( xlvii ) 



Descriptions and Types. 

 I once at a Meeting of this Society, in perhaps an incautious 

 moment, expressed a wish that as soon as a man had described 

 a species his type shovild be destroyed, so that no record of 

 the species should exist except his description. If his descrip- 

 tion were good and adequate his species would stand, but if his 

 description were bad and inadequate it would fall and hold no 

 rank of priority, unless it might be adequately described at 

 some subsequent date before anybody else had given it a 

 different name, and that even then its date to secure priority 

 would rank from the adequate description only. I may say 

 that I began to act on this principle in a " List of British 

 Diptera " which I publisjied in 1888, as in that List I expur- 

 gated 195 species already described by British authors from 

 specimens of British species, and were I to publish a second 

 edition I should expurgate at least 125 more of Walker's 

 so-called species, thus making a clean sweep of about 320 

 names which were simply burdening our lists and proving 

 a stumbling-block to those who wished to study certain 

 groups. These names are of no scientific value, and fortu- 

 nately in most cases the original type specimens have been 

 either destroyed by mites or mould, so that further identi- 

 fication is impossible. I intend in any future work which 

 I may be able to carry out, to continue this process in 

 Dipterology, and to remove all names of what I cannot 

 even dignify with the epithet of insufiicient descriptions, 

 but which I should call rubbish descriptions. I must 

 emphatically consider that such " rubbish " descriptions 

 possess nothing but waste-paper value, or are less than the 

 equivalent of "Catalogue names," even though the original 

 type specimen may be eventually identified and in the end 

 adequately described by some subsequent stvident. This leads 

 me to what I believe is considered another most heretical view, 

 and that is that in my opinion an author should in many 

 cases not identify his original type specimen. I know that 

 when I have ten or twenty or fifty specimens of a probable 

 new species before me I describe that species from the lot, and 

 not from one individual specimen, and I most distinctly object 



