TARSAL CHEMORECEPTORS OP BUTTERFLIES 195 



was placed in a holder and held on clean filter-paper for thirty 

 seconds to overcome any death feigning and to make certain that 

 the proboscis exhibited no sign of movement. The animal was 

 then transferred to the center of the watch-glass, where it was 

 held with all four walking legs in contact with the cotton pad. 

 There was enough solution present to immerse the distal ends 

 of the tarsi, if the legs were pressed firmly against the cotton. 

 In struggling, the butterfly frequently did this, and if it failed 

 to do so, a little pressure on the holder easily brought about 

 the same result. This manipulation, however, was seldom neces- 

 sary, as mere contact with the cotton was generally sufficient 

 to produce a response if any was given at all. As in previous 

 experiments, a failure to observe any visible extension of the 

 proboscis within a period of one minute was considered a 'no 

 response.' Immediately after each trial the feet of the butterfly 

 were immersed in distilled water and thoroughly washed, in 

 order to remove any adhering material which might contaminate 

 succeeding trials. A minimum interval of fifteen minutes was 

 allowed between the trials of a single individual. 



Four solutions and distilled water were tested in the above 

 manner. The solutions were 1 M saccharose, 1 M hydrochloric 

 acid, 1 M sodium chloride, and M/600 quinine sulphate. In 

 the introductory paragraph of this paper, especial stress was 

 laid upon the importance of using stimuli which the animal 

 encounters in its normal environment, and for this reason apple- 

 juice was employed in a number of experiments. Some of the 

 substances mentioned above are rarely or never met with by 

 butterflies in their natural state. It was necessary to select 

 these substances, however, because their solutions best met the 

 requirements of the experiment, viz., to afford widely different 

 contact stimuli along with uniform distance stimuli. 



It seemed possible by repeated trials on each of a large number 

 of butterflies, that the number of responses to each of the various 

 substances employed might be taken as a measure of its effective- 

 ness; and by comparison of these data any discrimination might 

 be demonstrated. Three experiments were performed, involving 

 a total of twenty-four butterflies, all of which had been reared 



THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 33, NO. 1 



