TEs WUD OF CANCER 21 
As I pointed out on a previous page, the cancer cells acquire at 
each transference a specific stroma, the connective tissue 
reaction they call forth in the host. A culture of isolated cells 
has not been obtained, since growth always proceeds in the 
form of a tissue with intimate vascular supply. We emphasised 
the distinction thus revealed between the cells of different 
growths nearly allied histogenetically, and the parallel it 
presented to the biological reactions whereby blood relationship 
has been established. At the same time we pointed out its far 
greater delicacy. In what I have briefly outlined with regard 
to protection the same delicacy, or specificity, appears. The 
cancer cell is dependent on the provision by the host of a 
connective tissue and vascular reaction, if it is to grow into 
a tumour. It would appear to be a piece of short-sighted 
generosity on the part of the host to supply connective tissues 
to subserve merely the needs of the greedy cancer cells, and yet 
this appears to be the relation which subsists when a tumour is 
growing well. The question arises, is the connective tissue 
reaction supplied only in response to the needs of the cancer 
cells, or is it really protective, as in the case of the similar 
reactions elicited by the tubercle bacillus or other organisms? 
When a tumour undergoes spontaneous absorption, the 
connective tissue reaction does become protective, and the 
acquisition of phagocytic properties by its cellular elements is 
the main factor in removing the remains of the tumour. Thus 
under certain conditions the organism is certainly capable of 
defending itself very actively through reaction to the presence 
of the cancer cells, and after the process remains protected. 
The conclusion that phagocytosis is the only agency by which 
protection is conferred would in my opinion be unwarranted 
at present. Spontaneous absorption appears to be frequently 
associated with a phase of depressed growth of the tumour 
cells themselves. We do not yet know exactly whether the 
cancer cells require to undergo changes—either spontaneous, 
or induced from the side of the host—placing them at times at 
the mercy of the connective tissue which, at other times, appears 
to be their willing servant. When a graft is introduced into 
a protected mouse, little, perhaps no, proliferation follows, and 
the tissue speedily dies. Whether it is killed by the fluids of 
the body, or simply starved to death, because it is prevented 
from exercising its chemiotactic influence on the connective 
