CHEMICAL SENSITIVITY OF TARSI OF PYRAMEIS 71 



In continuing periods of starvation as long as possible, the 

 risk of death became very great, and of eight specimens employed, 

 six died in the course of the experiment. The longevity of these 

 individuals was 8, 8, 9, 10, 19, and27 days, respectively. Inall these 

 eases death was most probably due either directly or indirectly 

 to the effects of starvation. In several of the shorter-lived speci- 

 mens, I might have saved them had I appreciated the gravity of 

 their condition soon enough, for anunals in a state of complete 

 collapse may sometimes be resuscitated in a surprisingly short 

 time by the administration of liVl saccharose. Two specimens, 

 however, survived the entire experiment, and several days after 

 their last trials appeared vigorous in every way. They were 

 killed for subsequent morphological study, having been under 

 observation for twenty-nine and thirty days, respectively. Of 

 the eight animals, I shall present the data from the four longest- 

 lived only. The data from the four shorter-lived individuals are 

 less complete and show nothing not shown by the others. 



The observations on the four butterflies which survived longest 

 are presented graphically in figure 3. The data for each animal 

 numbered, respectively, 11, 12, 13, and 22, are presented in the 

 form of four curves, each of which represents the responsiveness 

 to a single substance. In these curves, the total weight of daily 

 response, as defined on page 59, is plotted against age in days, 

 the nutritional state for each day being indicated. An examina- 

 tion of the four curves for any one animal shows, with one excep- 

 tion, viz., no. 22, that no two coincide. In other words, there 

 were differences of response to the different stimuU. It should 

 be borne in mind at the outset of this discussion that the fact 

 that a butterfly responds identically to two substances does not 

 necessarily mean that it fails to distinguish them. This may 

 merely indicate a positive response to both substances. On the 

 other hand, differences of response do not necessarily show^ dis- 

 crimination, unless they are differences of pronounced nature and 

 regular occurrence. Are the differences noted above such as to 

 indicate discrimination of the various stimuli or not? In order 

 to answer this question, let us examine and compare the responses 

 with respect to: first, relation to nutritional condition; second, 



THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 35, NO. 1 



