22 SCIENCE PROGRESS 



tissues, is a problem too hypothetical for discussion here ; it 

 will suffice to say, that interference with the chemiotaxis of the 

 cells of tumours of different kind would explain the phenomena 

 thus far observed, their specific and their common features. 



Is the protection to be interpreted as due to some virus 

 contained in the cancer cell ? I think not. We have pointed 

 out how mere inhibition of the connective tissue reaction would 

 suffice to prevent growth, since its specificity indicates a delicacy 

 of nutritive requirements on the part of the cancer cell far 

 exceeding that of any other known biological reaction. We 

 found that protection is conferred on the mouse only by mouse 

 tumour, not by the preceding inoculation of the tumours of 

 alien animals even so nearly related as the rat. One form of 

 mouse tumour protects against another. We found that pro- 

 tection is conferred on the mouse by the normal tissues of the 

 mouse and not by those of strange species. There is thus 

 something common to mouse tissue and mouse tumour which 

 is not common to mouse tumour and the tumours of other 

 animals. The protection cannot therefore be due to the presence 

 in the tumour cells of a virus common to vertebrate cancer. 

 Reviewing our experiments as a whole, I am of the opinion 

 they establish that the cancer cell is really a cell of the mouse 

 organism requiring the same food, the same kind of connective 

 tissue and blood supply, to which it was accustomed in the 

 mouse in which it developed. The only difference it exhibits 

 is a qualitative one, viz. in the powers of growth and assimila- 

 tion. Fortunately the cancer cell is very much at the mercy of 

 subtle changes in its environment, and it is by taking advantage 

 of this fact that we have been able to prevent its growth in 

 the living mouse. 



The study of cancer in man became increasingly difficult in 

 the course of time because of the mere multiplication of facts 

 defying classification. Malignant new growths in man and in 

 animals — even those of any particular organ and histogenetically 

 related — exhibit an extraordinary variety in their histology and 

 clinical behaviour. This variability is made yet more difficult 

 to comprehend by the inconstancy in the behaviour of tumours 

 histologically indistinguishable. Many attempts have been 

 made to group the phenomena, e.g. by pushing the histogenetic 

 subdivision of tumours to absurd lengths, by assuming that 

 progressive loss of histological differentiation proceeds pari 



