272 THE EVOLUTION OF CONTINUITY 



of granulations the marginal epithelium has dipped down 

 in an attempt to cover the wound ; that as the epithelial 

 processes grew longer, not only did their advancing ends 

 fail to find normal restraining granulations, but such in- 

 fluences as originally passed from the healthy connective 

 tissue near the ulcer would become weaker and weaker till 

 they were not felt. For normally the connective tissue 

 restrains a few layers of epithelium, and could not be 

 expected to control a process equivalent to a hundred layers 

 in depth. 



In the light of Champy's experiments one other possi- 

 bility suggests itself, namely, that epithelial cells might 

 break away from the margin and multiply, as it were, in 

 culture in the serum of the ulcer. 



All carcinomata do not develop in elderly people, though 

 these are specially liable to be attacked, and there is not 

 always an ulcer preceding the tumour ; but where the 

 tumour arises in normally arrested tissues it is probable 

 that it always follows on a breach of Continuity of some 

 sort, due to physical injury or inflammation. 



There is this peculiarity about all epithelium, and 

 especially such as is glandular : while we may say that as 

 a tissue it is permanently arrested in the sense that it never 

 gives rise to sexual elements, its cells, under normal condi- 

 tions, are constantly being released from the arrest of 

 multiplication, owing to friction, wear and tear, etc. More- 

 over, the very function of glandular epithelium entails 

 constant cell loss and fresh cell-division. Thus, epithelial 

 tissue as a whole is not so rigidly arrested as the other 

 tissues of the body, and we might suppose that on this 

 account it was specially apt to break away from control, 

 and that with advancing years its confirmed " habit " of 

 cell-multiplication became a danger to the body. 



We have taken epithelial tumours as examples of com- 

 plete escape from the control of Natural Arrest, but it may 

 well be that at times connective-tissue cells may also break 

 away from all control and originate sarcomatous tumours. 



An interesting question is that of the stroma of certain 

 cancerous tumours. This could sometimes be interpreted 

 as no more than cirrhotic connective-tissue due to the 

 breaches of connective-tissue continuity caused by the 



