320 ON GROWTH AND OVERGROWTH 



conditions tumours originating from the Malpighian layer of 

 the skin can in their growth lose their lepidic characters, i.e. 

 can become possessed of a definite stroma passing between the 

 individual cells, then we shall have to acknowledge that epiblastic 

 lepidomata can, in the course of their growth, revert to a more 

 undifferentiated sarcomatous type. 



In one very interesting tumour of the prostate, examined by 

 me some years ago, I was for a time of opinion that I could 

 recognize such a transition. Sections taken from the prostate 

 itself showed beautifully the existence of a very typical carcinoma 

 almost scirrhous in appearance ; passing towards the bladder 

 the type became that of carcinoma simplex. The abundant, 

 apparently rapid, growth, forming a projection into the bladder 

 itself, was so wholly cellular as to resemble a round-cell sarcoma, 

 but more careful examination showed that even here the cells 

 were arranged in alveoli, although the stroma between the 

 individual groups was so delicate as to be traced with considerable 

 difficulty. 



46. With mesothelial and endothelial lepidomata the same 

 is not always so ; the older or earlier portions of the tumour 

 may show distinct adenomatoid or carcinomatoid characters 

 occasionally the whole tumour is typically adenomatous ; but 

 more rapidly growing portions are peculiarly liable to depart 

 so far from type, so peculiarly liable to take on the appearance 

 of embryonic connective tissue, that it becomes impossible, 

 basing our terminology upon histological appearances, to say 

 whether we are dealing with a carcinoma, or a sarcoma, or a 

 mixed growth a carcinoma sarcomatodes, or a sarcoma car- 

 cinomatodes. More particularly the increased recognition of 

 the frequency of endotheliomata and peritheliomata has forced 

 us to see the difficulties in our present mode of classification. 

 The perithelioma when developing characteristically, apparently 

 as an endothelioma of the perivascular lymphatics, may strongly 

 resemble an adenoma in the regular columnar arrangement of 

 its cells, and yet other parts of the same tumour may be absolutely 

 sarcomatous in type. And while the ordinary endothelioma, 

 such as one so commonly meets with forming tumours in con- 

 nexion with the membranes of the brain, is in general char- 

 acteristically sarcomatous in structure, areas are to be detected 

 here and there indicating its origin as a squamous proliferation 



