Jan. 29, 1917 
Mechanism of Tumor Growth in Crowngall 
177 
I inserted tubes containing stronger solutions of monobasic ammonium 
phosphate (20 per cent) and from its vapors obtained many positive 
results (PI. 32-36). All of these experiments yielded positive results— 
that is, fine proliferations from surface cells of the pith cavity, not only 
in the opened internodes , but also in many unopened ones above and below , 
and consequently where the action had to take place at a distance 
through thick cross walls. 
From these experiments there can be no reasonable doubt, I think, 
that any soluble substance whatsoever, except a killing, a plasmolyzing, 
or an oxygen-absorbing substance, if continually liberated in excess 
locally in tissues would be competent to induce tumor formation. 
One of the most striking results of these experiments has been the pro¬ 
duction of a stem within a stem. Recently in The Journal or Agricul¬ 
tural Research (Apr. 24, 1916) in detailing various experiments made 
with the crowngall organism I showed that under some circumstances 
the bacterial stimulus caused the production of a stem within a stem 
(21, PI. 21). The plant used was the common tobacco ( Nicotiana taba- 
cum), and the new stem, or more properly stele (which I called a “tumor 
strand ” because at frequent intervals it induced the formation of sec¬ 
ondary leafy tumors), was developed in the outer cortex of the normal 
tobacco stem between tumors where I believe no such structure was ever 
seen before. 
I have now been able to produce this strange phenomenon of a stele 
within a stele in the absence of the crowngall bacterium by simply 
injecting one of its products—to wit, an ammonium compound—into 
growing tissues. Internodes of the castor-oil plant ( Ricinus communis) 
were used for all of these experiments. I selected this plant rather than 
tobacco because it is easier to work with than a plant having a solid stem, 
and also because the fluids could be injected in any desired small quantity 
without tearing the tissues, the only mechanical injury being the small 
hole made by the hypodermic needle. 
The results obtained have been very surprising and are, I think, quite 
suggestive not only of the growth previously obtained between tumors 
in a tobacco stem by means of the bacteria (the cortical stele formation), 
but also of the tumor formation itself. The photomicrographs show my 
results very clearly and only a few explanatory words are necessary. 
These results were obtained with the monobasic ammonium phosphate. 
Young, half-grown, and old intemodes were injected but striking results 
were obtained only once, when a very young internode was injected. 
The manner of experimentation was very simple. The needle of a 
glass-walled hypodermic syringe was thrust through the upper part of the 
wall of the intemode and the cavity (very small in the young intemode) 
was filled with distilled water in which 5 per cent of the ammonium salt 
had been dissolved. Sometimes the needie wound was closed with collo- 
