318 PROCEEDINGS OF THE ACADEMY OF [May, 



way tlie coui'se of the fibres. In these cases the albumose is plainly 

 not influenctid by the nucleus. If absorption goes on continually 

 for some time a mass of granules will be found after fixing, on the 

 coelomic side of the nucleus (Figs. 12, sixteen hours, and 4, 

 riate XVI, fifty liours). 



Fig. 12. — Absorbing cell X 600 from median portion of "mid-gut" of 

 Oniscus asellus, 16 hours after feeding, fixed with Hermann's fluid. 

 Intima is torn off. Alb. Or., albumose granules ; Pal., palisade ; Inter./., 

 intercellular fibre, showing continuity with the basement membrane. 



It has been shown by Fischer (Z. c, p. 36 et seq.) that the size of 

 the granules of albumose depends upon the strength of the solution, 

 and upon the strength — i.e. , precipitating power — of the killing fluid. 

 Thus he finds the average diameter of the granules precipitated by 

 Altmann's osmic-bichromate mixture from a 10 per cent, solution is 

 l;jL to 3,a; from a 3 per cent, solution, .Ifito 1.5//; from 1 per cent., 

 .5fi to .7/jL. Again, the average diameter from a 10 per cent, solu- 

 tion precipitated by Flemming's fluid is .7fi to 1;j.; by Hermann's, 

 .7n (fairly uniform); by 1 per cent, platinic-chloride, .7,« to Ifi; 

 by 7 percent, sublimate, .4,'i to l,a and smaller. Those represented 



