EXPERIMENTAL. 41 



solution was then decanted from a slight sediment, filtered clear, and heated 

 to 80, in order to coagulate any leucosin which might be present, and after 

 2 hours filtered from a very small coagulum which had gradually formed. 

 This filtrate was dialyzed in water for 4 days, and the globulin which sepa- 

 rated gave preparation 31. The solution filtered from the first dialysis 

 precipitates, which yielded preparations 28, 29, and 30, was further dialyzed ; 

 a little globulin, which separated, was filtered out and the filtrate dialyzed 

 in alcohol for 4 days. A precipitate was produced which, when washed 

 with absolute alcohol and dried, weighed 25 grams. This substance con- 

 sisted of protein which is described on page 47. 



Another series of fractional precipitations of this globulin-like protein 

 was made by extracting 4 kilograms of the oil-free germ meal with 27 

 liters of 3 per cent sodium-chloride solution heated to 67 at the time 

 it was applied to the meal. The mixture was thoroughly stirred and 

 thrown on filters. A clear filtrate of about 12 liters was finally obtained, 

 which was saturated with ammonium sulphate. The precipitate produced 

 was dissolved in water and its solution dialyzed for 48 hours ; where- 

 upon a large quantity of spheroids separated, which on settling united 

 to a coherent mass. This precipitate was washed by decantation with 

 water, dissolved in brine, and its solution made faintly alkaline to litmus by 

 cautiously adding decinormal potassium-hydroxide solution. In order to 

 separate phosphoric acid, a little calcium-chloride solution was then added 

 to this very slightly alkaline liquid, and the latter, though apparently free 

 from any precipitate of calcium phosphate, was filtered, whereby a little 

 suspended matter was removed. The solution was made exactly neutral to 

 litmus by adding 56 cc. decinormal hydrochloric acid and dialyzed for 18 

 hours. A gummy precipitate (A) adhering to the bottom of the dialyzer 

 then separated, from which the solution (B) was decanted almost completely. 

 The precipitate (A) was dissolved in about 200 cc. of 5 per cent sodium- 

 chloride solution and the liquid was poured into 800 cc. of water. The 

 resulting flocculent precipitate settled rapidly to a coherent deposit, from 

 which the solution was decanted. The deposit was repeatedly washed by 

 decantation with water, which caused it to lose its gummy character and 

 become opaque, white, and granular. It weighed 15.5 grams and was 

 preparation 32. The solution marked B was further dialyzed for 48 hours, 

 when a second precipitate formed, which, like 32, completely dissolved 

 in sodium-chloride solution to a solution perfectly neutral to litmus. This 

 precipitate was washed by decantation with water, but the finer part 

 settled so slowly that it was necessary to decant it together with the water. 

 The sediment, after exhausting with absolute alcohol and drying, weighed 

 2 3-5 grams, and formed preparation 33. On long standing the decanted 

 washings deposited the finely divided matter, which was then collected on a 



