32 TECHNICAL BULLETIN 134, U. S. DEPT. OF AGRICULTURE 
EXTRACTION OF TOXIC PRINCIPLES BY VARIOUS SOLVENTS 
To obtain information regarding the chemical nature of the toxic 
principle of red squill, a mLxture of two powders (P. C. 724 A and 
P. C. 725A) was taken for intensive study. Seventy-five-gram por- 
tions in Soxhlet thimbles were extracted with water, with 95 per 
cent ethyl alcohol, with acetone, and with chloroform, until there was 
no further appearance of color. The solutions were concentrated in 
a current of warm air, and the resulting solid extracts were dried in 
an electric oven at 80° C. and finally in a vacuum desiccator over 
calcium chloride. The residues were removed from the Soxhlet thim- 
bles and dried in the electric oven at 80°. Feeding tests were made 
with, the original mixed powder and with some of the extracts and 
residues. 
The water extraction, started on January 20, 1926, was run from 
8.30 a. m. until 5 p. m. daily for 19 days. The dark-red mucilaginous 
extract, possessing an odor of molasses, was concentrated on a steam 
bath, then dried in an oven, and finally allowed to stand over calcium 
chloride in a desiccator imtil constant in weight. It was ground to 
pass through a 40-mesh sieve. Both the extract (P. C. 751) and the 
exhausted residue (P. C. 750) were very low in toxicity, indicating 
that the active principle is decomposed by this treatment. 
The alcohoHc extraction for 19 days removed a much smaller 
quantity of mucilage, but apparently as much color as did the water 
extraction. The residue caked and became very hard upon exposure 
to the air. The soHd extract (P. C. 749) had a rather oil}^ feel, 
and was the most toxic extract in this series. The lethal dose of 
the residue (P. C. 748) exceeded 5,000 milligrams per kilo of body 
weight. The residue therefore is inert (nontoxic). 
Acetone extraction gave a small yield of lemon-yellow crj^stals 
(P. C. 755) and no mucilage or color. Feeding tests showed that 
the very fluffy residue (P. C. 754) was two-thirds as toxic as the 
original powder, indicating that a portion of the toxic principle is 
destroyed during the treatment ^\dth acetone. 
Chloroform gave the same result as acetone. The soHd extract 
(P. C. 757) was small and oily. The fluffy residue (P. C. 756) was 
somewhat less toxic than the original powder. The acetone and the 
chloroform extracts were too small to be used for feeding tests. 
Distilled water was repercolated through another 75-gram charge. 
Channels formed, producing incomplete extraction. So much muci- 
lage dissolved that after seven days further repercolation was found to 
be impossible. The tliick, molasseslike percolate had a sweetish odor 
and reduced Fehling's solution. Evaporated to a very dark-brown, 
tough, soUd extract (P. C. 759), it was readily eaten by rats. The 
lethal dose was 1,500 milhgrams per kilo of body weight. The incom- 
pletely extracted residue was not fed to rats. 
To study further the effect of water in extracting the toxic prin- 
ciple, 75 grams of squill powder were placed in a beaker wdth about 
800 cubic centimeters of water and warmed on a steam bath. Twice 
a day the solvent was removed in a Buchner funnel, and a fresh 
charge of water was added. After 22 days the extraction appeared 
complete. The aqueous solution was concentrated to a solid extract 
(P. C. 753). The residue (P. C. 752) was a dark-brown, tough sheet, 
heavily laden with ghttering calcium-oxalate crystals. Both the 
