Marcw 29, 1901.] 
ducts, but the liver-cells are also involved, 
but how it comes about that the liver-cells 
participate is, so far as I have been able to 
learn, by no means clear. From analogy 
with other tissues, we infer that it is im- 
probable that the large and specialized liver- 
cells ever resume an embryonic character. 
In short, I deem our understanding of the 
pathological differentiation of hepatic cells 
and of striated muscle fibers too imperfect 
to support a judgment. We can only say 
that the rarity of such differentiation con- 
cords with the degree of normal specializa- 
tion of the cells and fibers in question. 
Our very brief discussion of pathological 
differentiation seems to justify the following 
conclusions: rst, the process in its essen- 
tial features is identical with the process of 
normal differentiation; second, the charac- 
ter of a tumor depends primarily upon the 
layership of the cells producing it; third, 
normal differentiation impedes and limits 
the formation of tumors, precisely as it does 
of further normal structures; so that tumors 
arise most readily from undifferentiated 
tissues and may then be heteroplastic; and 
arise readily from differentiated tissues and 
are then always homoplastic; and arise un- 
readily or not at all from the most highly 
specialized tissues. 
Each of these three conclusions might be 
advanced as a law of normal development, 
if we substitute the term ‘ differentiated 
tissue’ for ‘ tumor.’ ; 
We now pass on to the final stage of 
eytomorphosis, necrobiosis and hypertrophic 
degeneration in their pathological manifes- 
tations. The consideration of the direct or 
simple death of cells need not detain us, 
nor need we pause long over the indirect 
forms of cell death. In fact, the analysis 
made earlier this evening of normal necro- 
biosis and degeneration forced us to recog- 
nize that all, or nearly all, the modes of 
indirect cell death which the pathologist 
encounters in morbid tissue recur under 
SCIENCE. 
497 
healthy normal conditions. To put the 
conclusion in its correct form, we need only 
to reverse it, saying: 
Most, and probably all, pathological ne- 
crobiosis and degenerations of cells are es- 
sentially identical with normal processes, 
and are pathological, owing to the abnor- 
mality of their occurrence in time and 
site. 
Death of a cell may, of course, occur at 
any moment as a consequence of conditions 
external to itself. To a given cell, as such, 
it is of no moment whether the term ‘ physi- 
ological’ or ‘ pathological’ be applied by us 
to the conditions which cause its death. 
The cell has its own inherent qualities, and 
its own cytomorphic possibilities. All that 
the environment of the cell can do, so far 
as we can at present understand, is to 
evoke, and perhaps to a minor degree 
modify, one of the possible structural 
changes of the cell. Hence we find actu- 
ally that the processes of cellular necro- 
biosis appear to us identical in normal and 
pathological cases. This affirmation does 
not imply that a given cell has only one 
kind of possible necrobiosis before it. Quite 
otherwise, it being reasonable to believe 
that any one of several forms of necrobi- 
osis, according to the circumstances, may 
ensue. 
All that has just been said might be re- 
peated in reference to hypertrophic degen- 
eration. One of the investigations which 
is most needed at the present time, and 
which promises results of extreme interest 
and importance, is the investigation of ne- 
crobiotic and degenerative cytomorphosis, 
carried out as a research upon cell struc- 
ture. At present we cannot discuss the 
subject except in terms the very vague- 
ness of which is a mortifying confession of 
ignorance. 
Time forbids the prolongation of the dis- 
cussion. But, although a more detailed 
study is thus for the present excluded, we 
