June 23, 1916] 



SCIENCE 



885 



subepidermal region as an invasive hyper- 

 plasia. The punctures were deep enough, 

 however, to have infected the subglandular 

 connective tissue which is also proliferating. 

 The sections were cut at the end of 27 days 

 and show transitions from a columnar 

 (glandular) epidermis into an irregular, 

 angular-celled, large nucleate, deep-stain- 

 ing mass of rapidly multiplying atypical 

 cells corresponding to an epithelioma 

 (slides) . The shape of these cells is exactly 

 that of proliferating epidermal cells from 

 ™y ■Koo mm- deep 72-hour inoculations on 

 tobacco stems. I have not yet obtained 

 metastases from such surface growths, but 

 I am only now beginning my studies of 

 skin and gland proliferation and there is 

 much to learn. 



We now come to embryomas. Before 

 describing the atypical teratoid tumors I 

 wish to make some general remarks. Con- 

 ceiving human and animal cancer to be due 

 to a parasite, I have been greatly interested 

 for the past ten years to see to what extent 

 the phenomena of such cancers, the cause 

 of which is unknown, can be paralleled by 

 crown gall phenomena the cause of which 

 is an intracellular schizomycete. By dis- 

 covery of a tumor-strand and of stem struc- 

 ture in leaf tumors (in 1911) my interest 

 received a tremendous accession from which 

 it had not yet recovered when the newer 

 discoveries of this winter converted it into 

 a white heat ! I am now persuaded that the 

 solution of the whole cancer problem lies 

 in a study of these plant tumors. At least 

 they must now be studied until the matter 

 is definitely settled, pro or con. 



If cancer is due to a microorganism, bac- 

 terial or other, we are not obliged theoret- 

 ically to conceive of all such new growths 

 as due to one and the same parasite, nor, 

 indeed, on first thought, is such the more 

 probable hypothesis. The first thought is 

 that probably there must be as many para- 



sites as there are kinds of tumors, yet cer- 

 tainly, on further reflection, the mere cell 

 differences between a sarcoma, let us say, 

 and a carcinoma do not necessarily involve 

 the conception of two parasites. The two 

 tumors can be explained theoretically .just 

 as well by the postulate of one parasite, 

 and in the light of our researches on crown 

 gall much better iy one. If the tissue re- 

 sponse depends on the kind of cell or cells 

 first infected, as apparently it must, on the 

 assumption of a parasitic origin, then, of 

 course if connective tissue cells only are 

 involved, we shall have sarcoma; if gland 

 cells only are invaded we shall have carci- 

 noma; or if both, then a tumor containing 

 both types of cancer. Whichever cell was 

 first invaded (the bacteria being impris- 

 oned) would be likely to continue its pro- 

 liferation as a tumor of a pure type, but 

 other elements might eventually become in- 

 fected by a surgical operation, or other- 

 wise, e. g., a sarcoma might follow a carci- 

 noma as in some mouse tumors, and also in 

 man, the connective tissue stroma becoming 

 infected. 



I now think the human embryonal tera- 

 tomata are cancerous not only potentially, 

 but actually from the beginning. Many 

 of them have been recognized to be so on 

 removal, and in the remainder the stimu- 

 lating blastomous portion may have re- 

 mained undiscovered owing to its rela- 

 tively small size, as was the case in hairy 

 root of the apple (and every particle of 

 such a tumor would have to be sectioned 

 and studied before one could deny it), or 

 it may have receded during the rapid de- 

 velopment of the non-blastomous purely 

 teratoid portions. All of them, whether it 

 be assumed that they have developed from 

 "cell-rests" or parthenogenetically, are, I 

 believe, due to the stimulus of a micro- 

 organism, but not necessarily of a schizo- 

 mycete, since other orders of parasites may, 



