Recenthj published Ornithological Works. 489 



The geographical limits assumed in the present catalogue 

 require a little explanation. They are described thus by Mr. 

 Ridgway : — 



•■'The geographical limits assigned to this catalogue 

 include the entire continent of North America down to 

 the southern border of the United States^ besides Green- 

 land, the peninsula of Lower California and the out- 

 lying islands of Guadaloupe and Socorro. Guadaloupe and 

 Socorro, like Lower California, are included for the reason 

 that their zoological relationships are much closer to North 

 America, as usually (but arbitrarily) restricted, than to the 

 tropical coast region of Western Mexico, their Avian fauna 

 in particular being decidedly of Nearctic affinity, with the 

 exception, so far as known, of only two species ; a Polyborus 

 peculiar to Guadaloupe, and a Conurus, found both in Socorro 

 and in Western Mexico. Indeed, the greater part of Mexico 

 itself (all, in fact, except the narrow coast region, or Tierra 

 caliente, and the lowlands of the southern portion) belongs, 

 ornithologically as well as geographically, to North America, 

 as might easily be demonstrated did space permit; but the 

 enlargement of our field to its proper limits would be quite 

 impracticable at the present time. For the surrender of this 

 our rightful territory, however, we have compensation in the 

 fact that the arbitrary line which we have drawn (i. e. the 

 United States and Mexican boundary from the Gulf of 

 Mexico to the mouth of the Colorado) gives a comparative 

 stability to the list which a greater southward extension of 

 the area, with indefinite limits, would render impossible.-'-' 



As regards the vexed question of trinomials, which, as 

 already stated, are adopted in 160 cases in the present cata- 

 logue, the author candidly acknowledges that the use of them 

 has caused " perhaps the greatest difficulty encountered in its 

 compilation,^^ it being " in many cases very difficult to decide 

 whether a given form should be treated as having passed the 

 varietal stage and therefore to be designated by a binomial, 

 or whether it is as yet incompletely dififerentiated and to be 

 subordinated in rank by a trinomial appellation.^^ This dif- 

 ficulty, however, as we are informed in a footnote, has ariseii 



