3G Rev. F. D. Morice on male terminal segments and 



sees. To estimate correctly (e.g.) the relative length of 

 two antennal joints seems a simple matter ; yet even here 

 the most practised eyes will sometimes be deceived. 

 Again, the comparative length and breadth of the gena 

 in a particular insect is a definite fact ; yet unless it be 

 placed in exactly the right position when we measure it, 

 the best micrometer will measure it quite wrongly. 



And as to infuscations, etc. of the wings, there are 

 but few Colletes-species in which such characters can be 

 employed to any purpose, and even in these species the 

 phenomenon is apt to be inconstant. 



Accordingly in constructing tables for Colletes based 

 on such characters, I am simply making the most that 

 I can of unsatisfactory materials. The fact is, that through- 

 out the whole genns the external structure is particularly 

 simple and uniform : strong paradoxical characters such 

 as help us to classify other genera are here scarcely to 

 be found * without dissection of the insects. And though 

 in quite fresh specimens the colour and disposition of 

 the pilosity in certain species may be adequate to dis- 

 tinguish them for certain, a slight amount of fading or 

 rubbing will make almost any Colletes practically un- 

 recognizable by such characters. Whatever its original 

 colours may have been, it speedily bleaches into a dull 

 uniform grey. The fasciie which now lojk white may 

 once have been distinctly yellow ; and those which are 

 now widely interrupted, may or may not have once been 

 entire. Hence, if we are to determine at all any but 

 exceptionally perfect specimens, we can do so only by 

 employing structural characters ; and these in Colletes 

 are, as we have seen, for the most part minute, easily 

 misinterpreted, and still more easily misrepresented when 

 we try to express them in a verbal description. 



If then my Tables prove to be of practical use to 

 hymenopterists, I shall be glad ; if not, I shall not be 

 much surprised. 



The length of the gena being perhaps the most con- 

 spicuous of the external characters, I have taken it as 

 my first ground of division. But I do not at all think 

 that it divides the species into real natural groups. (Cf. 

 the figures of gense in Plate VI, which, different as they 

 are, belong all to insects which I believe to be nearly 



* The scutellar appendages of graeffei, Alfk., are an isolated 

 exception. 



