166 Rev. T. A. Marshall's monograph of 



lacteus, Nees; and Hypera variabilis, Hbst., by A.sericeus, 

 Nees. A. hrcrive7itris, Ratz., has been reared from the 

 hymenojDterous Andriciis terminnlis, Fab. ; and in the 

 Mayr collection at Vienna, according to Reinhard, is a 

 series of A. albipennis, Nees, bred from Andriciis multipli- 

 catus, Gir. A. lineola, Hal., has without doubt been reared 

 from the larva of a dipterous insect, SyrjjJius pyrastri, 

 L. ; a,ud A. scahricidus, Jieinh., probably from the galls 

 of Cecidomyia rosaria, Loew. A. idtor, Reinh., by 

 Brischke from the larva of a Chrysopa. Reinhard, who 

 has collected most of the above notices, is of opinion 

 that as the greater part of these parasites have also been 

 bred from Lepidoptera, the observations connecting 

 them with other orders require further confirmation. 

 For example, Andricus midtiplicaias, Gir., makes galls 

 which are abundantly infested by a moth, audit is much 

 more likely that A. albipennis, Nees, is parasitic upon 

 this than upon the Cynips. Similarly A. ensiformis, 

 Eatz., issuing from the galls of Rhodites rosa, L., is not 

 thereby proved to be a parasite of that Cynips. The 

 references to economy in Dours' Catalogue have not 

 been copied, as the determinations of the species by 

 Ferris or Goureau are unreliable. The maggots of 

 Apanteles are themselves subject to the attacks of hyper- 

 parasites, belonging to the Ichneumonideous genera 

 Hemiteles, Pezomachus, and Mesochorus, as well as to the 

 family Clialcididce ; some instances of this we shall have 

 occasion to notice hereafter. The fine silk cocoons spun 

 by the larvae of Apanteles are well-known objects, and of 

 great use in the difficult task of determining species. 

 They are either white, yellow, or buff-coloured, and often 

 protected by an external web enveloping the whole 

 brood, except in the case of the solitary species ; in a 

 few instances they are arranged in an alveariform mass, 

 which always denotes that their victim was a geometrical 

 larva. Roughly speaking, we may lay it down as a rule 

 that all the species with a long terebra (Sect. II.) make 

 white cocoons ; as also those of Sect. IV. Yellow cocoons 

 belong to comparatively few species, of Sect. I. and III. 

 Bufl'-coloured cocoons are perhaps limited to two or 

 three species, as A. juniperatce, Bouche, and difficilis, 

 Nees. 



Haliday, in the Eut. Mag., 1834 (ii., pp. 242—254), 

 described 33 species of this genus, while the number 

 contained in the present work is 09. This increase is 



