1894-] ENTOMOLOGICAL NEWS. 35 



legs work in unison, not alternately, i. e. the middle legs keep 

 time with each other and with the hinder pair. The tips of the 

 legs rest on the water, and are not immersed in it so that the little 

 hollows near each, caused by the weight of the insect, can plainly 

 be seen. With captive specimens sunlight acted as a stimulant, 

 and evoked activity, which was lessened by shade. 



One Halobates was then placed in a tightly corked bottle, tilled, 

 when immersed, so as to make certain that no air was enclosed, 

 except the thin film which invested the insect. This was done at 

 1.22 p. m., and at 1.30 was witnessed what was then supposed to 

 be the final struggle with death ; after the expiration of four 

 minutes this recommenced, and was continued at irregular inter- 

 vals until 1.43, after which no more were seen until 1.48, at which 

 time a tiny bubble of air made its way from the cork, and at 1.58 

 another of these bubbles evoked a feeble struggle, the last. When 

 finally removed to the alcohol bottle, not a movement could be 

 detected, and the insect was undoubtedly drowned. This would 

 appear to antagonize the theory that they stay beneath the sur- 

 face in stormy weather. I also noticed, with several specimens, 

 that they could (or would?) only dive after being wet so that 

 "skimming" was impossible, but this does not agree with the 

 observations of other naturalists, and I recognize its practical 

 worthlessness as purely negative evidence. 



No more Halobates were seen during the trip, though on the 

 1 5th of July some little water bugs, at first taken for the larvae 

 of Halobates, were noticed at the extrance of the Spanish Wells 

 harbor, skimming about on the surface of the water. Several 

 pairs were observed in coitu and directed suspicion to the cor- 

 rectness of the reference -so Mr. Heidemann looked up the 

 matter and found them to be Rhagovelia collaris Burm., hitherto 

 supposed to be a fresh water species exclusively. Fortunately, 

 one couple was taken in the net, and did not separate on being 

 placed in a tub for examination, so the following details were ob- 

 served: The male rests rather tar back on the female, his head 

 reaching to a point immediately above the insertion of her mid- 

 dle legs. His anterior legs are then extended forward and bent 

 at the "knee" nearly at right angles, so as to clasp the female 

 thorax a little in front of the widest portion in such a way as to 

 prevent his sliding off. All progression is then performed by the 

 female, the male resting quietly on her back. His hold was very 



