i68 
Journal of Agricultural Research 
Vol. VIII, No. 5 
lithium (io, n), 1 and the excretions of the larvse of gall flies, of certain 
nematodes, and of various fungi. Other substances are known to cause 
cell proliferations in animals—that is, Sudan III and scarlet-red (12), 
repeated applications of tar (13), excretions of Bilharzia, etc. As bearing 
on the origin of neoplasms, these facts are extremely suggestive; but for 
obvious reasons chemicals which can not be supposed to occur in genuine 
neoplasms as the result of their own metabolism and can not be assumed 
to be excreted into them by parasites either actually or presumptively 
present, need not be considered here, however effective they may be as a 
cause of cell proliferation. In our search for the direct exciting cause of 
tumors we need consider only the excretions of known tumor-producing 
organisms; but since little or nothing is known concerning the growth¬ 
exciting substance, or substances, produced in animal neoplasms, or liber¬ 
ated in plants by the various gall-forming fungi, or by gall flies and gall 
nematodes, we may for the present confine our attention exclusively to 
the bacterial tumors of plants. This narrows down the probleiti to a few 
species of bacteria and even some of these—that is, Bacterium savastanoi 
(the cause of olive tubercle) and Bad. beticola (the cause of beet 
tubercle)—are negligible so far as the purposes of this paper are con¬ 
cerned, because I regard the tumors they produce as granulomas, not 
neoplasms. In fact, thus far in seeking for the physical-chemical origin 
of neoplasms I have sought only for answers to the following questions: 
(1) What are the products of the bacterial metabolism of Bad. tume- 
faciensy the crowngall organism ? 
(2) Are any of these products capable of inducing cell proliferation 
when injected into the growing plant ? 
(3) If so, is this substance a sufficiently common product of other 
bacterial and nonbacterial tumor-producing parasites so that it may be 
regarded hypothetically as produced by all of them and consequently as 
the exciting cause of all bacterial, fungus, nematode, cynipid, and other 
vegetable galls, or must we suppose that various chemical growth 
excitants exist ? 
(4) And finally to what extent can these facts be supposed to apply 
to human and animal neoplasms ? 
In case of the crowngall organism I had to begin with the simpler sub¬ 
stances, because I have as yet no knowledge of the more complex products 
of its metabolism, if any such exist, and certainly no very poisonous 
substances are produced in the crowngall, any more than in animal 
1 In this connection should be mentioned especially Caroline Rumbold’s recently published results of 
chestnut-bark injections (xi). With substances injected in the hope of controlling the Endothia chestnut- 
bark disease, and especially with lithium carbonate, she succeeded in causing numerous conspicuous islands 
of xylem to develop in the middle of the phloem. These caused bulges in the bark, visible externally. I 
have had the pleasure of examining her sections, and verifying her statements. The islands are developed 
from a cambium, the growth being always from the outer face of a bundle of bast fibers toward the surface 
of the stem. I could not satisfy myself as to the origin of this cambium from the normal cambium of the 
stem (Solereder), but it is plain that the occurrence of these islands of xylem in the phloem (they appear to 
be that rather than islands of phloem in xylem) is correlated with a scanty production of the normal wood 
of the season. 
