METHODS 53 



the approximate law a phenomenon of destruction super- 

 posed with a phenomenon of assimilation. 



3. The next step is even more interesting. In pure water, 

 to which the same quantity of phenic acid is added as to the 

 culture-medium in the preceding experiment, you sow the an- 

 thrax bacteridia and the assimilative reaction is suppressed. 

 Nothing remains but destruction, and the destruction is 

 entire at the end of a certain time. But if you stop the 

 destruction before the death of the bacteridia, you verify 

 that the partial destruction of the bacteridia has produced 

 the same attenuation of virulence as in Experiment No. 2. 

 We can say, therefore, that, in Experiments 1 and 3, we 

 have artificially separated the phenomena of assimilation 

 and of destruction, the superposition of which gave us the 

 result obtained in Experiment No. 2. This is Helmholtz's 

 method of decomposing a complex sound into several simple 

 sounds. 



In this particular case, by exact experiment, we have 

 caught in the act assimilative and destructive reactions 

 superposing themselves. Hence we can assert with more 

 confidence our approximate law of assimilation in cases 

 where we are not able experimentally to separate the destruc- 

 tive reactions which accompany constructive phenomena. 



This experiment of the attenuation of the virulence of 

 anthrax bacteridia teaches us yet another thing. 



The presence of phenic acid in the culture-medium not 

 only slackens the multiplication of the bacteridia some- 

 thing easily understood, because phenomena of destruction 

 were added to the phenomena of assimilation. But, more 

 than this, the bacteridia which we obtain have proper- 

 ties different from the original ones they have varied. 

 And, more still, if you next transfer them to a fresh bouil- 

 lon, they will multiply with their acquired variation ; and 



