Maeoh 3, 1916] 



SCIENCE 



295 



corpus luteum in the origin of cancer in 

 mice to which we referred above, but other 

 internal secretions will probably be found 

 to be of similar significance. External 

 mechanical factors can be recognized in the 

 well-known effect of chronic irritation. 

 How far certain parasites, especially those 

 in the class of vermes and insects, produce 

 cancer through chemical and how far 

 through mechanical means is not certain. 

 Previous observations in man in the case of 

 cancer of the bladder caused, directly or 

 indirectly, by bilharzia, and especially the 

 recent experiments of Fibiger make it, how- 

 ever, quite certain that such parasites may 

 be the cause of cancer. It is likewise un- 

 certain how far Eoentgen-ray cancer, fre- 

 quent in Eoentgen-ray operators, and also 

 apparently experimentally produced in a 

 few rats by Marie, is due to ulceration sub- 

 sequent to exposure to or to the direct stim- 

 ulating action of the rays. In some cases 

 perhaps chemical and mechanical factors 

 may cooperate in producing tumors; the 

 efficiency of such a combination in calling 

 forth tumor-like formations has been shown 

 by us in the case of deciduomata of the 

 uterus, which we produced experimentally, 

 a new formation which we included in a 

 class designated as transitory tumors. 



There are observations on hand which 

 indicate that growth stimuli may be espe- 

 cially active in animals with a hereditarily 

 determined tendency to cancer. Such an 

 observation we made in the case of a cancer 

 in a mouse belonging to a family rich in 

 tumors, where ulceration of the skin near 

 an adenoma of the mammary gland led to 

 the development of an epidermal carcinoma. 

 Further systematically conducted experi- 

 ments in this direction might lead to inter- 

 esting results. 



It is, however, not probable that in order 

 to obtain the production of cancer there 

 must be a definite quantity of prerequisite 



internal factors. On the contrary, there is 

 some evidence on hand which makes it 

 probable that internal and external factors 

 may vary in inverse ratio, and that if the 

 external factors are quantitatively very 

 strong, the quantity of internal factors may 

 be reduced. 



If we survey briefly the various types of 

 growth reactions known in vertebrates, we 

 may perhaps, according to the character of 

 the stimuli, which are usually in each case 

 the first members in a complicated reaction 

 chain, and according to the character of the 

 systems on which they act, provisionally 

 distinguish the following types : 



1. Various stimuli act for a short time on 

 complex systems, the egg-cells, and lead to 

 a long chain of growth phenomena which 

 ultimately cease. The experiments in arti- 

 ficial parthenogenesis of Jacques Loeb led 

 to a very far-going analysis of these phe- 

 nomena. 



2. Defects lead to a chain of growth phe- 

 nomena, which are of a temporary char- 

 acter, and which come to a standstill after 

 a certain quantity and kind of new-formed 

 tissue has more or less completely filled out 

 the defect. 



3. Chemical substances stimulate the 

 growth of certain tissues to which they bear 

 a more or less specific relation. These 

 growth phenomena come to a standstill with 

 the activity of the stimulating substance or 

 very soon afterwards (corpus luteum and 

 mammary gland) . 



4. A combination of factors 2 and 3 

 may lead to tumor-like growth phenomena 

 when either factor alone would cause only 

 a slight proliferation. Here again the 

 effect is temporary (experimental deciduo- 

 mata of the uterus). 



5. Chemical (fat soluble?) bodies which 

 do not show a specific relation to the organs 

 affected stimulate various tissues to a tem- 

 porary proliferation; fat soluble stains 



