100 JOHNE'S DISEASE 



reacted, so as to produce some more acid for its own 

 use. But we know that in the process of repair there 

 is usually over-production, a rule that is well illus- 

 trated in the formation of antitoxins ; so that one 

 would expect an over-production of the acid, and this, 

 again, would be more favourable to the bacilli, or, if 

 these had been killed, to a second infection of bacilli. 



Although a great deal of what has been said is 

 theoretical, it is not entirely so, as our experiments 

 alone suggested that the essential substance might be 

 an organic acid ; so we again started to investigate the 

 alcoholic extract of B. phlei. The extract obtained 

 from loo grammes of dried bacilli as a whole showed 

 but slight acidity; but when separated into different 

 portions with the solvents already mentioned, we 

 found that some were markedly acid, and, further, that 

 those portions that were most acid made the best 

 media for Johne's bacillus. Using that portion of the 

 extract which is insoluble in chloroform and soluble in 

 water, we attempted to precipitate the acid with baryta 

 water. Only a small precipitate formed, so the fluid 

 was evaporated to dryness, and the sticky mass so 

 obtained heated with methyl alcohol. Most of the 

 mass dissolved, but a very small insoluble portion 

 (about 0*007 gramme) remained. This was incor- 

 porated in 30 c.c. of egg medium, which was dis- 

 tributed in tubes, and sterilized as usual. A good 

 result was obtained — that is, Johne's bacillus grew 

 well on the medium, and no growth took place on 

 a control medium made up at the same time with the 

 same mixture of eggs. The test was repeated on 

 several tubes of both media with another strain of 

 Johne's bacillus, and the same result was obtained. 

 Control media, containing small quantities of barium 

 hydrate, gave negative results. This experiment is 



