1893.] 231 



annexed sketch, taken from a specimen of H. princeps set in the 

 attitude of repose). On the approach of the finger, or a pencil, they 

 dive readily, and swim with great facility beneath the surface, the air 

 entangled in the pubescence giving them a beautiful appearance like 

 that of a globule of mercury or polished silver. This supply of air 

 must be essential to the existence of the insects, which I feel sure 

 must pass a large part of their life beneath the surface of the sea, 

 diving into undisturbed water in rough or even moderate weather, and 

 coming up again only when it is absolutely calm. 



I am not able to give any information as to the food of Salohates, 

 further than that I have once or twice seen several specimens congre- 

 gated round small floating objects, such as fragments of seaweed, &c., 

 as if deriving nutriment from it. Nor do they seem to be themselves 

 preyed upon by other marine creatures ; at all events, I have seen 

 plenty of small fish about when Halohates has been on the sea surface, 

 but have never seen the fish " rise " to the insect. Their integuments 

 are so tough and leathery that they can scarcely be very dainty 

 morsels. 



The union of the sexes takes place on the surface of the sea, and 

 the eggs are unquestionably carried about by the $ , attached to the 

 extremity of the abdomen, for some time before she parts with them. 

 Among my specimens from the Marquesas Islands I have found two 

 females of H. Wiillerstor-ffi , one of which has three ova, and the other 

 a single one, thus attached. These eggs are cylindrical in shape, with 

 rounded ends, of a deep ochreous-yellow colour, and the envelope is 

 of very tough and firm consistence ; their size compared with that of 

 the parent is enormous, as they are just over 1 mm. in length. "Where 

 they are eventually deposited still remains a mystery, though Prof. 

 Uhler (Standard Natural History, vol. ii, p. 269, Boston, 1884) suggests 

 that the insects may use Sargassum or other fioating seaweeds as a 

 nidus. I have found young larvsD in the tow^-net in the Central 

 Pacific many hundreds of miles from land, but I am inclined to think 

 that oviposition is usually effected within a much less distance of the 

 coast ; and all my experience tends to show that the creatures are 

 much more numerous near the land (or at any rate, more frequently 

 observed in large numbers) than in the open ocean. 



Dr. Buchanan White, in his " Eeport," enumerates eleven species 

 of Halohates as known to him, and two others have since been de- 

 scribed, H. Whiteleggi, Skuse (Eecords of the Australian Museum, 

 p. 174, plate 1), and H. regalis, Carp., both from Australian w^aters. 

 Some of these appear to be very restricted in their distribution, while 



