352 ENTOMOLOGICAL NEWS. [Oct., 'l2 



tristis DeGeer; that the protective odor of Euschistus fissilis 

 sometimes killed the hardy weevils Anthonomus grandis Boh. 

 in the adult stage (at other times failing to do so), and that 

 the vapors emitted by Brochymena annulata killed adult mus- 

 cids and stupefied a centipede but were variable in their effects. 

 That variation of the efficiency of these secretions occurs ful- 

 fills the expectations, but has not of course been fully measured 

 by these experiments, if measured at all. The end result, 

 namely actual protection, after all, is the only measure of effi- 

 ciency in this respect, since the fact that the protective vapor 

 is seriously injurious to one species of insect and not to an- 

 other, under the unnatural conditions of these experiments is 

 no criterion upon which to state that it affords protection 

 against the one and not against the other. Because a toad is 

 killed once and not another time, under the same stated con- 

 ditions, is no reason whatever for stating that variation in the 

 protective efficiency of the vapor in question occurs. The 

 actual protection must consist in being distasteful or obnoxious, 

 not in causing serious injury, stupor or death to the attacking 

 animals, since the latter must involve, under natural condi- 

 tions, actual handling or even swallowing of the protected in- 

 sect in order that the protective vapor would have time to be 

 effective, even if then effective. In the case of these evil- 

 smelling Heteroptera, however, birds and amphibians mum 

 be the enemies to be feared and it is in giving the individuals 

 of attacking species of these animals the memory impression 

 of distastefulness that the protection consists. It works 

 through the memory of the young attacking animals, not stu- 

 pefying or killing them but giving them such an experience 

 of unpalatableness that they soon learn to avoid attacking. 

 Therefore, it is obvious that the experiments recorded by 

 Conradi and myself show no more than what is stated in the 

 first three sentences of this paragraph and have little or no 

 bearing on the relative efficiency of the protective odors in 

 question. 



The foregoing experiments were performed in a field la- 

 boratory at Centralia, Illinois, U. S. A. 



