-48- 
vere made by shaking the sample with acetone for 7 hours, using such 
■oronortions that each- cubic centimeter of the filtered extract repre- 
sented 0.2 gm. of the plant materiel. Onl*y one determination was 
made of each sample each day. The materials were applied as an 
acetone spray at a concentration of 8 mg. of plant material ner cable 
centimeter of solvent. Six tests were made, each on 38 samples. 
Each set of 38 consisted of composite samples from seversl -orogenies 
of each of the 10 parent plants from each of the 4 growth stages, 
except that in 2 cases materiel was not available. Clonal progenies 
of 10 parent nlants of Tephrosie virginiene "ere grown under culti- 
vation in northeastern Texas to study the changes in the amount of 
rotenone and chloroform extractive present in the roots of such pro- 
genies and their toxicity to houseflies at 4 seasonal stages of 
growth. The roots were dried and ground, the amount of chloroform 
extractive end rotenone determined-, and the toxicity of acetone 
extracts tested on houseflies. The results indicated that at the 
full-bloom stage -the roots were significantly more toxic to house- 
flies than at the dormant end emergencestages, hut their superiority 
over those at the mature-seed stage was less pronounced. The chloro- 
form extractive and. rotenone content was also highest at the full 
-bloom stage. The toxicity of the roots of the several clonal pro- 
genies of the same -oerent did not very significantly, but signifi- 
cant differences were found in this, respect between the progenies of 
d'ifferent parents. 
The United States Department of Agriculture, Bureau of Entomology (2 
and Plant Quarantine, Division of Control Investigations,/ in 1940 
reported that rotenone in s^-frol (20 m-r. per cubic centimeter, dis- 
persed by dropping on a hot plate, caused. 15-percent knock-do^n of 
houseflies in 10 minutes, and killed 65 oercent in 2 days and 83 
nercent in 3 days. 
■ The Whitmire Reseerch Corporation ( 319 ) in 1940 published an 
account of repellency tests against houseflies. A series of repel- 
lent re Ik tests develoned by Whitmire and Denny, at St* L^uis, '.'o. , 
has demonstrated the effectiveness of this method in checking the 
toxic end repellent values of contact insecticide sprays. The test 
wrs conducted by painting 125 mg. of spray on 8.75 sauare inches 
around the tot) of the arc of a circular screen roach cage. The base 
of the cage was a funnel rhich extends into the observation ca~ . 
Flies so affected with leg paralysis that they ceonot fly or walk 
fell through* the small opening at th. t tton of- the funnel into the 
observation cage, "here they had access to food and water. fun- 
nel opening could be kept closed by a cork stooper in order to 
record the percentages of knock-down at definite time intervals. 
The livestock spray used was a 10-percent solution Of a mixture of 
tephrosin, deguelin, sumatrol, toxicer >1, end other highly toxic 
unidentified substances extracted from derris resins. The eage was 
painted 1 hour before the flies -^ere introduced and the following 
average results were recorded: After 10 minuter 26 percent - re 
down; after 30 minutes 74 peremt -ere dorn, at LOUT 98 percent 
v "ere dorm, rnd at the end of 2t h mrs 99 percent • vre dead. 
