- 35 - 
"In the inalysis of finely powdored samples of c'erris and cu.be 
roots a method involving treatment with chloroform at roon temperature 
follower 1 by removal of an aliquot of the filtered extract gives satis- 
factorily coriplete extraction of the rotenonc. 
"Fineness of the sanple is an exceedingly important factor in 
obtaining complete extraction by any method. If coarse samples are 
ground so that at least 95 per cent passes a 60-r.csh sieve, they will 
usually give ratisfo.ctory extraction by the aliquoting procedure, 
"Samples containing a high ratio of ro tc.ocnc to total extrac- 
tivos uerc found to be core difficult to extract than thoec with lower 
percentages of x-otenono. When the ratio of rotor one tc total extract 
'./as about 40 per cent or over, particularly in the case of dcrris 
roots, it was necessary to employ extractiOxi at roon. tenperature v/ith 
successive lots of chloroform in order to obtain satisfactory extrac- 
tion of the rotcnonc. This nether 1 should also be employed as a check 
v/henevor there is doubt as to the eor.rolctenese of extraction by the 
o.l i quo o i ng p ro c o dur o . 
"Cube roots in general arc more readily extracted of their ro- 
tcnonc content than are dorr is roots, 
"The moisture content of dcrris on'", cube roots as received in 
this country has not been found to be sufficiently great to interfere 
v/ith tnoir analysis, and hence preliminary drying of samples seems 
unnecessary." 
Jones and Graham (229) in February 1936 published the procedure recom- 
mended by them for the determination of rotenone in dcrris and cube. This 
is as follov/s: 
"Weigh 30 grans of the finely powdcrec root into a 500 cc. glass- 
stopocred Erlenrieyer flask. Ac"d 300 cc. of CHCl 5 measure 1 at a defin- 
ite ran temperature; ulacc the flask on a shaking machine and fasten 
the stoppers securely. .Agitate vigorously for not less than 4 hours, 
preferably interrupting the shaking v/ith overnight rest. (As an olter- 
native procedure the flask nay be shaken continuously overnight.) Re- 
move the flask from the n:,chino and alio - ;/ to c :• 1 in a refrigerator 
for o.t least an hour. Filter the nixtv.rc rapidly into a suitable flask, 
using o. fluted paper without suction and keeping the funnel covered 
with a v/atch-glass to avoid loss from evaporation. Stopper the flask 
and adjust the temperature of the filtrate to that of the original 
CHC1 3 . 
"Transfer exactly 200 cc. of this solution to a 500 cc, Pyrex 
Erlcnmcyer flask and distil until only about 25 cc. remains in the 
flask. Transfer the extract to a 125 cc, Erlcnneyer flask, -using CC1 4 
to rinse out the 500 cc. flask. Evaporate almost to dryness on the 
steam oath in a current of air. Then remove the remainder of the 
solvent under reduced pressure, heating crutiously on the stean bath 
when necessary to hasten the evaporation. (The suction nay be .applied 
chirectly to the flask.) Dissolve the extract in 15 cc. of hot 
CCI4. and again, in a similar manner, remove all the solvent. Repeat 
