( xliv ) 



of the cells as " pitcher-shaped " is hardly borne out by the 

 orio-inal account, and that no mention is made of the mud 

 being " mixed with small pebbles " as stated by Smith 

 (" Brit. Bees," 2nd Ed.). 



It is to be observed that all six species, except the last- 

 named, employ ready-made holes or covered spaces, or 

 excavate in very soft material, possessing in a more generalised 

 form the same kind of instinct as that of the shell-using species. 

 Furthermore, all the four British species in which the instinct 

 has been recorded, except spinulosa, of which very little is 

 known, are described as also making use of other holes for 

 nidification and thus possess the generalised form of the 

 instinct as well as the specialised — a necessary combination for 

 any shell-using species reaching a locality where snails are 

 not readily to be found. Even bicolor, with an instinct as 

 complex and specialised as any known in the genus, is said 

 to burrow in banks. 0. rufa, so commonly known to possess 

 the generalised instinct in this country, is said to seek shells 

 in France, but such an interesting example of geographical 

 modification requires confirmation. The Rev. F. D. Morice 

 thinks that there must be a mistake due to the confounding 

 of rufa with rufohirta, — Fabre's "VOsmie rousse." 



The fact that both Osmia and Anthidium include species 

 with the specialised instinct led me to ask Dr. Perkins whether 

 a common ancestry was the probable explanation. He kindly 

 replied as follows :— 



" Osmia and Anthidium, are not really closely allied : 

 though belonging to one family (Megachilidae) they represent 

 different sub-groups. Both genera are of huge extent and 

 evidently not really single genera. Osmia is particularly 

 composite, and the name as at present in general use rather 

 represents ' Papilio ' when this name comprised Pieris, 

 Colias, etc., etc., as well as Papilio. At least, that is my 

 opinion. In some respects Anthidium more nearly approaches 

 the leaf-cutters (Megachile). The similarity in choice of 

 nesting- place (shells) of several species of each genus evidently 

 does not indicate a structural affinity — it is not even a constant 

 habit in the case of some species. 



