346 THE NERVOUS CYTOLOGY OF PERIPLANETA ORIENTALIS. 



some cases of polychrome or of Van Gieson's 5% solution of methylen blue for the 

 Nissl solution. 



2. Observations. The appearances of the cells obtained in this way were indeed 

 striking. The blue cells (Figs. 20-23), teased out of their connective-tissue 

 spaces, and entirely devoid of a cell-wall, exhibited an even outline, giving no evi- 

 dence of shrinkage or other distortion. The nuclear membrane, though buried in 

 the cytoplasm, could be made out intact, with even outline, like that of the whole 

 cell. The nucleus itself presented a homogeneous blue substance, entirely filling 

 its membrane and containing one or more nucleoli, that took a deep-blue color. There 

 was no evidence of any reticulum in the nucleus. The cytoplasm was occupied through- 

 out its entire extent by a sharply defined anastomosing blue network with slight concave 

 nodal enlargements, but no granules or masses either at the nodes or on the strands 

 (Fig. 24). Its spaces were of equal size and its strands of equal thickness throughout 

 the entire cell. Different cells, however, varied considerably both in the thickness 

 of the strands and the size of the meshes, but the smaller cells by no means regularly 

 presented the smaller-meshed network. Except for this structure, the cell-body 

 was entirely clear and transparent. There was no division into the two zones, seen 

 after the various fixing methods (PI. XXVI, Figs. 20, 21; PL XXVII, Figs. 22-26). 



After about one hour the cell began to shrink and the network to lose its defini- 

 tion. These changes were regularly preceded by the shrinkage of the nucleus within 

 its membrane. The above appearances were uniformly obtained in a number of 

 similar subsequent preparations. 



The nerve-fibres were seen to be composed of fibrilla? disposed for the most part 

 longitudinally, but anastomosing with each other, the nodal points being prominent; 

 in other words, like the cell-body, they contain a network of fibrillse (PI. XXVII, Fig. 

 26) . In preparations of teased specimens fixed in formalin vapor, as will be described, 

 there were no granules or masses of deeply staining particles to be seen upon 

 the strands of the network in the nerve-fibre, although these strands stood out with 

 great sharpness. It was also to be noted that in such preparations the fibrilla? in 

 the nerve-fibre anastomose at very acute and not at right angles (Fig. 28). 



Satisfactory permanent preparations of whole cells in approximately this condition 

 were obtained by two methods. In neither was the network as clear as in the fresh 

 condition, but slightly obscured by an irregular granular deposit at the nodes and 

 along the strands. The first one was more simple and yielded better results. 



3. Technique. (1) The slide with the teased cells on it was exposed to the vapor 

 of strong formalin (10% or over), stained, dehydrated, and mounted in the usual 

 manner. 



