INTRODUCTION. XIX 



Some of these species may doubtless be distinguished by their 

 size (E. villosa), or by the examination of their wings (E. Z- 

 cincta) ; but in most cases, if it is not known where they come from, 

 there is no possibility of determining them, not even of referring 

 by sight to their females. 



The same fact may be noticed in a very developed state among 

 other species of Elis which present another system of coloring. 

 and which forms a part of the subgenus Trielis. Thus the 



Elis interrwpta (Europe) — texensis (Texas) — Xantiana (Cali- 

 fornia), 



though offering females very different from each other, coincide 

 in the males. 



I do not pretend to affirm, however, that all these males are 

 perfectly identical. I only mean that they are so close neighbors 

 that I have not succeeded in finding between them differential 

 characteristics which appear to distinguish them. 



The preceding will suffice to show that the femaie ought, at 

 all events, to be taken as type of the species. 



Preparation. 



In a great many collections, it is usual to spread the wings 

 and legs of the hymenoptera. This is mere amateur's work, of 

 no utility for study, sometimes even quite opposed to the purpose 

 in view, by dissimulating the character of the insects instead of 

 exposing it to view. This practice is to be regretted, moreover, 

 by its increasing the value of the insects, on account of the time 

 and expense wasted thereby, so that one is loath afterwards to 

 place them in the softener when it becomes necessary to dissect 

 the mouth. 



It is usual with me to prepare each of my insects only when 

 about to study them, because by so doing I can, after having 

 softened them, give them whatever arrangement may be best cal- 

 culated for bringing into relief such special characteristics as are 

 particularly useful in the group to which it belongs. As the use- 

 ful characteristics are not the same in each group, the preparation 

 ought to vary according to the species. Of the Epijwna % the 

 mandibles are to be opened, the legs to be stretched downwards, 

 so that it may be possible to examine distinctly the haunches and 

 thighs; of the Odynerus I, approximate to the group of 0. 



