Effect of Oil of Citronella on two species of Dacus. 413 



a distance of about five yards and vigorously shaken and 

 waved in the air to dislodge and disturb the flies ; it was 

 then replaced, and the flies which had returned to it are 

 shown in Plate XL, which is a photograph taken exactly 

 three minutes after its first removal, or perhaps two 

 minutes after it was replaced. 



In both the photographs it will be noticed that the flies 

 are congregated not actually on the moistened patch but 

 round its margin. This is their usual custom, and was 

 taken advantage of in catching the flies with fly-papers. 

 If citronella is put in the middle of the fly-paper (on the 

 gummy substance) many flies escape capture by sitting on 

 the edge of the paper which is free of adhesive ; if, instead 

 of this, the citronella is put on the edge of the paper, they 

 will not sit on it, but settle on the sticky surface : a very 

 sensible difference is thus made in the number of flies 

 caught. Fly-papers treated with citronella were exposed 

 in the orchard during the months of March, April, May 

 and June. A careful estimate of the number caught 

 during part of this period gave approximately eighteen 

 thousand, and among these not more than fifty females 

 were seen, or 03 per cent. Since the reaction was con- 

 fined to the male sex and did not appear to be in any 

 way connected with feeding habits, it seemed most reason- 

 able to suppose that the smell might resemble some 

 sexual odour of the female which in natural conditions 

 served to guide the male to her. 



Six or seven freshly killed females were therefore 

 placed in a clean glass tube which was closed for about an 

 hour with clean cotton-wool. On smelling the tube a 

 faint odour resembling that of citronella was just per- 

 ceptible, but although the presence of the smell was 

 confirmed by my assistant, it was so faint that I feared 

 the influence of unconscious " auto-suggestion " on our 

 judgment, and repeated the experiment with about twenty 

 living females which had emerged from the pupa from 6 

 to 24 hours previously. In this case the smell was 

 distinctly perceptible and closely resembled the citronella 

 smell ; its presence and nature were confirmed by an 

 independent observer who did not know what smell was 

 being looked for or expected. When a similar number of 

 males were tried in the same way, no smell of citronella 

 was detected. 



It seems probable, thereforej that this smell is the 



