338 TUMORS. 



mucous tissue, and the new growth of gland epithelium simply keeps 

 pace with the growth of the stroma which it covers. Such growths are 

 sometimes classed among the adenomata, but do not, strictly speaking, 

 belong there. 



Many of the adenomata are benign tumors, being slow of growth 

 and localized, but there are very important exceptions. Some of the 

 adenomata of the stomach and intestines belong to the most malignant 

 of tumors in rapidity of local extension, in the formation of metastases, 

 and the development of cachexia. Certain of the adenomata of the 

 uterus, mamma, and thyroid are also very malignant. It should be re- 

 marked, however, that, as a rule, the malignant adenomata are those 

 which, in structure, lie close upon the border line between tumors of 

 this class and carcinomata (see Fig. 184), and by such observers as in- 

 cline to lay more stress upon clinical than morphological distinctions 

 they are usually classed among the latter. 



II. CARCINOMA. 



The tumors of this highly important class are composed, as above 

 stated, of a connective -tissue stroma, forming more or less well-defined 

 communicating spaces or alveoli, in which lie variously shaped epithelial 

 cells arranged in an atypical manner. The stroma, containing few or 

 many cells, may be, especially in the advancing portions of the tumor, 

 composed largely of the old connective tissue of the part. It may, how- 

 ever, be entirely new formed. The cells which lie in the spaces or alve- 

 oli sometimes bear a very close, but sometimes only a general resem- 

 blance to epithelium (Fig. 182). 



Within the last twenty years the opinion has found general accept- 

 ance that new epithelial cells in tumors, as well as under other conditions, 

 normal and pathological, arise after embryonic life exclusively from old 

 epithelium. 



The apparent occurrence of primary epithelial tumors in parts of the 

 body in which epithelium is not normally present, as in bone and the 

 lymph-nodes, has been recorded ; but these may have been metastatic 

 tumors, in which the primary growth was small and overlooked, or they 

 may have been displaced embryonic remains, which, according to Cohu- 

 heim's hypothesis (see page 298), would explain their heterologous oc- 

 currence. These possibilities of error should be taken into the account 

 in the apparently exceptional cases, and it is to be remarked that they 

 are becoming less and less as our knowledge increases and our technical 

 facilities for research improve. 



A considerable number of the tumors formerly described as heterolo- 

 gous primary carciuomata are now known to be formed by proliferation 

 of eudothelium, or mesothelium, * and hence to belong to another class 

 endothelioma although sometimes considerably resembling the carci- 

 nomata in structure. 



1 See reference Minot, p. 303. 



