362 ROBERT S. MCEWEN. 



the other hand, there is no significant difference between the 

 records for the lined and unlined tubes in either the horizontal 

 or vertical tests. In this connection it might further be stated 

 that not only does the data based upon timing indicate that 

 slipping is not a significant factor in retarding the vestigials, 

 but that observation of their actual behavior leads to the same 

 conclusion. Thus in the vast majority of cases the writer was 

 unable to see that the vestigial insects had the slightest difficulty 

 in crawling upon the unlined glass. Regarding the matter from 

 yet another aspect it is instructive to note the records for corre- 

 sponding groups where the only difference involved is the position 

 of the tube. Here it appears that in the case of the wild flies 

 in the unlined tubes the vertical position is slightly superior 

 both as regards the arithmetical average and the mode. In the 

 lined tubes on the other hand, only the modal value for the 

 vertical position is superior. As regards the vestigials, the lined 

 tubes again show only a slight superiority for the vertical records 

 but in the case of the unlined tubes the vertical position is 

 markedly superior, both as to average and mode, even without the 

 correction for slipping. This of course is exactly opposite to what 

 one would expect according to Cole's results and interpretation. 

 Besides this analysis of the data as a whole, it is also pertinent 

 to call attention to certain facts brought out by observation of 

 individual insects. Thus in the c'ase of the horizontal tests of 

 vestigial flies in the unlined tubes, where the behavior of the 

 animals could be perfectly observed, the following feature was 

 noteworthy. There were 14 insects which were noted on the 

 records, not only as having no tendency to slip, but as being 

 unusually active; i.e., they crawled rapidly and constantly about. 

 Yet the average time for these 14 cases was 23.5 seconds, virtually 

 the same as that of the general average for the test. There 

 were likewise 16 vestigials in this test which were recorded as 

 "quite" active, with an average of 24 seconds, and 12 recorded 

 as "fairly" active with an average of 27.4 seconds. Hence the 

 mere activity of these animals does not seem to have materially 

 affected their tendency to move toward the light. The only 

 feasible explanation of such cases would seem to be that such 

 flies are only weakly phototropic. This interpretation is given 



