TRANSACTIONS OF SECTION D. 599 
Malignant new growths, in common with benign, increase their characteristic 
parenchyma entirely from their own resources. As soon as a malignant new 
growth is recognisable as such, it is marked off anatomically and physiologically 
from its surroundings. This phenomenon, now well established, is sufficiently 
remarkable when it is borne in mind that, histologically, the independent tissue 
may be indistinguishable from that among which it takes its origin. To a 
recognition of this fact is due the acceptance accorded to Cohnheim’s hypothesis 
and all its variants. These variants were introduced because of the necessity that 
was felt to account for the close dependence of the type of growth on the 
characters of the surrounding tissue, especially when the latter presents well- 
marked differences at different periods (Thiersch, Ribbert). They are all attempts 
to account for the behaviour of malignant new growths as independent new 
organisms, and, whatever acceptance we may accord to the various hypotheses, 
the fact they seek to explain is incontrovertible. In discussing the experimental 
investigations some reasons for considering these hypotheses as inadequate will be 
referred to, 
The cells of malignant new growths increase in number by division. Amitosis 
certainly occurs, but mitotic division is by far the more common, especially in fully 
developed tumours. Multipolar mitoses are common, but not universal, The 
active growth and extension of the malignant tissue, as manifested at the growing 
surfaces of the malignant new growths we have so far examined, is effected by 
cell-divisions, which, so far as they are mitotic, conform to the ordinary type met 
with in early development. Apart from multipolar divisions, the number of 
chromosomes entering the equatorial plate is found constant in each species, and 
they undergo the ordinary longitudinal splitting. Passing from the growing 
margin towards the older parts of the growth two sets of changes occur. Many 
cells undergo the characteristic histological changes peculiar to the tissue among 
which the tumour has arisen, while others prepare for further mitosis. In some 
of these the resulting mitosis is characterised by the presence of bivalent 
chromosomes (heterotype), in number half that found in the younger parts. From 
the position of these heterotype mitoses in relation to the growing surface of the 
tumour in which they occur, they must be regarded as a late phenomenon in the 
life history of the cells, contemporaneous with the histological differentiation going 
on around them. We have not found evidence of continued proliferation of the 
immediate descendants of the heterotypical division, and the analogy of animal 
spermatogenesis suggests that the heterotype initiates a terminal phase in the life 
history of the cancer cell as in the spermatocyte. 
From a consideration of these facts the most divergent conclusions have been 
drawn. Professor Farmer and his colleagues, who first described the occurrence 
of heterotypical mitoses in malignant new growths, consider that we have here 
a transformation of somatic tissue into a kind of reproductive, ‘gametoid,’ 
tissue, which, gua its gametoid character, is independent and capable of unlimited 
further growth. This view of the nature of the change which marks the distinc- 
tion between somatic tissues and malignant new growths had already been advanced 
by Beatson as a result of clinical considerations. Against this view the general 
objection may be raised that, while it would explain the occurrence of hetero- 
typical mitoses in malignant new growths, as regards the powers of growth and 
self-propagation and independence which they manifest it is no explanation at all. 
In the vertebrates, where, until now, malignant new growths have alone been 
found, we have no evidence that gametes, or the tissue which precedes them, 
possess powers of growth at all comparable to those seen in cancer. The analogies 
drawn from the vegetable kingdom all concern the interaction of independent 
organisms, not of different tissues of the same organism. 
When, however, the attempt is made to attack this problem by experiment, and 
artificial propagation of reproductive tissue in animals is tried, the results are in 
no way different from those obtained with other tissues. The power of in- 
dependent growth is very limited, and it is found that the power of regeneration 
of which the testis is capable (along with the thyroid) is confined to the stages 
before differentiation of eametes has commenced. 
