1865.] 211 



Ou the other hand, in other lehneumonidous genera there exist homo- 

 logous spots, but often different in number or difiPering slightly in their 

 location, though they are manifestly modifications of the same primordial 

 pattern. For example in Tror/itu^ where the pentagonal areolet oi Ich- 

 neumon becomes rhomboidal by the elimination of the upper side, A 

 generally covers the whole angle formed above by the union of the 

 inner and out<3r sides of the areolet, and the other four bullau are placed 

 as in Jrhnrumon. lo Fimpla and Ephialtcs, which also have a rhom- 

 boidal areolet, A is placed above B on the upper end of the outer side 

 of the areolet, and in Plmpla is generally separated from B only by a 

 very small space, and sometimes entirely confluent with it, the other 

 three bullae being located in both genera nearly as in Ichneumon^ ex- 

 cept that E is usually closer to the angle of the first recurrent vein. 

 In Cri/piua there are normally but four bulla), C and D being conflu- 

 ent and the others placed as in Jrhueumon, except that B is located 

 higher up on the cross-vein which it bestrides. In Glijpta^ where the 

 areolet is represented by a simple cross-vein, A and B are absent, but 

 as C and D are not quite confluent, being divided by a slender black 

 line or black dot, there are three bulla), C, D and E. Odontomerus^ 

 Xi/lonomus, Acsenitus and Aretes difier from Ghjpta chiefly in Cand D 

 being separated by a very wide space, and have the same number of bulliB. 

 And in Ophlon and Anomalon, where the areolet is also represented 

 by a simple cross-vein, A and B are absent, but C and jD being per- 

 fectly confluent, there are consequently but two bullae. So far as I can 

 discover, on a careful examination, there is no Ichneumonidons genus 



either the shape of the "central area" of the metathorax, or the number of 

 joints contained in the antennal annulus, a constant and reliable specific cha- 

 racter. In very many well-marked species of which I possess numerous speci- 

 mens, the width of the "central area," as compared with its length, varies 50 

 per cent, with all the intermediate grades, i. e. varies from " transverse" to 

 "quadrate" or from " quadrate" to "elongate;" and in other such species the 

 length of the antennal annulus varies very considerably, often by three or four 

 and in one species by as much as seven joints, with numerous intermediate 

 grades. It is even the case that in one undi'seribed species, allied to ^jarix'S 

 Cresson, of which I possess seven %, a single % has two or three of the inter- 

 mediate antennal joints marked with yellowish-white above, while the remain- 

 ing six % have no such markings whatever. And Westwood records the fact tliat 

 "two 9 of Ciyptus bellosus were reared by Mr. Thwaites, one of wliich had the 

 antennsB annulated and the other entirely black." {Intr. II. ]>. 1.38, note.) If 

 all specimens that differed in the above characters were considered as distinct 

 species, the number of species in my collection would be very largely increased, 

 and my argument strengthened so much the more. 



