THE NERVOUS CYTOLOGY OF PERIPLANETA ORIENTALIS. 355 



both the connective-tissue spaces and the reticulum of the cell-body, besides causing 

 a bulging of the nuclear membrane. How formalin solutions cause this taking up 

 of water is, I believe, not yet explained, although their swelling action on all masses 

 of tissue has been recognized since their first use in anatomical technique. 



The diffused alcohol and graded formalin methods, although far superior to 

 those previously used and described, each causes some slight mechanical violence 

 to the cell-structure, that is, the protoplasmic elements are not fixed with sufficient 

 firmness until osmosis and diffusion have caused some damage. They both cause 

 a nuclear network to appear which does not exist in the fresh state. 



It is unquestionably true that a somewhat new set of methods has to be worked 

 out for each new kind of material studied, but I venture to predict that vapor of con- 

 centrated formalin, or any other reagent possessing similar advantages, will prove 

 a most satisfactory fixing reagent for small and readily penetrable masses of tissue, 

 being entirely free from the dangers of mechanical damage by rapid dehydration, 

 besides leaving the nucleus approximately as in life. For larger and less penetrable 

 masses some combination of graded formalin and diffused alcohol methods, yet to 

 be elaborated, may prove thoroughly satisfactory. Some eight years ago I worked 

 with Dr. G. H. Parker in determining the requisite proportions of these two reagents 

 for preserving mammalian brains in exactly their natural proportions (see Parker 

 and Floyd, '95). 



A comparison of the actions of the fixing reagents of different groups is shown 

 in Figures 50 to 54 (PI. XXVIII) 



XIX. SUMMARY. 



This research demonstrates that the nerve-cells in the thoracic ganglia of the 

 roach possess no evident cell-walls, that their nuclei, though exhibiting a reticulum, 

 with enlarged nodal points and irregular amorphous deposits, after most fixing reagents, 

 are homogeneous in appearance in the fresh condition and after fixation in formalin 

 vapor. They contain nucleoli and are surrounded by an entire nuclear membrane, 

 the instances of its apparent interruption in fixed specimens being due, I believe, 

 either to its lying in the plane of section, on account of wrinkling, etc., or to its rup- 

 ture by the fixing reagent. The cell-body contains a fine anastomosing reticulum 

 whose fibrils show no more ultimate structural divisions. Its interstices show no 

 structure or staining affinities in the fresh condition. There are, however, one or 

 more substances, presumably existing in the cytolymph in the normal living cell, 

 that may change in character and form deposits upon the cytoreticulum under 



