242 
AMERICAN JOURNAL OF BOTANY 
[Vol. 8 
the combination of a liquid acid with albumin. He puts forth the following 
evidence in support of this conclusion: (i) the small amount of poison 
(0.000005 g.) necessary to produce itching and burning on the skin; (2) 
similarly to a toxalbumin, it loses its toxicity by heating to 50° C. on a 
water bath; likewise at 75° C. and 100° C. The toxicity was tested by 
rubbing with a probe on the skin of the upper arm. In no case was irritation 
evident. This evidence is inconclusive of the poison's being a toxalbumin, 
for: (i) other substances besides toxalbumins are poisonous when in such 
small amounts; (2) the poison remains toxic if heated on glass in a steam 
autoclave for one hour under twenty pounds' pressure per square inch 
(temperature 126.2° C); and (3) the poison contains no nitrogen. 
The work of Acree and Syme is probably erroneous for: (i) all three of 
the so-called constituents of the poison are found in two non-poisonous 
species of Rhus; (2) the natural glucoside yielding fisetin, rhamnose, and 
gallic acid is non-toxic; and (3) there is not sufficient evidence that the 
poisonous substance which Syme attempted to decompose was not a 
complex containing a poisonous body and one or more non-toxic glucosides 
in addition. McNair (30), working with R. diver siloba, concluded that the 
poison of this plant is not a glucoside of rhamnose, fisetin, and gallic acid. 
A di,fferent method was used for extracting the poison, and none of these 
substances could be obtained on hydrolysis. 
The specific cause of skin poisoning from R. Toxicodendron L. and its 
two sub-species, R. diversiloba T. & G. and R. radicans L., has thus far been 
ascribed successively to: an emanation of vapor; a hydrocarbon gas; a 
gum resin, mixed with a ''subtil" acid principle, qualified to combine with 
hydrocarbon gas which emanates from the plant after sunset; a volatile 
alkaloid; a non-volatile gum resin; a volatile organic acid (toxicodendric 
acid); an infection by bacteria {M. toxicatus, Burrill, 7); a non-volatile oil 
(toxicodendrol) ; a glucoside of fisetin, rhamnose, and gallic acid (toxicoden- 
drin); a toxalbumin; and finally to something other than a glucoside of 
fisetin, rhamnose, and gallic acid. 
The Transmission of Rhus diversiloba Poison 
My investigation of the transmission of the poison has been carried on 
from three standpoints; botanical, chemical, and pathological. The 
following chemical experiments were carried out : 
1. One half pound of fresh, finely chopped poison oak leaves were dis- 
tilled normally at different temperatures up to the point of decomposition 
of the leaves. As a result, both the distillate and the residue were non- 
toxic. 
2. Another lot of leaves similarly prepared was subjected to steam 
distillation. The distillate was non-toxic, but the residue in the retort 
remained toxic. 
