114 IVAR TRAGARDH— ON THE CFIEMOTROPISM OF INSECTS 



collectors who employ traps with strongly smelling baits. Numerous observa- 

 tions have shown how coprophagous Lamellicornia and carrion-bettles are 

 capable of detecting a long way off the smell of excrement and carrion ; 

 similarly in the case of insects, the females of which oviposit upon substances 

 or plants that serve both themselves and their larvae as food, it has been 

 recognised that it is the odour that guides them. 



But in the case of insects, the larvae of which have quite another diet 

 from that of their parents, e.g., parasitic Hymenoptera which oviposit upon a 

 certain insect, or butterflies that lay their eggs upon a single species of plant, or 

 upon the plants of a single genus or family, we have been satisfied with saying 

 that their actions were instinctive, without attempting to analyse which definite 

 elementary instinct or instincts, i.e. tropism, had guided their apparently 

 purposeful action. One of the reasons for not thinking of chemotropism as one 

 of the active factors in these cases, is probably that it was impossible for a 

 human being to detect any odour. And yet it is evident that it is chemotro- 

 pism which in the first instance guides the majority of these insects to the right 

 spot for depositing their eggs, even if, after they reach it, their touch or some 

 similar sense guides them, when it is a case of perhaps finding the very spot 

 upon the plant where the egg has to be deposited. 



If we consider that there are some Ichneumon idae that lay their eggs deep 

 down in the trunks of trees upon Sir ex larvae, it is of course quite evident that 

 neither sight, nor hearing, nor touch is able to assist them in ascertaining that a 

 larva is living in the tree ; but it must be a certain odour given off by the larva 

 which is sensed by the parasitic insect. In the same manner it is probable that 

 those parasites which attack leaf-mining larvae in the first instance are guided by 

 their sense of smell. And finally we have every reason for assuming that it is 

 the same sense which on the occasion of ovipositing guides all those insects whose 

 larvae are herbivorous to their respective food-plants. 



This assumption is confirmed by some experiments made during the last few 

 years. In the year 1910 Ed. Verschaffelt 2 published in Amsterdam his investi- 

 gations into the factors that determined the choice of food of certain butterfly 

 larvae. He experimented with the caterpillars of Pieris brassicae and P. rapae, 

 which prefer cruciferous leaves to all others. They eat, however, also leaves of 

 Tropaeolum and Reseda, and the composition of their bill-of-fare throws a good 

 light upon the determinative factors of their choice, for a group of glucosides, 

 the mustard-oils, is common to all these. 



In order to ascertain how strongly the mustard-oil attracted the larvae, 

 Verschaffelt smeared a dough of Bunias orientalis upon some leaves of other 

 plants which the larvae had previously refused to eat. Thus treated, the leaves 

 were attacked and devoured with evident appreciation. In some instances the 

 attempts failed, however, e.g., with leaves of Salvia officinalis and Menyanthes 

 trifoliata ; but this was certainly due to the occurrence of other substances" in 

 these leaves which exercised a more strongly negative reaction than the JBunias- 

 dough had been able to effect in a positive direction. In order to obtain con- 

 vincing evidence that it was really mustard-oil and not perchance some other 



* In the Salvia leaves there occurs a volatile terebene, called tujon. 



