40 WARFARE IN THE HUMAN BODY 



growth, or that its undue increase makes such growth abnor- 

 mally large, we are assuredly dealing with phenomena 

 closely connected with the existence of epithelial or con- 

 nective-tissue neoplasms. In both sets of phenomena the 

 root fact is failure of proliferation or its excess. If we delve 

 deeply enough into causes it will not seem absurd to put 

 cancers and giantism or acromegaly into related sub-classes. 

 That the latter are due to abnormal glandular activity we 

 know. With normal pituitary influence no overgrowth 

 occurs. Connective-tissue proliferation ceases at a point 

 when the glandular system becomes balanced. This is 

 obviously the case with what we see in repair of normal 

 epithelium and connective tissue. When the epithelium is 

 stripped away, and the underlying structures damaged, the 

 connective-tissue cells proliferate rapidly. As the young 

 epithelial cells invade the edges of the wound the under- 

 lying cells become fibrous and deep scar-tissue. Excessive 

 and unhealthy granulations only arise when the epithelium 

 does not do its work. Histologically there is a great like- 

 ness between round-celled sarcoma and granulation tissue, 

 and, after all, granulation is no more than connective- 

 tissue cell proliferation growing outwards into a wound 

 where normal tissues are wanting. A sarcoma might 

 almost be called inverted ingrowing granulations. That the 

 varied phenomena of malignancy exceed in variety those 

 attributable to merely defective or hypertrophied glands 

 is only what might be expected. Such glands are highly 

 specialized epithelium with very definite work. The 

 general epithelium of the body is much less differentiated 

 and nearer the embryonic type. It is found practically 

 everywhere. Connective tissue is the somatic network; 

 in no part is it absent. It exhibits a remarkable capacity 

 for many forms of rapid specialization, and may be looked 



