352 THE NERVOUS CYTOLOGY OF PERIPLANETA ORIENTALIS. 



strated, although I am inclined to believe they are real. The picroformalin prepara- 

 tions were, as with those produced by Faradization, too much distorted to yield any 

 definite results. 



XVI. CELLS FROM ROACHES POISONED BY CORROSIVE SUBLIMATE. 



1. Technique. A roach was fed on equal parts of a 5% solution of corrosive 

 sublimate and molasses. Fifteen hours later the animal was found on its back in 

 general flexion, with faeces oozing from the anus, and exhibiting spasmodic movements 

 of the limbs from time to time. The alimentary tract contained molasses. The 

 ganglia were removed and prepared by the same three methods as in the other 

 experiments. 



2. Observations. None of the preparations exhibited any appearances differing 

 from normal material. 



XVII. CELLS FROM ROACHES POISONED BY ARSENIC. 



1. Technique. A roach was fed on molasses with which a considerable amount 

 of Paris green had been mixed. The animal was moribund in one-half hour. The 

 alimentary tract contained Paris green. The ganglia were fixed by the same three 

 methods as in the other experiments. 



2. Observations. The graded formalin and teased preparations showed nothing 

 abnormal. The picroformalin specimens, however, showed a most unusual condition. 

 Perinuclear condensation of the chromophilic material and peripheral laceration and 

 distortion of the reticulum were practically absent, though some tendency to the usual 

 artificial condition after the use of this reagent was seen in the cells with compara- 

 tively scant amounts of deeply staining deposit. For the most part the network and 

 its deposit were evenly disposed throughout the cell-body (PL XXVIII, Figs. 48, 49). 

 In other words, the deposit had occurred during life and so reinforced the reticulum 

 as to protect it from the mechanical violence of the fixing reagent. There was, of 

 course, no more time for post-mortem changes to occur in this case before the appli- 

 cation of the reagent than in any other, as this particular dissection was completed 

 in five minutes. 



