THE CAUSATION OF DISEASE. 345 



are agreed that the primary carcinomata increase, for the most 

 part, by an actual transformation into malignant tissue of the 

 tissue primarily affected, and in no case is this more beautifully 

 shown than in carcinoma of the kidney. But some patholo- 

 gists doubt whether the sarcomata ever originate in this way, 

 the tumour growing, in their view, simply and solely by a 

 multiplication of the few cells initially constituting it. 



It is, with our present knowledge, almost impossible to decide 

 this question. Inasmuch, however, as all are agreed that primary 

 carcinoma is a transformation — at all events, in part — there is 

 no need to assume in carcinoma a group of previously existing 

 embryonic cells, as some have done, because we have very distinct 

 proof that the tumour is, to a large extent, developed from 

 mature tissue. Although many of the cells constituting a 

 primary carcinoma are derived from the normal cells of the 

 parts, yet there is not the slightest doubt that very many of 

 them arise from a multiplication of the cells already con- 

 stituting the tumour. Indeed, this seems to me to be more 

 certain of the carcinomata than of the sarcomata, for connec- 

 tive-tissue elements are everywhere present ; therefore, in no 

 matter what direction sarcoma spreads, or in what parts it is 

 secondarily produced, the proper elements are always at hand 

 to share in the formation of the growth ; but it often happens 

 that carcinoma spreads to tissues, either directly or by secondary 

 reproduction, in which there are no epithelia — the growth in such 

 cases evidently occurring by a multiplication of cells already 

 belonging to the tumour. Squamous epithelioma, to wit, may 

 directly extend to bone, or it may be secondarily reproduced 

 in a tissue containing no squamous epithelia. The reason 

 why the primary tumour contains squamous epithelial cells is 

 certainly because the tissue in which the growth started is 

 largely made up of such cells. The tissues secondarily affected 

 contain none such. Now, if the cancer had first attacked the 

 tissues thus secondarily affected, the tumour would certainly 

 not have contained squamous cells, for a primary squamous 

 epithelioma never starts in tissues devoid of these cells. From 

 this, two conclusions seem to follow — (1) The primary tumour 

 derives its squamous epithelial cells from previously existing 

 similar cells ; (2) the squamous epithelium of the secondary 



