ON THE ANTITOXINS OF DIPHTHERIA. 201 



precipitates. Wassermann (28) has described an elaborate 

 method for concentrating antitoxin from milk yielded by 

 goats which have been rendered immune to diphtheria. The 

 whey is separated by acidification with hydrochloric acid 

 and rennet, then shaken with excess of chloroform and 

 allowed to stand. By this means a clear supernatant liquid 

 free from fat and bacteria is obtained. On treatment with 

 33 per cent, ammonium sulphate proteid bodies together 

 with antitoxins are precipitated. This filtrate is collected 

 and placed on slabs of unbaked clay in a vacuum. The 

 excess of ammonium sulphate is pressed out of the pre- 

 cipitate, and this when dissolved in an amount of water equal 

 in bulk to the original whey gives a solution ten times the 

 original strength of the milk, so that '9 gramme of toxin 

 is neutralised by "125 ccm. of the solution. Interesting as 

 are such observations they do not essentially advance our 

 knowledge of the antitoxin. No one has ever seen this 

 substance, and the term itself is possibly of doubtful value. 

 Under Nencki's direction a renewed study of antitoxic 

 serum has been made by Smirnow (29), who, on finding 

 that the results obtained by the use of oxidising and re- 

 duction processes were only negative, employed a current 

 of about 160 milliamperes to electrolyse the fluid. 

 Directly the current traverses the serum bubbles of gas 

 develop at both the negative and positive electrode, and a 

 turbid precipitate collects at the former. This ultimately 

 clears up to a large extent, and an acid reaction is de- 

 veloped in the serum at this pole, while at the positive 

 electrode the fluid is perfectly clear and alkaline. This 

 observer, moreover, considers that serum-albumin in the 

 blood is modified rather than the globulins, and after 

 electrolysis antitoxic serum has wholly lost its power as a 

 protective measure against diphtheria. Without throwing 

 any doubt upon the facts which have been detailed, it is, 

 however, most unlikely that any information as to the 

 nature of antitoxin can be gained by such a method, for 

 serum is a fluid of the greatest complexity, and as to the 

 molecular structure of the proteids in that fluid absolutely 

 nothing is known. A continuation of Smirnow's researches 



