appearance is 

 seldom met with — indeed, the granule is often larger than the alveole 

 (figs. 9, 11, B, 9, Plate XVI). Again, in favorable places (figs. 6, 

 8, Plate XYI), the alveolar walls can be seen intersecting the space 

 about the granule. ^* The space does not correspond, therefore, to an 

 alveole, but represents the area in the interalveolar substance influ- 

 enced by the kiUiug fluid to form a single granule. It appears, 

 therefore, that the soluble food is independent of the cytoplasmic 

 structure. 



Fig. 14. — Outliue drawing (with camera X 90) of au optical section of 

 anterior cells, "mid-gut" of Oniscus asellus, fed with raw beef; intes- 

 tine mounted in blood and stained 15 minutes in dilute Bismarck brown 

 (1 : 20,000) 16 hrs. after feeding. The granules about the nucleus take 

 the stain and appear to be coming from the nucleus. This is precisely 

 the appearance of zymogenesis in the hepatopancreas when treated in 

 the same manner (figs. 20 and 21). 



Having distinguished the albumose granules from the rest of the 

 cell contents, we may inquire whether further chemical stages of the 

 food appear in the fixed and stained sections in a form distinguish- 

 able by the microscope. It was mentioned farther back (p. 317) 

 that evidence might be found for regarding some small granules 

 of feeble staining capacity as a stage in the transformation of food. 

 In the first place, it must be stated that the digestion of the food 

 (raw beef) is pancreatic (i. e., alkaline'^) (see Sec. ix, a); and 



" Cf. Hardy's figures, loc. cit. 



^^This we have on strong comparative evidence also. Thus Plateau 

 (24) demonstrated (a) that the reaction in the crop of the Carabidse and 

 Dytiscid* may be neutral before feeding, but is always alkaline after; 

 (6) that the reaction in the "mid-gut " of the carnivorous chilopod Litho- 

 bius is alkaline, in that of the herbivorous diplopod Julus may be acid. 



