176 
Journal of Agricultural Research 
Vol. VIII, No. s 
grow, and therefore the products of their metabolism do not come into 
play; (2) most of the parasitic forms cause a loss of turgor, requisite 
for cell-division, either (a) by obstructing water-conducting channels, 
causing the plant to wilt, or (6) by excretion of plasma-killing toxins 
which pave the way for the bacterial advance into the tissues, or (c) 
by excretion of enzyms or other solvents of the middle lamellae of the 
cell-wall, in which case we have a softrot—that is, a complete or par¬ 
tial disintegration of the tissues; (3) moreover, the case is not so unique 
as some persons suppose, because various other bacteria than the crown- 
gall organism do actually produce overgrowths, large or small—that is, 
granulomas (often of large size), tyloses in vessels or small surface intu¬ 
mescences (I know of half a dozen or more such bacteria and new ones 
are coming to light every year); (4) a very considerable number of fungi 
also cause overgrowths; (5) the number of insects that cause overgrowths 
is legion. 
I will only say here, further, that there is really no reason to be dis¬ 
turbed by the variety of substances shown to be capable of causing cell 
proliferation, since probably only such substances can cause neoplasms 
as are thrown off locally and continuously into the tissues by cancer 
organisms, and that too intracellularly , for otherwise only granulomas 
should result in plants, while in animals the blood stream would quickly 
remove and effectually dispose of any irritating substances not contin¬ 
ually replenished. Moreover, in animals the tissue naturally most 
exposed to foreign substances—viz, the epithelium—is by nature the 
most resistant—that is, the last to yield to cancerous proliferation. 
Having come to the conclusion that I was dealing with a general 
physical problem rather than with a specific chemical problem, I next 
tried the effect of injecting various foods and poisons as follows: 
Ten per cent tannic acid (PI. 11), 10 per cent ethyl alcohol, 2 per 
cent sodium chlorid, 2 per cent sodium carbonate, 5 per cent sodium 
bicarbonate (PI. 12), N/20 sodium hydrofdd, 5 per cent ammonium 
bicarbonate (PI. 13), clear lime water (PI. 14), milk of lime (caustic), 
1 to 10,000 mercuric chlorid, 0.5 per cent carbolic acid, chloroform water, 
1 to 1,000,000 copper sulphate water, 5 per cent grape sugar (PI. 14), 5 
per cent cane sugar (PI. 15). 
These experiments also yielded positive results and very striking ones 
in some cases, even the grape-sugar and cane-sugar ones. 
Next, I undertook to determine the effect of very minute doses of 
ammonia—doses so weak that surface, woundings were not to be ex¬ 
pected. For this purpose I used the feeble alkaline vapors arising from 
dilute solutions of urea, ammonium Carbonate, and the two ammonium 
phosphates (PI. 16 and 17), these substances being inclosed in open 
small test tubes sealed into the pith cavity of Ricinus stems by means 
of collodion or surgeons' adhesive strap. I also injected 1 to 100 and 
1 to 1,200 ammonia water with striking results (PI. 18-31). Subsequently 
