182 REPORT 4, UNITED STATES EKTOMOLOGTCAL COMMlSSIOii. 



guide ns. . Plants that are poisonous to man and the higher animals are 

 usually harmless to insects j others may possess insecticide qualities 

 only at a certain season or only when they are dried or otherwise pre- 

 pared 5 in still other plants only one part, e, g., the flower or the root, 

 may prove useful, perhaps only a;t a certain stage of development, and 

 the rest worthless. Hence experiment in this direction must be more 

 or less tentative, and practical results will follow only, if at all, the 

 most extensive trials. Other difficulties also present themselves. Two 

 observers, while experimenting with the same material, with the same 

 instruments, and under precisely similar conditions, often report vary- 

 ing results ; or the same observer after a favorable result in the first trial 

 obtains a quite difterent result in subsequent trials. In the case of 

 tlie Cotton Worm, with its rapid development, there is, also, great risk 

 of drawing false conclusions, the most frequent cause being the appar- 

 ent sudden disappearance of the worms, or at least of a large portion 

 of them, after the application of the "remedy." The worms may 

 have simply changed to chrysalides or may have been carried away by 

 enemies. Again, the presence of a number of dead worms upon and 

 under the plants is often misleading, as such worms may have been 

 killed by their numerous enemies. Finally, very young worms, or such 

 as are just in the act of molting, are often killed by the mere force of 

 a spray thrown upon them, while others that have fallen to the ground, 

 and are more or less covered with, soil, exhaust themselves in their 

 frantic efforts to get rid of the incumbrance. 



In this connection, also, it must be remembered that experiments 

 made indoors are without practical value unless confirmed by repeated 

 experiments in the field. 



The plants with which we have experimented are partly such as, were 

 recommended by correspondents as being efficient for the Cotton Worm 

 or other insects;* partly such, as from their immunity from insect at- 

 tacks might be presumed to possess insecticide properties. Prevalence 

 and cheapness being desiderata in any plant that might prove to have 

 destructive qualities, many of the weeds that by their abundance, per- 

 sistency, and power of adaptation are so obnoxious to agriculture in the 

 South were tested. In order to test such plants thoroughly they should 

 be used both fresh and dry, and in flower, fruit, leaf, and root. This is 

 a work of considerable magnitude which cannot possibly be accomjjlished 

 in one or two seasons, and for the purpose of this investigation we 

 deemed it advisable, with few exceptions, to first try alcoholic extracts 

 and decoctions, mostly made from the fresh plants, leaving for the future 

 more complete experiments with such as give promise of success. While 

 the results thus far obtained seem discouraging, considering the outlay 

 in money and time, and while in no case was the effect upon the worms 

 such as would warrant the recommendation of any of these plants for 



* A few of these plants could not be recognized from the popular name given by the correspondent, 

 And could, therefore, not be tested. 



