191 



them to the Chalcidoid series, and Ashmead, the latest writer 

 whose works I have used, agrees with him. On the oither hand 

 such famous entomologists as Westwood, Foerster and Thom- 

 son all agree in placing them in the Proctotrupoid family; and 

 I believe that this great difference m opinion is a true indicaton 

 of the anomalous character of the Mymaridae. While they 

 cannot, judging from the forms that I have studied for the pur- 

 poses of this paper, be possibly included amongst the Proctotru- 

 poids, it must be admitted that, if they be referred to the Chal- 

 cidoid series ,they form an unusually distinct family in the lat- 

 ter. In fact they are probably more distinct from any family in 

 this series, than is any other of its families, from that to which 

 it is most nearly allied. Unfortunately in the literature ac- 

 cessible to me the structural characters of the genera are for 

 the most part merely given in dichotomous tables, and therein 

 characterized in such very brief fashion, that I cannot determine 

 with certainty whether the species here described really belong 

 to the genera, to which the tables assign them, or not. None 

 of these tables make mention of characters, which I believe to 

 be of great importance for purposes of generic division. 



The family Mymaridae has been divided into two sub-families 

 according to the number of the tarsal joints, and each of these 

 is further divided into two 'tribes' according to the nature of the 

 basal articulation of the abdomen. For purposes ot identi- 

 fication of genera these divisions are convenient ; as a natural 

 classification their value appears to me very dubious. The 

 variability in the number of tarsal joints in those families of the 

 Chalcidoid series, which are most nearly allied to the Mymaridae, 

 i3 known to all students of parasitic Hymenoptera to be of in- 

 ferior value, this number in some cases difTering even in the 

 sexes of one species. It is quite probable that the nature of 

 the articulation of the abdomen, and possibly the dififerences m 

 the structure of the thorax, will prove of much greater import- 

 ance than the number of tarsal joints. 



To me, by far the most remarkable character in the Mymarids, 

 that I have studied, is the extraordinary difference in the rela- 

 tion of the base of the abdomen to the posterior end of the tho- 

 rax, even in species that resemble one another in their general 

 habits. In the species that I describe under the genera Aiiagnis, 

 Paranagrus and Alaptus the abdomen is truly sessile, adapted at 

 the base to the thorax, the rigidity being further increased by the 

 great thoracic mesophragma, which penetrates well back into 

 the abdomen. Forms that I refer to Ooctonns, Polyncma etc., 



