124 BACTERIOLOGY 



tain of these bodies could not be called pro- 

 teids at all. To illustrate, if certain bacteria 

 are cultivated upon proteid-free nutrient media, 

 such as ammonium lactate or asparagin, 

 poisonous substances are usually produced, 

 but proteid reactions are not always obtained 

 from the filtered solutions. On the other 

 hand, it is positively established that the 

 poison always constitutes only a slight frac- 

 tion of the whole solution. In solutions con- 

 taining proteid there is always some poison 

 side by side with much proteid, and the inten- 

 sity of the reaction has often been falsely at- 

 tributed to the quantity of the latter, for we 

 know of no chemical method by which we are 

 able to distinguish between the poisonous 

 active proteid and the non-poisonous passive 

 proteid of the same group. A very slight 

 amount of poisonous proteid, however, in an- 

 other solution may elude our test for proteid. 

 So, for example, one part of rennet curdles 

 two million parts of casein ; according to Vail- 

 lard, there is contained only 0.0002 5 gr. of the 

 solid tetanus toxin in a dose fatal to a guinea- 

 pig ; and according to Kobert and Stillmark, 

 0.00003 gr. of ricin per kilogram of body- 

 weight of the animal is fatal when injected 

 into a vein. Such slight quantities may well 

 escape the coarse reactions of proteid chemis- 



