A. D. Imms and M. a. Husain 279 



which the odour of a particular substance may exercise chemotropic 

 properties. According to Howlett (1915) the radius of attraction of 

 citronella oil for certain fruit flies may be taken as about half a mile: 

 with many Lepidoptera the female probably attracts the male from a 

 much greater distance. It is evident, therefore, we should guard against 

 placing substances to be tested too near to each other. For every agent 

 to be tested suitable controls have to be decided upon, and spaced as 

 far apart as may be possible. 



In our field experiments we have to obtain the following data: 



(1) Power of attraction of a substance in various dilutions and 

 under varied atmospheric conditions. 



(2) Optimum attractive strength. 



(3) Relative powers of attraction as compared with other sub- 

 stances. 



(4) The species attracted, their numbers and the proportion of 

 the sexes. 



(5) Meteorological data. 



There are two methods of discovering substances likely to exhibit 

 chemotropic attraction. The one is to undertake experiments with any 

 substance that normally occurs in the environment of particular species, 

 e.g. a constituent of the food of the larvae or adult, or likely substances 

 chemically allied to the same. The second method is to test a large 

 number of substances, necessarily somewhat indiscriminatingly at first, 

 in the hope of discovering attractive agents. VerschafEelt (1910) observed 

 that the alkaloid that gives mustard oil is the attractive agent in the 

 larval food of Pieris brassicae ; he started with the food plants and worked 

 with the ingredients thereof. Richardson, who has studied the reactions 

 of the house-fly to various agents, also followed this same method. He 

 first observed that the house-fly is attracted to the manure heap by the 

 odour of certain substances, which were being liberated during the early 

 stages of fermentation. On this basis, he experimented with various 

 inorganic and organic compounds that occur in the manure heap. This 

 method is full of possibilities, in so far as individual species are concerned, 

 but the second of the above methods has its advantages for preliminary 

 quantitative work. It has been observed that a large number of insects 

 are attracted towards chemicals that are neither associated with the 

 smell of their food substances, nor occur in their normal environment. 

 It is, moreover, in the discovery of such substances that some of the 

 most striking results have so far been achieved. We have therefore 

 followed this method of exposing a large number of chemicals and 



