192 



on the other hand have a mobile, pedicellate abdomen and in 

 these and petiolate forms, the very structure prohibits the pre- 

 sence of such a mesophragma in the abdominal segments. Yet 

 in spite of these profound differences we find insects of the first 

 class ovipositing^ in the eggs of leaf-hoppers laid in the leaves of 

 cane or grass ; tho'se of the second doing precisely the same, as 

 well as piercing the eggs of Heteroptera, imbedded in a similar 

 manner in the cane leaf. It does not appear to me improbable 

 that, when a special study is made of the peculiarities of the 

 Mymaridae, and this study is based on a rich and well prepared 

 collection, the group will prove to be of more than family rank 

 and itself to contain some very distinct families. 



The Mymaridae, or at least such as I have studied in con- 

 nection with this paper, appear to me to be most nearly allied 

 tf> certain of the Eulophidae. Some of the latter have to a large 

 extent the same structure of the thorax as the Mymarids, and in 

 this point amongst others the latter at once differ from any true 

 Proctotrupids known to me. It is further a somewhat remark- 

 able fact that various Mymarids of very different groups have 

 a pattern and arrangment of markings identical with that of 

 certain Eulophidae. The wings are of the most characteristic 

 form, and are very different from those of the egg-parasites of 

 the family Trichograiiimidac, from which I exclude Ashmead's 

 sub-family Oligositinae, believing it will be sooner or later asso- 

 ciated with the Eulophidae in spite of the three-jointed tarsi. 

 In the presence of a great mesophragma extending far into the 

 abdomen, the sessile-bodied Mymarids agree with many Eulo- 

 phids and Trichogranmiids, indeed in some of the latter this 

 structure is enormously developed and extends almost to the 

 apex of the body. 



Ashmead who has published the latest classification of the 

 Chalcidoid families, and has probably spent more time on this 

 study than anyone living and'whose views are therefore worthy 

 of the most careful consideration, gives as the leading characters 

 to define the Mymaridae : 



"Hind wings exceedingly narrow, linear, pedunculate at base ; 

 ovipositor issuing from beneath just anterior to tip of abdomen ; 

 antennae without a ring-joint, the scape rather small, short, 

 compressed:" 



All the species considered here agree with this definition in 

 two particulars ( 1 ) the hind-wings are narrow and pedunculate, 

 (2) the antennae have no ring-joint. In other respects the 

 characters do not agree with any of my species. Thus the ovi- 



