18S8.] 253 



metathorax and scopa of OoUetes, see Trans. Ent. Soc. Lond., 1878, 

 pi. vi, figs. 5, 6, 12, 12rt) : also both Prosopis and CoUetes have the 

 lingua beautifully and finely ridged transversely, and set with bristles, 

 a peculiarity which Dr. Miiller attributed more especially to the higher 

 Hymenopiera {vide p. 56), although it exists also conspicuously in some 

 of the sandwasps. 



One can hardly help feeling a doubt as to whether the habit of 

 lining the brood cavities with slime can have had much to do with the 

 shape of the tongue. It is hard to see why Sphecodes, &c., should 

 require " deeply seated honey " more than Prosopis, also (unless this 

 desire be granted for Prosopis as w^ell) why if their ancestors, the 

 sand wasps (Jide Dr. Miiller), had short bifid tongues, without any 

 slimy habits, the slimy habit of Prosopis should be relied on to account 

 for its conservative tendencies. 



Following Dr. Miiller to his remarts on Sphecodes, at p. 50, 

 one reads : " In Sphecodes the whole body is sparingly covered with 

 hairs which show the first traces of feathery branching." As I have 

 already pointed out, far more distinctly branched hairs exist in genera 

 which in Dr. Miiller's scale would be placed decidedly lower down ; 

 I may here mention a peculiarity in Sphecodes and Halictus, which, so 

 far as I know, exists in no other genus of the Anthophila, viz., that 

 they have no apparent apparatus for projecting the mentum and 

 lingua ; in most of the genera this office is performed by what are 

 called the lora (or retractors. Dr. M.), these are two joints, each of 

 which swings on the apex of one of the cardines, and are united at 

 their apices, so as to form a A-shaped body ; the mentum, which bears 

 the lingua at its apex, is attached to the centre of the A, and is 

 retracted or projected according as the A is inverted or not. In some 

 genera, Andrena, Colletes, Prosopis, and Panurgus, the lora do not 

 form distinct arms, but the membrane which covers the space between 

 the cardines is chitinized and darkened towards the apex, where its 

 convexity forms an apical arch ; to the centre of this arch the mentum 

 is attached, and the membrane being elastic, the arch performs the 

 same functions as the lora, only, probably, in a less perfect way ; but 

 in Sphecodes and Halictus,! have failed to find either lora or chitinous 

 arch : and in this respect I should say they are decidedly lower down 

 in the scale than Prosopis or Colletes. 



I have mentioned Panurgus above as one of those genera in which 

 no distinct lora occur, and in this respect it would appear to be 

 decidedly less advanced than Macropis, Dasypoda, and Cilissa, although 

 in the elongate lingua, and the form of the labial palpi, as well as in 



