Records of Bees. 493 



According to this scheme the several forms may be 

 tabulated thus : — 



Face-markings. Marks at side Hind basi- Fifth ventral 



(Colour.) ofclypeus. tarsus. segment. 



A. zonata — — (2) + — 



A. korotonensts . . + —(--') — — 



A. calccifera .... — — (1) — -f- 



A. cahlwelli + — (li — — 



A. subccendea, . . , -f- + -j- -|- 



According to this plan^ A. suhccerulea seems to be the 

 stem-form, while the others represent various degrees of 

 reduction. A. kuroto7iensis would seem to have been derived 

 from the Chinese caldivelli ; while zonata and calce/fera may- 

 represent independent derivatives from subccerulea. It is 

 possible, however, that the forms inhabiting the same region 

 (as zonata and calceifera in Formosa) may freely cross, with 

 Mendelian results, in which case calcnj'era cannot be con- 

 sidered a valid species. Against this last supposition is the 

 fact that the several characters seem to go together, with the 

 single exception in regard to the clypeal marks mentioned 

 under calceifera. 



Still another form from North-west India may be 

 described as follows : — 



Anthophora comberi, sp. n. 



Hair of head and thorax above pale reddish mixed with 

 black, as in zonata ; abdominal bands very brilliant shining 

 turquoise-blue; face-markings light yellow ; size of zonata. 



d . — -Lateral markings o£ clypeus reduced to small pyri- 

 f orm sutural spots (a third reduction stage) ; hind basitarsus 

 with a few white hairs at extreme base ; fifth ventral segment 

 elevated in middle and broadly emarginate (Nasik, = type). 



. ? . — Clypeal marks large and elongate-quadrate, shaped 

 as in suhc^rulea, and with the median yellow band broad ; 

 spots at basal corners of labrum small ; hind basitarsi with 

 the hair all black ; light hair of legs white, with a brown 

 stain apically on middle tibise. Hab River, Karachi, Sept. 

 1909. 



Hah. Nasik and Hab River {E. Comber). British 

 Museum. 



In the above scheme this will stand as -h, — (3), -f ?, — , 

 This is a considerably larger insect than A. subccerulea. 



