3 



to be found on umbelliferous flowers. Some of the species are 

 very little known, and there is a great likelihood of two or three 

 fresh species being discovered. The largest known species of any 

 of the Hymenoptera are some of the tropical members of this 

 family, which are able to overpower the very largest spiders. 

 Though of very similar habits in general, one species at least, P. 

 niijer, occasionally preys on caterpillars, and the two species of the 

 genus Af/euict frequently nest in old posts. 



Of the two species of Aatatuit, the larger one, A. hoops, is by far 

 the commoner. Both, however, occur where there are good-sized 

 patches of bare sand. The smaller, A. stirnna, I have only taken 

 on the sandhills at the mouth of Brading Harbour. Both species 

 rest on the sand in the hottest sunshine, and it requires a very 

 quick stroke of the net and a fully average amount of patience to 

 effect a capture. The males of both species have exceptionally 

 large eyes, which meet on the vertex of the head, and they certainly 

 make good use of them. The prey of A. stiguia is as yet unknown, 

 as far as I am aware. A. hoops usual prey appears to be the larva 

 of one of the larger plant bugs belonging to the genus Pentatoma, 

 though, according to Smith's observations, it also preys on another 

 kind of Sand-wasp, O.ri/belits nny/ltniiis. The closely-allied genus 

 Tach]/tes comprises three species, two of which are very rare in this 

 country. The third, 2'arhiftes pectinijies, is abundant almost every- 

 where. The usual prey on the continent is believed to be larvae of 

 Orthoptera, and Smith took it at Weybridge with a small grass- 

 hopper. Shuckard, however, says that he has frequently taken it 

 with a small, sandy-coloured caterpillar, and this entirely accords 

 with my own experience. Another very closely-allied insect is 

 Dinctiis pictiia, which, I believe, rests only on old records from 

 Ascot and Windsor, in the early part of last century. The small 

 species of Miscophns occur on sandy commons and prey on spiders, 

 one of them ili. luaritimiis, being found only on the sandhills at 

 Deal. Now we come to the three species of Trjipoxyhn, with their 

 long slender bodies, which put one in mind of Anuiiophila. The 

 neuration of the wings at first sight resembles that of the genus 

 Crabro, but on closer inspection, extra cells are seen, though 

 enclosed by such very fine nervures as to easily escape observation. 

 They are all fairly common, and nest either in the ground, in old 

 posts or palings, or in bramble stems, preying on spiders. The 

 least common of the three, T. attenuatiiiii, I have several times bred 

 from perforated bramble stems. 



The very widely distributed genus Aiiiiaophila is represented in 

 this country by four species. The most abundant is A. sabnlosa, 

 which is to be found in most sandy localities in the south of 

 England, its long legs and long slender body putting one somewhat 

 in mind of a Crane-iiy. The petiole of the abdomen is so slender 

 as to give the rest of the abdomen quite a detached appearance. A 

 very similar species, but with a stalk or petiole to the second sub- 



