16 SCIENCE’ PROGEKESS 
who find it profitable to breed that kind of dog. There is a 
multiplication of the number of dogs breeding, so in the artificial 
propagation of cancer we multiply the number of cells pro- 
liferating. I discuss later the possibility of biological alterations 
leading to increased rapidity of cell growth. Propagation segre- 
gates in the first instance the cells which can accommodate 
themselves to the artificial conditions, and then provides for 
their nutrition and proliferation. Thus there are more of the 
suitable cells in each of the later grafts, and within a given 
period, this primary advantage gives bigger tumours from each 
single graft, and a higher percentage of tumours over the total 
number of animals inoculated. 
A violent change in environment is effected when a tumour 
is removed from the animal in which it developed. The manner 
in which the change is effected is important. Whereas we have 
succeeded in transplanting 66 per cent. of the sporadic growths 
we have discovered, Ehrlich has transplanted 14 per cent. of his 
tumours successfully. The discrepancy is certainly due to the 
difference in method. Asa rule only the tumours best suited to 
propagation survive Ehrlich’s procedure, while we have obtained 
a broader basis for the study of the growth of tumours, since 
our material includes tumours of all degrees of transplantability. 
One group of tumours does not survive the process at all. A 
second group grows for a variable time and then dies, being 
unable to adapt themselves to a strange and presumably an 
unfavourable environment. The tumours of a third group 
are able to adapt themselves gradually. The fourth group 
adapt themselves quickly or at once, and this is the small 
group which are easily propagated. It is possible that among 
them there are tumours surviving transplantation, not only 
because of the inherent properties of their cells, and their 
capacity for adaptation, but because the change in environ- 
ment is in reality to a more favourable soil. In the third and 
fourth groups there must be also more and more cells suited to 
propagation in the grafts at each additional transference, till a 
maximum is reached. Although the group of tumours giving 
maximal success on propagation are of practical importance in 
experiment, because of the ease with which they permit of 
controls to attempts to modify growth, their theoretical value 
depends on their relation to the varied degrees of energy of 
growth exhibited by the whole group of tumours. It is 
