254 MICROBES AND TOXINS 



which agglutinate spontaneously or do not agglutinate at all. 

 Strains can be artificially produced which refuse to agglutinate : 

 this resistance of the microbe to the action of the serum 

 is an example of the immunity of the microbe towards the 

 animal, the converse case of the immunity of the animal 

 towards the microbe. 



Agglutination is applied to as a test for suspected bacteria 

 found during cholera scares in suspected water or choleraic 

 diarrhoea ; the serum employed is obtained from an animal 

 prepared by several immunizing injections of a definitely known 

 cholera vibrio. 



Agglutination has also been applied to the diagnosis and 

 prophylaxis of bacillary dysentery and epidemic cerebro-spinal 

 meningitis. In tuberculosis it has only been applied as a 

 means of controlling the treatment by tuberculin. As a 

 diagnostic agent the tuberculin test is much more convenient 

 and certain (R. Koch). But agglutinin is no better than the 

 other antibodies (vide p. 208) for the estimation of the resist- 

 ance of the body towards tubercle.^ 



Precipitation. 



If we take a culture of B. typhosus and filter it, we obtain 

 a clear fluid free from bacteria. If a Iiltle of a very active 



^ Agglutination is . the touchstone in the study of bacterial strains and 

 their variations. Recently Bordet and Sleeswyk, working with the bacillus 

 of whooping-cough discovered by Bordet, have created in the laboratory 

 varieties of this analogous to the varieties of the dysentery bacillus. 



What fixity do these varieties or strains possess, created as they are by 

 growth on different media and separated by their different reaction to 

 such-and-such a serum ? What is of interest here is the qonclusion of these 

 workers. They maintain, at least as far as the agglutinating action of 

 serum on bacteria is concerned, that sera " do not act on the fundamental 

 bacterial substance, which is inherent to its life and whose presence is 

 indissolubly bound with the nature and constitution of the species, but 

 that they act on substances to some degree accessory, of possible but 

 facultative presence, whose production is in no way one of the bundle of 

 hereditary, immutable characters which give to a living creature its own 

 physiognomy and autonomy." 



It is obvious that this interpretation of the facts is not in favour of a 

 chemical theory of immunity : the more one considers the agglutination as 

 a surface reaction, the more probable becomes the physical explanation, i.e., 

 the explanation of Bordet. 



