156 A REVIEW OF THE GOLGI METHOD. 



dure appears to me especially fitted to give the pieces in less time 

 that particular hardening most favourable to the best reaction with 

 silver nitrate. I mention, for example, a case where I have 

 obtained graduated reactions of surprising fineness on pieces fifteen 

 to thirty days after they were placed in bichromate at a tempera- 

 ture of 15^ to 20^ C. (in autumn), the pieces having been subjected 

 to the above treatment. 



" The injection is performed in the usual way (with a simple 

 syringe or with a siphon in which the pressure is regulated by the 

 height of the vessel containing the injection fluid) either through 

 the carotid, when one wishes to limit the hardening to the cere- 

 brum and cerebellum, or through the aorta when the fluid should 

 also extend to the spinal cord. 



" It is superfluous to state that when the bichromate and gela- 

 tine is injected it must be warmed so that it will remain fluid. In 

 this case it is especially important to perform the operation imme- 

 diately after the death of the animal, before the tissues are cold. 

 Only in this way does one secure the finest and most widespread 

 injection. 



"After the injection the nervous parts are removed from their 

 cavities, cut into pieces, and brought as usual into bichromate, 

 where they are carefully treated as dealt with above. 



" (b) Hardening in bichromate at a constafit temperature. The 

 circumstance, pointed out several times, that the uncertainty about 

 the time at which the pieces must be brought from the bichromate 

 into the silver solution depends for the greater part upon the tem- 

 perature of the medium, leads to the idea that the best means of 

 avoiding this inconvenience would be the employment of a con- 

 stant temperature for the bichromate in which the pieces lie. For 

 this purpose the warm chambers used in investigations upon 

 micro-organisms seem best adapted. 



" I have used the chamber of Wiesnegg, in which I maintained 

 a temperature of 20^ to 25°. This had good success, but only in 

 the direction of considerably abbreviating the period of hardening 

 in bichromate, so that the reaction could be obtained much sooner 

 than formerly and in a tolerably constant period of time. Thus, 

 the reaction in a warm chamber appeared after eight to ten days 

 and proceeded to completion up to fifteen to twenty days. This 



