262 Rev. T. A. INIarslmll on Catalogue of 



perhaps a different judgment in each case, the cataloguer 

 has aimed at exchiding all matter so doubtful as to be 

 useless, and to render the references complete in all cases 

 of certainty. If the line waves more or less, he must 

 shelter himself under the plea that it could not be other- 

 wise. In Avorking out these views, it must occasionally 

 happen that the references do not travel back to the earliest 

 inventor of a now unrecognizable name, but stop short at 

 the first describer of an unmistakeable thing, or in most 

 cases, Gravenhorst, and sometimes not the ancient autho- 

 rities he quotes. Priority has been a first object or hobby 

 with the compiler, but the hobby has not been ridden to 

 death. 



Mr. Walker has remarked, at the end of his " Xotes on 

 Chalciditij," pul)lished in the j^resent year, that " some 

 alterations are required in the arrangement of the families, 

 and the genera and their respective species have yet to be 

 examined in detail." Pour evcourager les autres, the 

 same judgment, or something very like it, may be pro- 

 nounced upon the Ichneumonidi^. To descend no further 

 than to the division of genus, the Avant of an uniform 

 standard is very conspicuous. Some genera are eminently 

 exclusive, and others in the highest degree latitudinarian. 

 The 1,186- species of Ichneumonidte are comprised in 136 

 genera ; while the Braconidte, numbering only 439 spp., 

 are distributed among 125 genera, only II fewer than 

 those of the tribe preceding them. This disproportion, 

 the result of a totally different idea of Genus in different 

 minds, is mainly due to the labours of Forster, who has 

 established a very great number of generic divisions among 

 the BraconidaB, founded frequently u\)on minute characters, 

 not involving general appearance and structure, and which 

 to others have seemed only of specific value. We have 

 then at present a mass of very unequal composition, tending 

 both ways into extremes, about half-way between which 

 the truth in other matters is commonly considered to lie. 

 General resemblance and structure (interpreted Avith a 

 certain moderation) is probably the central point at Avhicli 

 these oscillations must cease. Thus, Enicospilus and 

 Opliion fall conAxniently into one genus, Opliion ; ScJii- 

 zoloma, Exochilum, Ilctcropehna, Anomalon, Agrypon 

 and Trichomrna,iii present only separable Avitli a poAverful 

 lens, fall easily into Anomalon, and so forth. But these 

 considerations, being beyond the province of a compiler, 

 have not been alloAved to a^jpear in the Catalogue. 



