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



or damaged specimens and very aberrant varieties, in which 

 the ordinary characters are wanting or obliterated ; and 

 sometimes to reach a conclusion, satisfactory to myself at 

 least, as to the propriety or otherwise of uniting or distin- 

 guishing particular species. 



More than this specific value I do not claim for them. 

 I long cherished the hope that they might be utilised for 

 the establishment of good natural groups or subgenera — so 

 urgently needed in this interminable genus, with its ever 

 increasing "rudis indigestaquc moles' of specific descriptions. 



But I have had — most unwillingly — to lay that hope 

 aside. It does not seem to be at all the universal rule that 

 other correspondences between species should involve agree- 

 ment in the terminal characters. Nor do the latter always 

 agree among themselves ; i.e.^ associations suggested by the 

 type of the armature may be rendered less probable by 

 that of the ventral segments and vice versa. Further, 

 though the extreme forms are readily distinguished, so 

 many intermediate conditions occur, that even if it were 

 certain that the former indicated real subgeneric differences, 

 it would be extremely difficult to fix the limits of the sub- 

 genera ; and again, if every striking difference were to be 

 treated as subgeneric, the number of subgenera required 

 would make the system practically useless. Certainly 

 when two species show an extreme difference in all ter- 

 minal characters, it would require strong evidence to make 

 me believe that their i-elationship was of the closest kind. 

 But assuming, what seems to me almost axiomatic, that a 

 large group of species like Andrena embraces many minor 

 groups, each developed separately from an antecedent 

 species, I see no reason why a parallel differentiation of 

 terminal characters should not have been produced by 

 similar causes in several groups independently. 



But however this may be, the inquiry has acquainted me 

 with much that has been new and interesting to myself, 

 and may be so, I hope, to others. Should the publication 

 of these notes lead some better qualified hymenopterist to 

 examine the facts for himself, I have no doubt that he 

 will see in them much that has escaped my notice, and will 

 obtain more valuable results. 



The characters that I have examined lie, as has been 

 said, in the $ genital armature and the last two ventral 

 segments, the 7th and 8th (not reckoning the " transferred 

 segment "). Those of the 7th, however, do not seem likely 



