240 Rev. F. D. Morice on Armatures, etc. of Andrena. 



In most, if not all cases, where the geniculation is 

 lobate the process is at least a little dilated laterally at 

 this point (see Fig. db). This is very noticeable in Fig. IQb 

 (drawn without the pilosity to emj^hasise the outlines), 

 where another very curious phenomenon is to be observed, 

 viz., that the thickened (dark) part of the process behind 

 has an outline of its own which is cruciform (that of the 

 process itself being hexagonal). If for any cause (as by 

 absorption or folding inwards of its frail transparent part) 

 the actual outline of the segment should become identical 

 with that of the thickening, we should get exactly such a 

 cruciform process as we find in Plate VII, Figs. 6, 7, 8, 9. 

 Comparing these Figures, and also Figs. 10 and 11 of the 

 same plate, with Figs. 9 to 15 of Plate VI, it appears to 

 me that we can trace the gradual stages in a process of 

 development, through which it is possible at least that the 

 " cruciform process " may have arisen. I do not mean to 

 suggest that, as a species, scita (Plate VI, Fig. 16^) is to be 

 placed necessarily between ovina (or humilis) and scnccionis, 

 but only that, in this particular matter of the lateral 

 development of the " process," it shows a condition which 

 is certainly intermediate and may be transitional. As a 

 species it may belong to quite a different group. But if a 

 species, really belonging to the group of humilis and 

 senecionis, and really transitional between the two, exists 

 and should hereafter be discovered, I think it is highly 

 probable that its " valvula ventralis " will resemble that of 

 scita. 



I have now only to add a few stray notes on sundry 

 details in my figures (Plate VI). 



Fig. 1, tlioracica, Fab., shows a slightly paradoxical type, 

 with strong but not lobate geniculation (la x), rather short 

 broad and cup-like process, and pilosity developed as a 

 pair of definite lateral fringes whose outlines run some- 

 what parallel to the sides of the process. 



A similar type, combined with strong similarity in the 

 armature, occurs in nitida, Fourcr., morio, BrulL, alho- 

 punctata, Rossi, etc. 



Figs. 2 and 8 are introduced mainly because the species 

 will be mentioned in the paper following this. I will 

 here say only that they are tolerably normal forms, except 

 that the process in Iraunsiana is wide for its length, and 

 that the apex in rufo-his'pida, Dours, is rounded not bilo- 

 bate, at any rate in this particular specimen. 



