62 J. B. JOHNSTON, 
these preparations I have sought to improve upon the usual pro- 
cedure of cutting the brain into small pieces for impregnation. 
This method gives very incomplete and broken series which are 
chiefly useful for the study of nerve elements in restricted regions, 
and are not well adapted to the study of fibre tracts and the re- 
lation of various parts to one another. All of the GoLGr prepar- 
ations were made in the following manner. From the head of the 
living or recently dead fish there was cut out a block of cartilage 
about 10 mm square and 50 mm or more in length, containing the 
brain and the roots and part of the ganglia of the cranial nerves. 
The entire block was treated by the rapid — osmium-bichromate —- 
method. Sections were cut 60, 75 or 90 microns in thickness accor- 
ding to the character of the impregnation and were carefully seriated 
and covered with dammar. The surrounding cartilage was of distinct 
service in handling the brain, and perhaps also in preventing the 
formation of precipitates in the brain tissue. These series when 
mounted were as complete as series prepared by ordi- 
nary histological methods. Such GOuLai series are distinctly 
adapted to the study of fibre tracts and have to a large extent ser- 
ved the purpose of WEIGERT sections (which could not be prepared 
for want of material). In tracing fibre tracts the individual fibres 
constituting each tract have been studied at the origin and ending 
of the tract; and in all regions where the course of the tract was 
in doubt, final judgement has rested solely on the course of indi- 
vidual fibres whose character was certainly known. In offering the 
opportunity to apply this rigid test to the tracing of fibre tracts, 
complete GOLGI series have a peculiar value. 
To the technique of GoLGI impregnation I have little to con- 
tribute. Strrone’s method has been followed in the main, the os- 
mium-bichromate solution being varied by the addition of new fluid 
or of crystals of potassium bichromate as seemed to be necessary 
during the course of each staining process. Silver nitrate of one 
per cent strength was used, and large quantities of both fluids 
were employed. After the sections are arranged on the slide they 
are washed with one change of 96°, alcohol, and one of absolute 
alcohol. They are then covered with very dilute collodion which is 
allowed to evaporate until it begins to appear dry and wrinkled. 
The treatment with absolute alcohol allows the dilute collodion to 
fuse completely with the collodion surrounding the sections, so that 
they are cemented together with a continuous film of collodion, 
