XVI INTRODUCTION. 



Marshall in the Braconidous genus Pambohis, Hal. Even when of normal 

 length, the wings of some of the genera of the Cryptinae have the penta- 

 gonal areolet imperfect, through the absence or weakness of the external 

 nervure, and in a few species even the internal nervure is so short as to 

 render the areolet very irregular. 



In the classification of the Cryptinae, I have in the main followed 

 Prof. Thomson's scheme, though in placing the Phygadeuonides before 

 the Cryptides I think that Taschenberg, who treated the sub-family more 

 as part of a whole than have the later authors, had more regard to the 

 natural order. In their alar neuration and usually fully developed meta- 

 thoracic costae the Phygadeuonides certainly bear a closer analogy to the 

 Ichneumoninae than do the Cryptides, on account of the often sub-quad- 

 rangular areolet and obsoletely costate metathoraces of the latter. It has 

 been truly said that Hemiteles is Phygadeuon in miniature, and wherever 

 the latter be placed it cannot be separated from the former : if the Phyga- 

 deuonides be placed first, Hemitelini must next follow. As I have said 

 elsewhere, Pezomachus is divisible into several genera, though I think it at 

 present inexpedient to adopt the divisions ; the antennal conformation must 

 form the basis of these genera, some of which (the Sylvicola group) will be 

 most closely allied to Phygadeuon, while others (the Instabilis group) will 

 be closely related with the Spinolia group of Hemiteles. Till these genera 

 be erected, however, it is better to place Pezomachus, on account of its 

 lower development and the antennal conformation of the majority of its 

 species, next after Heniiteloides. The question of the natural position of 

 the Stilpnides is much more puzzling, if indeed such do at all exist ; the 

 shortly, and in some cases not at all, exserted terebra allies them with none of 

 the present sub-family, but the apically incomplete areolet and small size 

 resemble Hemiteles, while the curious conformation of the metathorax, 

 with its comparatively few areae, the elongate form, and (in Atractodes) 

 slender antennae are more comparable with the Cryptides. These latter, 

 in their genera Xylophurus, Nyxeophilus and Acroricnus, bear such strong 

 Pimplid facies that it appears most natural to place them at the end of the 

 present sub-family, immediately preceding the sub-petiolate Xorides ; and 

 here, perhaps, I should also have included Helcostizus, a most specialized 

 insect, forming, as has several times been pointed out, a connecting link 

 between the Phygadeuonides and Cryptides. 



If, however, one follow up this latter connection it leads to the 

 conclusion that the Stilpnides are placed first or last, immediately after 

 Ichneumoninae, with which they possess nothing but the concealed 

 terebra in common or next before the Pimplinae, from which they too 

 materially differ. And, further, that the curious apterous Pezomachi fall 

 out of sequence entirely. On the whole, I consider it better to take the 

 semi-circular view ; the Phygadeuonides at the beginning falling in 

 development to the Hemetelini and, finally, the Pezomachini (to the 

 indeterminate condition of whose specific characters I have referred), 

 then rising through Stilpnus with its short and stout antennae, Atractodes 

 with its incomplete metathoracic areation and imperfect areolet, the 

 Mesostenini with irregular areolet and peculiar areation, to the Cryptides, 

 which, although highly developed, do not bear the full complement of 

 metathoracic areae, and are consequently certainly most closely related to 

 the Pimplinae. Our present classification is, however, notoriously super- 

 ficial, and the above, it will be seen, is based purely upon facies. 



