236 Rev. ¥. D. Morice un Armatures, etc. of Andrena. 



a bold and almost angular curve (see Fig. 4 f?c^), begins to 

 develop itself almost from the apex of the stipes downwards, 

 the interior margin of the process passing so gradually 

 into the apical outline of the lobe, that it is impossible to 

 say whereabouts, oven, one ends and the other begins. On 

 the interior outlines, also, the lobes merely touch at one 

 point and then instantly diverge, leaving a great gap 

 between their bases. The little knob-like twist at the 

 apex of each process is altogether eccentric. And finally 

 the armature is quite enormous in proportion to the size of 

 the insect, e.g., it is twice as broad, and more than twice as 

 long, as that of cctii, Schrank, in a number of examples of 

 each species that I have measured. Genevensis, Schmied., 

 a species superficially a good deal resembling it, has 

 an utterly different, small and simple type of armature, 

 rather resembling that in Fig. 7. Cctii, Schr., has a some- 

 what unusual form of lobe, the angles being distinctly 

 obtuse, but its other terminal characters are not re- 

 markable. 



As to the characters of the process, they arise chiefly 

 from the different manner in which, according to the 

 species, it is bent or creased over inwards near the apex. 

 The part beyond this crease forms a sort of " dog's-ear " 

 (Fig. 4< a g h) which in the dorsal view faces the observer 

 almost directly, and so looks dilated in comparison with 

 the part below it {h h c), which is seen more edgewise. 



When the crease is near the apex, the narrow looking 

 part becomes long and stalk-like (Fig, 9) and the process 

 assumes a sort of petiolate character, which is hardly to be 

 recognised in Fig. 4., etc., where the " dog's-ear " is longer. 

 But whatever the position of the crease, a process may look 

 petiolated, if the lobe is so sinuated (Fig. 19) as to increase 

 the apparent length of the process. 



In the most normal forms of the process its exterior 

 outline forms a single convex curve, and deviations from 

 this always give a peculiar look to the armature. Thus a 

 very abrupt creasing of the process nearly transversely 

 (ordinarily it is diagonal), substitutes a distinct angle for 

 the simple curve. I believe I could always recognise an 

 analis, Panz., armature (Fig. 21) by its heptagonal look, due 

 partly to its angled process, and partly to similar features 

 in the exterior outline of its lobe, Florca, Fab., again 

 (Fig. 7) seems to me very distinct by its nearly straight 

 processes, with a very slight sinuation inwards, the exterior 



