Maech 3, 1916] 



SCIENCE 



299 



efforts of this kind do not appear promising 

 at present. There seem, however, still other 

 ways open through which one may ap- 

 proach this problem in an indirect manner. 



To decide between the two alternatives 

 which we mentioned does not only concern 

 cancer research in the more restricted sense, 

 but is of the greatest importance for gen- 

 eral biology. In return for much that it 

 received from neighboring sciences, cancer 

 research has given something important to 

 biology; the serial endless experimental 

 propagation of tumors has enriched biology 

 with a valuable instrument of research and 

 new outlooks on the life and character of 

 somatic cells have been gained. We may 

 briefly mention the following facts estab- 

 lished or very strongly suggested: In the 

 course of our early transplantations, we 

 found that the energy of tumor growth 

 can be experimentally increased as well as 

 decreased. Ehrlich explained the increase 

 as due to a selection of rapidly growing 

 tumors; we, however, believed from the 

 beginning that it was partly produced by a 

 mechanical stimulation of the tumor cells 

 and in addition was possibly due to chem- 

 ical stimulation caused by the transfer into 

 a new host with a different constitution of 

 the body-fluids ; in some cases perhaps proc- 

 esses of immunity may also enter into this 

 phenomenon. 



In conjunction with M. S. Fleisher, we 

 noted that chemical bodies which inhibit 

 tissue growth at a certain period in the life 

 of tumors do not have this power at other 

 periods. Especially are they powerless in 

 the case of very young tumors, an observa- 

 tion confirmed by Keysser. But we found 

 that such early injections produce an im- 

 munization against the later action of these 

 substances. The proof thus given that an 

 immunization takes place against substances 

 (and apparently also against physical 

 agencies) inhibiting tumor growth is, as we 



pointed out on previous occasions, of great 

 importance in our attempts to arrive at a 

 rational treatment of cancer. Our experi- 

 ments suggest, furthermore, very strongly 

 that this immunity is of a twofold char- 

 acter, that it originates in the host as well 

 as in the tumor cells themselves; that this 

 cell immunity can be transferred to a cer- 

 tain number of later cell generations and 

 is to some extent specific for the substance 

 which had called it forth. While our re- 

 sults, based on the observation of a very 

 large number of animals, strongly suggest 

 these latter conclusions, we nevertheless 

 think it desirable to add new evidence in 

 order to guard against a complication with 

 variable factors. 



Are we in all these cases dealing with 

 indirect actions on the cells and with direct 

 actions on accompanying microorganisms, 

 or with direct actions on the cells? We 

 rather incline to the latter view and we 

 would suggest that an increase in chemical 

 activity in the tumor cells — an increase 

 perhaps restricted to certain activities — 

 renders the latter a much finer balance in 

 their response to certain environmental 

 conditions through variations in growth 

 energy than are the normal tissue cells. 



As we pointed out in 1901 on the basis of 

 Morau's and our own experiments, cancer 

 cells are potentially immortal in the same 

 sense in which Protozoa and germ cells are 

 potentially immortal. All, or at least the 

 large majority of all normal tissue cells are 

 potentially cancer cells, and we may there- 

 fore with full justification conclude that 

 ordinary somatic cells are likewise poten- 

 tially immortal. Like the majority of 

 tumors, they can not be indefinitely propa- 

 gated in other individuals of the same spe- 

 cies because of the injurious action of what 

 we may term homoiotoxins. On the other 

 hand, thanks to their increased growth 

 energy and perhaps a lessened sensitiveness 



