1 82 Journal of Agricultural Research voi. xvir, No. s 



feeder, which was so covered with wire that the bees could not waste 

 any of the food. After the lo feeders, containing supposedly poisoned 

 food, had been placed in as many wire-screen cases, 50 normal bees 

 were introduced into each case; the bees were thereafter observed care- 

 fully and the dead ones were counted at regular periods. As a control, 

 honey containing the same amount of alcohol as mixed with the other 

 food was used; and whenever the bees required more food, pure honey 

 was given to them. These experiments were repeated and were so 

 arranged that the probable errors were minimized. Reference to Table 

 II (extracts No. 246-254) shows that all of these extracts, except the 

 water extract, are almost equally toxic to the honeybee within 48 

 hours and that there is practically no difference in toxicity between the 

 extracts obtained with the use of heat and without it. The water 

 extract apparently had no effect on the bees tested. Similar results 

 were obtained by using the same extracts against aphids, fall webworms 

 (Hyphantria cunea Dru.), and tussock-moth caterpillars (Hemerocampa 

 leucostigma S. and A.) (see No. 246-249, 252-253, Table IV, and No. 

 253, Table V). The water extract from the powder of Derris sp. (filtered 

 mixtures) killed only a small percentage of the aphids sprayed (see 

 lower half of Table IV), while the nonfiltered spray mixtures, consisting 

 of powder and soap solution, were efhcient against aphids. 



To determine whether the solvents had removed all of the toxic 

 principle from the powders extracted, these five powders (No. 240-244 

 in Table II) after having been thoroughly dried were fed to other honey- 

 bees in the same manner as already described. In these tests ys gm. 

 of powder was thoroughly mixed with 5 cc. of honey. Reference to 

 Table II shows that the powders exhausted with ether, chloroform, 

 and alcohol had very little effect on the bees tested, while the powder 

 exhausted with water killed 94 per cent of the bees within 48 hours. 

 The results pertaining to the powder exhausted with petroleum ether are 

 not reliable (see note at bottom of Table II). 



To ascertain the effect of powder exhausted successively with i to 4 

 of the solvents and also the effects of the resulting extracts, other experi- 

 ments were performed. Reference to Table II (No. 260, 261, 264, and 

 266) shows that powder successively extracted is only slightly less 

 effective than powder extracted once, and that the third and fourth 

 successive extracts (No. 263 and 265) have no effect at all. These 

 results agree in only certain respects with the successive quantitative 

 extractions, expressed in Table I. 



To determine whether any poisonous volatile substance can be 

 removed from Derris by steam distillation, 50 gm. of the powder were 

 so treated and the distillate was collected. Later some of this distillate 

 and a portion of the distilled powder, after it had been dried, were tested 

 on silkworms. The distillate had no effect whatever, but the powder 

 was as poisonous as ever. 



