366 BACTERIA. 



Brieger and Fraenkel found Millon's reagent gave a 

 white precipitate, which on heating became brick red in 

 colour ' this indicating its proteid nature ; it is precipitated 

 by magnesium sulphate in saturated solution ; it is there- 

 fore not an ordinary albumen ; whilst on the addition of a 

 drop of dilute sulphate of copper solution and a slight 

 excess of potash solution (the so-called biuret reaction) a 

 rose red and not a violet colour is given, indicating that 

 this material belongs to the albumose rather than to the 

 globulin group. There are a number of other tests which 

 it is not here necessary to describe. 



It must be remembered, however, that these proteid 

 poisons and the ptomaines are very closely bound up 

 with one another. Martin indeed holds, that as some of 

 the albumoses are less toxic than the alkaloids with which 

 they occur and as they also have a marked alkaline reaction, 

 the alkaloid may really be bound up in a nascent condition 

 in the albumose molecule. This may undoubtedly be the 

 case with certain vegetable alkaloids and albumoses, but in the 

 cases of diphtheria and tetanus, it would appear that some of 

 the so-called poisonous alkaloids owe their specific properties 

 to the presence of minute traces of an enzyme, or proteid 

 poison that is present along with them. It has even been 

 suggested that the alkaloid may be the poisonous agent 

 formed by bacteria, whilst the albumose is to be looked upon 

 as the protective agent. This, however, can scarcely be 

 maintained except in certain cases, as in the case of tetanus 

 the poisonous agent is certainly not of an alkaloidal nature. 

 Still, these points should be borne in mind by any one going 

 to work at the question. It seems to be undoubtedly the 

 case, that in accordance with the well-known fact that 

 certain products of micro-organisms, such as those of acid or 

 alcoholic fermentation, act deleteriously on the bacteria that 

 produce them, especially as they accumulate in large quan- 

 tities. On the other hand, it may be that these, or similar 

 products in a more or less dilute form, may be necessary 



■ Millon's reagent is prepared by adding one part of mercury to two 

 parts of strong nitric acid, gently warming until the mercury is thoroughly 

 dissolved : to one part of this mixture two parts of water are added ; a 

 precipitate is formed ; the supernatent fluid only is used. A few drops 

 of this solution give the above characteristic reaction with all proteid 

 materials lield in solution except in the presence of common salt. 



