tate diagnosis by suggesting the plants that should be 
looked for when certain symptoms are observed and plant 
poisoning is suspected. 
-A word may not be inopportune on the need of re- 
search on poisonous plants. The indefiniteness with 
which many statements have had to be made and the 
dearth of positive experimental work upon which to base 
conclusions have been keenly felt in the preparation of 
these pages. Two lines of attack are evident, one based 
on the detailed chemical analysis of the poisonous plant, 
the other on feeding experiments with the plants them- 
selves. These two lines are complementary and both very 
necessary. The toxic substance having been determined, 
it is often at once evident what the chemical antidote 
should be. Moreover, much is known or could be easily 
learned of the conditions of its action. For example, how’ 
much is immediately apparent when it is known that 
prussic acid is the chemical responsible for Cherry or 
Sorghum poisoning. Knowing the volatile character of 
this toxic substance we can at once see why the dried 
fodder is harmless. Again, when the chemical responsi- 
ble for Hemlock poisoning was found to be polymerized 
by heat much of the erratic character of the poisoning 
in this case was understood. It was quite evident why 
this plant was less dangerous in the-heat of summer. 
Still all the emphasis must not be placed on this line of 
research, essential though chemical analysis assuredly is. 
The constitution of a plant is exceedingly complex and 
several factors may co-operate in producing the poison- 
ous effects. As mechanical injury in the case of Skunk- 
tail Grass opens the way for bacterial infection, so in the 
case of Horsetail poisoning, it is quite possible that irri- 
tation of the digestive tract by the indigestible silica may 
pave the way for the attack of a chemically poisonous 
substance. It is thus only by feeding experiments, ap- 
