8 Trans. Acad. Sci. of St. Louis 



It seems also that in certain cases a cancer itself may 

 stimulate neighboring tissues which had been normal 

 to assume cancerous growth. In this case we have prob- 

 ably to deal with a chemical stimulus exerted by cancer- 

 ous cells. 



There still remains one kind of cancer to be considered 

 which differs in some respects very markedly from all 

 other kinds which have been so far investigated. There 

 occurs in fowl a cancer of the connective tissue and car- 

 tilage which behaves in its course not unlike the typical 

 cancer; it makes metastases, leads to the death of the 

 animal and can be transplanted into other individuals of 

 the same kind. It differs, however, very markedly from 

 other kinds of cancer in three respects; 1) while in other 

 kinds of cancers, notwithstanding many experiments in 

 this direction, some of them carried out as early as twen- 

 ty years ago, it has not been found possible to separate 

 from the cells an agent which would cause the cancer 

 when injected into suitable individuals, this attempt 

 was successful in this particular cancer of fowl. By 

 filtration through filters so fine that tumor cells cannot 

 pass, the agent was separated from the cells. The same 

 result was obtained through drying the tumor and simi- 

 lar devices which killed the tumor cells without caus- 

 ing a marked injury to the agent. As stated, the same 

 means previously used in other kinds of tumors had led 

 to the result that the inoculability of the tumor had 

 been lost. 



Here then exists an agent distinct from the tumor 

 cells, which is presumably an ultramicroscopic microor- 

 ganism. 2) It is possible to produce in birds an immune 

 serum which protects fowl against the agent, as distinct 

 from the tumor cells. This has never been possible in 

 the case of the typical mammalian cancer. Here im- 

 mimization is always directed against the cancer cells 

 themselves and only succeeds in transplanted, not in 



