594 
Journal of Agricultural Research 
Vol. XXI, No. 8 
little food and a meager water supply and hence be under the worst 
possible conditions for developing tumors and especially tumors con¬ 
taining shoots, and because owing to this procedure one would not then 
be able to distinguish between the specific crowngall stimulus and the 
general stimulus of separation, which in Bryophyllum sets all the leaf 
buds growing. 
Mr. Levine's most serious criticism is the statement that the shoots in 
crowngall develop from tumor cells. According to my observations the 
crowngall stimulus does not create totipotent cells out of tumor cells, 
but only sets growing those totipotent cells which already exist in the 
invaded tissues. The tumor cell is a disoriented degenerate cell, given 
over to a hasty vegetative growth. It is not an embryo cell, and I know 
of no evidence going to show that it can develop subsequently into normal 
tissues, organs, or the whole plant; on the contrary, it tends steadily 
toward decay. Moreover, since the tissues are not killed, what becomes 
of the bud when inoculations are made in the leaf axil if the numerous 
shoots which appear in various parts of the subsequently developing 
tumor (as for example in Plates 102 and 103) are not growths from dis¬ 
lodged fragments of the bud ? Furthermore, when a deep crowngall de¬ 
velops under the normal cortex, the cortex is lifted up and grows with the 
growth of the tumor without being actually a part of the tumor tissue. 
Certainly its cells have normal orientation, function normally, are no 
part of the malignant tissue and have not originated from it, although 
they are borne on it, as in Plate 101. See also Bulletin of the Johns 
Hopkins Hospital, September, 1917, figures 70, 71, 72 (okra tumors), 
and many other figures which I have published. 
Concerning the tissue of vessels and mature cells contained in the 
substance of crowngalls, one may call it stroma if he likes. Every tumor 
has a stroma (supporting tissues), and it is not likely that the stroma in 
plant tumors would be exactly like that in animal tumors, although the 
amount of stroma is extremely variable in the latter. The crowngall 
stroma is organized along with the tumor and forms an intimate part of 
it, but I can not think that it is necessarily developed out of infected cells. 
We shall not know positively, perhaps, until we are able to stain the 
bacteria in situ. As in malignant animal tumors, the crowngall stroma 
appears to me to be a growth of normal tissues (vessels and connective 
cells) stimulated by the presence of the abnormal cells, just as the 
stimulated roots and shoots are outgrowths of normal cells; yet the lat¬ 
ter, being no part of the actual tumor tissue, have normally arranged 
tissues whereas the tracheids of a crowngall, being a more intimate part 
of its structure (stroma), are often contorted into the most bizarre forms 
by the growth of the tumor. A crowngall in its growth often surrounds 
normal cells just as a cancer may or bears them on its surface. The 
number of vessels in a crowngall depends not on the activity of the 
tumor so much as on the nature of the tissue invaded—that is, a tumor 
