THE STUDY OF CANCER 23 



passu with prolongation of growth (anaplasia), and, for 

 apparently identical tumours, by assuming differences in the 

 soil afforded by different individuals, and also by the several 

 organs of any single individual. At the outset of our investi- 

 gations, therefore, the growth of cancer under natural and 

 experimental conditions was selected for study, as covering 

 ground common to all the varied manifestations of tumours. 

 The experimental study of the growth of cancer proved the 

 permanence of the histological characters, and revealed that the 

 cells possessed other qualities of a more or less permanent kind, 

 escaping demonstration by other means. At the same time, 

 experiment shed light on the relations subsisting, on the one 

 hand, between the different properties of cells similar morpho- 

 logically, and on the other hand the soil in which they are 

 situated. Only by experiment has it become possible to study 

 the growth of one tumour in different soils, and, what is equally 

 important, the growth of different tumours in the same soil. In 

 this way the varying clinical and histological features of the 

 cancers of the mouse's mamma, reproducing as they do in 

 miniature all the difficulties encountered in man, have been 

 relegated to subsidiary importance for the present, in favour of 

 the study of the problem of cell assimilation, which, being 

 common to all these tumours, permits us to unite in one general 

 conception all the gradations and mutations exhibited in the 

 growth of cancer. The experimental study of growth has 

 already enabled us to compare tumours together as regards 

 their biological behaviour (clinical behaviour). It has been 

 proved that cancer grows better in young mice than in old ; 

 that some mice of the same age afford a more suitable soil than 

 others ; that some sites of the body favour growth while others 

 hinder it ; that tumours seemingly identical (when examined by 

 the aid of other methods) can be differentiated from one another, 

 since they possess distinct properties by virtue of which they 

 behave differently in the same soil ; that one and the same 

 tumour exhibits increased proliferation alternating with 

 diminished proliferation in a similar nutritive environment, and, 

 this being so, the fluctuations in proliferation are possibly 

 expressions of an alternation in the assimilative powers of the 

 cells of the tumours which exhibit them. Experimental pro- 

 pagation further showed that proliferation can be maintained by 

 normal bipolar division and that other forms of cell division are 



