302 Conditions for Tumorigenesis 



other hand become cancerous, and, as already mentioned, is transplantable. 

 Important data are available, however, indicating that things are not so clear- 

 cut as originally supposed with reference to the autonomous growth of these 

 tumors induced by hydrocarbons. Mider and Morton^" reported regression of 

 cutaneous papillomas elicited in mice by a single painting with 0.5 per cent 

 of methylcolanthrene. Rous and Kidd*"' have recently described at length the 

 transient condition of many of the new growths induced in the skin by tarring 

 of the ear of the rabbit. They not only disappear when tarring is stopped, but 

 they can even do so while tarring is continued. Similar statements were re- 

 cently made by Flory^* regarding papillomas elicited by tobacco tars. Our 

 own observations with the guinea pig (unpublished results) give evidence of 

 regression of hyperkeratosis of the skin induced by painting with benzpyrene, 

 though the painting was maintained. This has recently been corroborated by 

 Miss Moreira in this laboratory. Examples of regression of papillomas or 

 "carcinomatoids," to use the terminology of Rous and Kidd, are found in 

 the original work of Yamagawa and Ichikawa as early as 1915-1918 (quoted 

 from Rous and Kidd^^). In regard to sarcoma due to carcinogenic hydrocarbons 

 it may be noted that Des Ligneris^° observed that the new giowths started as 

 simple granulation tissue and became sarcomatous only subsequently. 



Such results with tumors induced by tarring or painting with carcinogenic 

 hydrocarbons remind us that not too much stress can be laid on regression 

 of newgrowth elicited by estrogens. One must suppose rather that there are 

 two different phases in experimental production of tumors: the first one is 

 that of active, atypical proliferation of cells which are as yet physiologically 

 normal; the second phase is that of transformation into cells capable of 

 autonomous growth. In the case of carcinogenic hydrocarbons the first phase 

 is in general, though not always, succeeded by the second. With estrogens the 

 first phase is in general not succeeded by the second; but this depends ap- 

 parently upon experimental conditions,— all the more so as under certain ex- 

 perimental conditions even the second phase can become independent from 

 the first one, as shown by Rous and Kidd" with tar warts, where newly ac- 

 quired "neoplastic potentialities" may remain in abeyance or dormant. 



4. Differential Behavior of Homologous Territories in Different Species. 

 When one compares the response of the skin in mice and guinea pigs toward 

 carcinogenic hydrocarbons it is evident that the same, or homologous, terri- 

 tories react differently in different species. This is true also for tumorigenesis 

 resulting from estrogens. Comparative results with prolonged treatment with 

 estrogens in guinea pigs and rats are given in table i. 



Pluristratified metaplasia and subsequent cornification of the endometrium 

 may be seen in the rat treated for three months with estrogens. In contrast 

 to this, cornification was never seen under similar experimental conditions in 

 the guinea pig. This was true even when treatment with great quantities was 

 prolonged for a year (see pi. 1, fig. 1 and pi. 3, fig. 9) and even though pluri- 

 stratified metaplasia of the endometrium was easily established in the guinea 



