164 
ing difference between granulomata and 
cancers, but the true nature of the can- 
cerous development becomes more evident 
in the secondary tumors. The mere fact 
that a primary cancer has developed on 
some part of the body does not constitute 
the chief danger, since one might have 
such a tumor for a long time without death 
supervening, unless the primary growth 
happened to be situated in or near a vital 
organ. What constitutes the peculiar ma- 
lignaney of cancer is the tendency to form 
secondary growths in various parts of the 
body, including the vital organs, and it is 
this clearly recognized danger which 
in modern times has led to the universal 
recommendation on the part of competent 
physicians and surgeons of the early ex- 
tirpation of suspicious growths, the hope 
being that the surgeon may be able to dis- 
sect out all the infected tissues and thus 
free the patient from the disease. This is 
the reason why, for instance, in cancer of 
the breast the surgeon so carefully removes 
not only the infected breast, but the lym- 
phatics for long distances away, that he 
may, if possible, reach beyond the unseen 
growing cancer strands. This also is why 
delayed operations for cancer are seldom 
successful. 
In ease of granulomata, as we have seen, 
it is the parasite which migrates. In case 
of cancers it is the cancer cell itself which 
migrates, 7. e., some of the body cells which 
under some unknown stimulation have 
been taken out of the physiological con- 
trol of the body and have become thus, as 
it were, parasites on their fellow cells. 
There are two ways in which secondary 
tumors are derived from the primary 
tumor in cancer: (1) The primary tumor 
growing peripherally sends out roots or 
strands which bore their way through 
normal tissues of the body, sometimes for 
long distances, developing from certain 
SCIENCE 
[N.S. Vou. XXXV. No. 892 
portions of these strands secondary tumors. 
(2) Small groups of cancer cells are dis- 
lodged from the parent tumor and carried 
as floating islands in the blood stream or 
lymphatics to develop secondary tumors 
where they lodge. The first of these ways 
has been definitely established by observa- 
tion; the second by inference, no connect- 
ing strand having been discovered. Nat- 
urally these secondary tumors, being de- 
rived from the primary tumor, tend to par- 
take of the nature of the tissue from which 
the primary tumor has developed. For ex- 
ample, if the primary tumor be in the 
stomach, the secondary tumors are likely 
to contain glandular cells resembling those 
of the stomach, wherever they may be de- 
veloped. This is such a striking peculiarity 
that it is often possible for the animal 
pathologist to tell from the study of his 
sections whether the cancer is primary or 
secondary, and, if secondary, in what organ 
the primary tumor is located. In case of 
tumors located in an organ containing all 
three of the embryonic layers or developed 
out of cell-rests of this nature we might 
have in the tumors a jumbled-up mass of 
all sorts of tissues—skin, bone, teeth, hair, 
muscle, nerve, ete. This at least is one 
method of explaining the embryomata. 
Having found no parasite in the cancer 
cells, a majority of the animal pathologists 
have given up the idea that cancer can be 
of parasitic origin. For a generation the 
research workers fell back upon Cohn- 
heim’s hypothesis that cancers were due to 
the development of small fragments of 
tissue cut off from the parent layer during 
embryonal growth, to be enclosed in other 
tissues and lie dormant until acted on ab- 
normally later in life by some unknown 
stimulus. But while studies of the animal 
body show that such separation of small 
portions of tissue from the germinal layer 
is not uncommon, research workers on 
