4 PROCEEDINGS OF THE NATIONAL MUSEUM vol.80 



pressed as a synonym of DoUchozele. And the latter name, in my 

 opinion, must in turn fall as a synonym of Macrocenti^us. Apart 

 from the very short ovipositor, the long calcaria of the posterior 

 tibiae, and the long acute metapleural tooth, Dolichozele does not 

 differ from normal Maci'ocentrus^ and these differences are appar- 

 ently only comparative. Paniscozele atreitarsis, one of Enderlein's 

 originally included species, was described as havin,g an ovipositor 

 nearly as long as the body; while the metapleural tooth, though 

 usually less prominent and acute, occurs in nearly all species of 

 Macrocentnis, and the long tibial calcaria are found in some. Accord- 

 ingly, since it has seemed impossible to distinguish clearly between 

 Dolichozele and Macrocemtrus^ I have placed the former, and its 

 synonym Paniscozele^ in synonymy under Macrocentrus. 



I have also seen the type of Fhogra i^romaculata Cameron, 

 which is in the British Museum, and that of Metapleurodon ceyloni- 

 eus Enderlein, which is in the Pomeranian Museum of Natural His- 

 tory, the genotypic species, respectively, of Fhogra Cameron and 

 Meta'pl&ufrodon. Enderlein. In my opinion both belong in 

 Macrocentrus. 



Apparently all species of Macroceivtrus are internal parasites of 

 lepidopterous larvae, and at least in most cases the host larvae live 

 more or less concealed, principally as leaf rollers or as borers. Most 

 of the species seem to be solitary parasites, but some are gregarious, 

 and in the case of the latter type it has been frequently observed that 

 all adults obtained from a single gi'oup of cocoons are usually of one 

 sex. This naturally has suggested the probable occurrence of poly- 

 embryony, and recently Parker ^ has shown that this method of repro- 

 duction does occur in Macrocentrus gifuensis Ashmead, a gregarious 

 parasite of the European corn borer {Pyrausta nuhilalis Hiibner). 

 No doubt it is common to many species of the genus. 



Species of MacTOcentrus appear to be rather less specific in host 

 selection than those of many other groups of Braconidae, and the 

 same host species may be attacked by several different forms. Four 

 species of Maci^ocenti^us^ for example, are known to parasitize the 

 larvae of La^^'peyvesia inolesta Busck, the introduced oriental fruit 

 moth ; while 9 different host species have been recorded for M. ancyli- 

 voinis and 12 for M. delicatus. A list of the hosts of Macrocenti^ns 

 recorded in this paper is given on pages 53-54. 



Owing to the extent of variation occurring within species, it has 

 seemed desirable to make the key rather full and to give detailed 

 descriptions of the species. 



« U. S. Dept. Agr. Tech. Bull. 230, 63 pp., Illus., 1931. 



