202 Corrcs-pondence. [April 



slight to be detected, but which are either too slight or too inconstant to 

 require recognition. While theoretically it is possible to recognize 

 'varieties of varieties,' in practice this rarely occurs, and should never be 

 countenanced; if a form is different enough to be recognized, it should 

 stand as a variety of the common stock, not as a variety of a variety, 

 although it may be more nearly related to some one of several varieties than 

 to any of the others.* Again, the objection has been raised that the rec- 

 ognition of varieties is subject to the caprice of any dabbler vs^ho may feel 

 disposed to set them up : theoretically this also is true, but in practice such 

 work falls where it should — to experts, who occasionally err in judgment, 

 or through inadequate material, but in the main are safe guides, and as 

 such are followed, even by their peers when these have not them- 

 selves the same or a better opportunity to review the group in question. 

 The recognition of a variety is a matter to be as carefully and conscien- 

 tiously considered as the recognition of a species, or any higher group. 



Hoping that our remarks may serve to throw a little further light upon 

 the points at issue, we again take leave of the subject. — ^J. A. A.] 



The Ornithological Report in the 'Cruise of the Corwin.' 



To THE Editors of The Auk: — 



Sirs: I observe that in his notice of my ornithological paper in the 

 'Arctic Cruise of the Revenue Steamer Corwin,' Dr. Coues indulges in 

 some severe strictures on the typographical errors and mechanical execu- 

 tion of the report. 



It must be conceded that the number of these errors and their atrocity 

 renders his critical remarks justifiable enough. Had, however. Dr. Coues 



* In this connection it see'ms not out of place to refer briefly to a point raised by Dr. 

 Stejneger in his article in this number of 'The Auk' on the genus Acanthis. He 

 alludes fp. 150) to Mr. Seebohm's practice of forming trinomials of the names of the 

 conspecies most nearly related, as tending to better express their true affinities than 

 does the method, adopted by American writers, of taking for the second term of the 

 trinomial the name first given to the group of conspecies as a whole or to any of its 

 forms ; and adds : "This .... is a point which merits earnest consideration." We 

 believe, however, that there are two unquestionably strong objections to Mr. Seebohm's 

 method of constructing trinomials. First, it leaves the construction of conspecific 

 names subject to individual opinion as to what two forms of a given group of inter- 

 grading forms are most nearly related — a point about which there must, in the nature 

 of the case, be often a diversity of opinion. Second, and of far greater importance, it 

 ignores the law of priority — the fundamental principle of our nomenclature — and 

 therefore opens the way to instability of names and endless confusion. It seems to us 

 perfecfly evident that the law of priority should be considered as equally imperative in 

 relation to conspecific — or subspecific — names as to specific and generic names. In 

 other words, the name first applied to any form of a group of conspecies should be 

 the designation, in a specific sense, for the group as a whole, and should also form the 

 second term of the trinomial for each of its conspecies, whatever may be their relation- 

 ship ««/«' Je/ and that the slight gain accruing in special cases by Mr. Seebohm's 

 method is much more than offset by the ill results that must inevitably follow from 

 disregarding the law of priority in constructing conspecific names. 



