18 SCIENCE PROGRESS 
natural variations, an increase in its rate may be one result 
of our artificial selection in long-continued propagation. In 
passing I may merely mention it is now well established that 
the amount of proliferation can be influenced in the opposite 
direction—viz. diminished by exposure of a tumour to heat. 
The environment may be of great importance when a tumour 
is propagated for years. Theoretically the long-continued 
growth of cancer cells in the soil provided by the mice of one 
stock or country may handicap them for growth in the different 
soil provided by mice of another; but, whether this be so or 
not, augmented adaptation to an accustomed soil may result 
in an enormously increased proliferation. In this direction 
therefore the cells of later generations may be biologically 
very different from those remotely antecedent. What occurs 
when a tumour is propagated by transference through many 
generations of mice is determined not so much by the number 
of transferences, as by the long duration of the particular 
environment, breeding as it were a cell of particular quality. 
If one follow the process of artificial propagation back step 
by step to the primary animal, there is no reason why one 
should stop there, since the gulf between the growth of 
cancerous tissue and normal tissue after transplantation has 
been bridged over. What takes place in artificial propagation 
may well be but an artificial reproduction of what had long 
been going on naturally in the animal in which cancer 
developed. However, it is inadvisable to pursue such specula- 
tions further at present, for the conditions of the growth of 
cancer in the spontaneously attacked may be different from 
those in normal animals, although the lesions are the same. At 
any rate, the differences between individuals may be of prime 
importance in determining the nature of growth in the primary 
transplantation as well as in the individual spontaneously 
attacked. It seems, at least, that the preceding considerations 
give us a deeper insight into the nature and clinical behaviour 
of the cancers of man during their continued growth, and when 
they exhibit what appears to be a change to a more malignant 
condition. 
The phenomena of growth can all be explained without the 
assumption of microbic interference. And this brings me to the 
further consideration of some matters of clinical importance. 
When the growth of a tumour under artificial propagation is 
