24 BULLETIN 800, IT. S. DEPARTMENT OF AGRICULTURE. 
gave no reaction with ferric chloric!. Two half-gram portions of it 
were suspended in water and drenched into a guinea pig without 
effect. 
Alcohol extract. — Four thousand three hundred grams of the dried 
whole plant were extracted with alcohol, the solvent removed, and the 
residue treated with hot water, which dissolved a large quantity of it 
and left a resinous mass undissolved. Both the water solution and the 
residue were toxic ; the solution produced a peculiar narcosis and the 
resin caused the characteristic spasmodic symptoms observed in range 
poisoning. 
The resin. — This was divided into 4: fractions by treatment with 
petroleum ether, benzol, and alcohol in succession, when a small resi- 
due remained which was insoluble in the ordinary organic solvents, 
but dissolved in dilute sodium-carbonate solution and was nontoxic. 
The petroleum-ether fraction was nontoxic ; the benzol fraction pro- 
duced the spasmodic type of intoxication; the alcohol fraction was 
narcotic. 
The aqueous solution. — A portion of this solution was investigated 
for the presence of toxic saponins, by treatment with barium hy- 
droxid, and testing the various fractions obtained on guinea pigs. 
Nothing in the nature of a saponin was detected. 
The main portion was made alkaline and extracted with chloro- 
form. This treatment yielded a small quantity of a nontoxic alkaloid 
aifd a glucosidal substance which has strong narcotic properties. The 
aqueous liquid which remained after the chloroform extraction was 
still toxic, producing narcosis in experimental animals, and from it a 
second toxic glucosidal substance has been isolated. 
The marc. — The marc remaining after the alcohol percolation was 
then extracted with boiling alcohol and this extract was kept separate 
from the first. The marc remaining after this treatment was thor- 
oughly dried to free it from alcohol and tested for toxicity. On 
August 2, 1918, 43.5 grams of marc (60 grams of dried plant) were 
forced fed to Sheep 479, weight 93 pounds. This feeding produced 
no apparent effect, and on August 4 the same sheep received 87 grams 
of the dried marc, also without effect. On August 6 a further quan- 
tity of 174 grams of marc were forced fed to the same sheep, again 
without effect. On August 7, 269 grams of the marc were forced fed 
to Sheep 497, weight 81.5 pounds, and produced no effect. The marc 
evidently did not contain toxic matter responsible for the cases ob- 
served on the range. 
SUMMARY OF CHEMICAL EXAMINATION. 
The plant material used was identical with that employed in the 
feeding experiments. The fresh green plant and the dried plant were 
