172 Dr. Bashfbrd, Mr. Murray, and Dr. Cramer. [Dec. 10, 



A clear conception is required of the necessity for interpreting the results 

 of the experiments about to be described as bearing directly on the condi- 

 tions of growth only and their artificial modification. By inference, they 

 have also important bearings on the nature of cancer ; but we desire to 

 enforce caution in anticipating prophylactic or therapeutic actions as likely 

 to follow the use of those measures we have found efficacious in raising the 

 resistance of animals to transplantation. 



(2) The Importance of Natural Variations in the Eesistance of Mice 

 to Inoculation, and of Inherent Variations in the Energy of 

 Growth of the Tumour Cells. 



In previous communications we have directed attention to variations in 

 the susceptibility of individual mice to inoculation with a number of different 

 mammary tumours of the mouse. The natural insusceptibility to inoculation 

 to which we particularly wish to draw attention in the present communication 

 is that which we have stated was still present after the disturbing influences 

 of race, age, dosage, and fluctuations in the powers of growth of the 

 cancerous cells, have been allowed for in artificial propagation of trans- 

 plantable tumours. In earlier papers we used the natural resistance of the 

 animals as an indicator of the variations of the power of proliferation of the 

 cancer cells. We shall now conversely use the tumour-cells as indicators of 

 the susceptibility of the inoculated animals. In such experiments the 

 number of animals necessary is inversely proportionate to the percentage of 

 successful inoculations attainable. Where the tumour gives a maximal 

 success in normal animals, the number necessary in each experiment can be 

 reduced, and must be correspondingly increased when the percentage of 

 success in normal animals is lower. 



Experiments in which less than 60 per cent, of the normal animals develop 

 tumours are open to grave objections, and will not be employed in this paper. 



Until we can experimentally determine the development of cancer de novo, 

 we are necessarily confined to a study of its growth and the conditions which 

 favour or hinder it when cancerous tissue is inoculated into mice. The 

 differences between mice of different susceptibility are expressed quantitatively 

 in differences in the percentages of successful inoculations. 



In all these experiments we have to take account of two variable factors of 

 prime importance which are beyond experimental control : — 



(1) The varying natural susceptibility of the mice ; 



(2) The varying energy of growth of the tumour cells. 



These vary independently of each other, and under normal circumstances, 



