132 Lectures on Bacteria. [§ xn. 



during two years by Koch, Gaffky, and LoflBer without any 

 change or any return to virulence. 



The behaviour of Bacilli which have been rapidly attenuated 

 at 47-50° C. and at still higher temperatures is different; they 

 recover their virulence when cultivated under the most favourable 

 conditions. 



It is true that a return from the attenuated to the virulent 

 condition is not altogether excluded, even in forms which have 

 been slowly attenuated. Pasteur aflSrms that if matter, which is 

 not fatal to full-grown guinea-pigs but kills young individuals 

 a day or two old, is taken from one of the latter and used to 

 inoculate other guinea-pigs successively older, a degree of viru- 

 lence is ultimately attained sufiBcient to kill old animals. Koch 

 and his colleagues have not found these statements confirmed by 

 their results, and though their experiments were somewhat differ- 

 ently arranged, yet they seem to show that no such fixed rule ob- 

 tains in this matter as might be gathered from Pasteur's account. 

 On the other hand the same observers have distinctly proved 

 that there is a return to increased virulence in individual and 

 not very similar cases. 



Lastly, they have also proved that cases of the reverse kind 

 occur, in which the virulence of a culture suddenly diminishes 

 of its own accord, that is without any ascertained external cause; 

 spores from material which killed rabbits and guinea-pigs pro- 

 duced eight weeks later a generation which did not injure these 

 animals but was still fatal to mice. To this class of phenomena 

 belongs perhaps an observation recently communicated by 

 Prazmoswki, according to which a pure culture of Bacillus 

 Anthracis in a nutrient solution entirely lost its virulence with- 

 out apparent cause. I have myself seen the same or a similar 

 case. Buchner's investigation, which bears on this point, will 

 be noticed presently. 



I have already observed that the form of the Bacillus remains 

 unchanged in the virulent and attenuated states. In the main 

 points this is always the case, though some modifications have 



