296 Surgeon-Captain D. Bruce. Disappearance of [Mar. 15, 



I may mention here that for practical purposes and without pre- 

 judice as to their origin, I adopt the classification of the leucocytes of 

 the rabbit's blood into four varieties : 



A. The Eosinophilous, constituting on an average 2 per cent, of the 

 total leucocytes, with an irregular-shaped nucleus and large oval 

 shaped grannies, which stain readily in eosine. 



B. The Polynuclear, 51 per cent., with intensely staining poly- 

 morphous nucleus, having the appearance of several nuclei united by 

 narrow thread-like processes, the protoplasm of the cell containing 

 fine granules which also stain in eosine. 



C. The Myelocytes, 16 per cent. Difficult in every case to rigidly 

 separate from the fourth variety, but defined as mononuclear cells, 

 the nucleus of which stains badly, and is surrounded by compara- 

 tively a large amount of non-granular protoplasm. 



D. The Lymphocytes, 31 per cent. Mononuclear cells, little larger 

 than a red blood corpuscle, with intensely staining nucleus and narrow 

 rim of protoplasm. 



The preliminary enumeration of the leucocytes in the blood of the 

 ear or heart was made by diluting the blood 200 times with an 8 per 

 cent, magnesium sulphate solution, to which sufficient gentian violet 

 had been added to stain the white blood corpuscles, and not by 

 estimating their proportion to the red blood corpuscles, which seems 

 to me a most unsound method. The number of white blood cor- 

 puscles in 300 squares of a Gower's haemocytometer were then 

 counted and this number multiplied by 333, which gives approxi- 

 mately the number in a cubic millimetre. 



Three samples of blood are taken, two of which are counted, and, 

 if the counts are sufficiently close, an average made. In the event of 

 there being a marked discrepancy, the third sample is counted and an 

 average of the two most alike taken, but, with care, it is seldom found 

 necessary to use the third sample. 



In regard to the fixing and hardening of the tissues for cutting, 

 alcohol, Mllller's solution, Foa's, and Flemming's solutions were used, 

 but the best results were obtained by fixing for twenty-four hours in 

 Flemming's strong solution, washing in running water for twenty- 

 four hours, then through successive alcohols for forty-eight hours. 



The tissues, after infiltration with paraffin, were cut as neatly as 

 possible of the same thickness, and stained in various ways, the most 

 successful results being got by staining for twenty-four hours in the 

 Ehrlich-Biondi triple stain. 



On taking the sections out of the stain I place them for a minute 

 in a saturated solution of corrosive sublimate, as this seems in some 

 way to prevent them becoming too much decolorised during the sub- 

 sequent manipulations. 



The white blood corpuscles contained in twenty fields of the micro- 



