Production of Heritable Variations in Bacteria 91 



ditions, in which the products of metabolism are not allowed 

 to gather, they usually at once produce the normal amount 

 of mucus ; the change was not a hereditary one. But if the 

 organisms are kept for a long time under the unfavorable 

 conditions (four weeks or more), some produce no mucus 

 at all; and if these are restored to normal conditions, they 

 and their descendants continue to be without the mucous 

 envelope. The change has become hereditary. But it is still 

 not permanent, for by special means the organisms can be 

 caused to begin anew to produce the normal amount of 

 mucus. This is most completely brought about through 

 allowing the organisms to live for a time in a living animal, 

 by infecting a white mouse. After passage through the 

 animal's body the bacteria have regained their normal 

 powers of producing the mucous envelope. 



By long continued cultivation with the products of metab- 

 olism, using special methods, Toenniessen produced other 

 changes that were permanently hereditary. The organisms 

 gradually produced less and less mucus, so that successive 

 gradations could be distinguished. At least three of the 

 grades were independently hereditary; one had a mucous 

 envelope a little smaller than normal; a second had a very 

 small envelope; the third had no envelope whatever. Long 

 continued cultivation under normal conditions left each of 

 these three grades unchanged; even passage through the 

 body of animals did not restore the organisms to the normal 

 condition. The alterations produced were permanently in- 

 herited. 



In this case, as in the former, we observe the striking fact 

 that what seems outwardly the same modification may appear 

 sometimes without being hereditary; sometimes as heredi- 

 tary for a number of generations ; sometimes as permanently 

 inherited. The difference appears to depend on the length 



