REVISION OF THE CLAPPER RAILS — OBERHOLSER 315 



Neither one of the external characters of plumage above mentioned, 

 nor any difference in size or proportions, is entirely trenchant when 

 all the races of Rallus longirostris are included. However, even though 

 individually the two species may apparently intergrade in their char- 

 acters, they sometimes breed in the same areas, as for instance in 

 Texas, Louisiana, Cuba, and North Carolina, where their ranges, at 

 the edges at least, overlap; and it is interesting to note that in these 

 particular places the characters of the two birds are such as to render 

 them easily distinguishable. For the most part Rallus elegans is a 

 fresh-water bird, and Rallus longirostris an inhabitant of the salt 

 marshes, but Rallus elegans sometimes breeds in salt or brackish areas 

 where also there are clapper rails ; and some of the subspecies of clapper 

 rails, such as Rallus longirostris yumanensis and Rallus longirostris 

 tenuirostris are strictly fresh-water birds. Furthermore, there is in 

 Cuba a subspecies of Rallus elegans, known as Rallus elegans ramsdeni, 

 although it is separated geographically from the typical Rallus elegans 

 elegans by two or three forms that are certainly subspecies of Rallus: 

 longirostris, these being Rallus longirostris insularum, Rallus longi- 

 rostris waynei, and Rallus longirostris scottii. The two groups can b© 

 separated more or less satisfactorily, however, by the use of a combina- 

 tion of characters, in that Rallus elegans presents the combination of 

 a very reddish bend of the wing with a decidedly ochraceous or rufes- 

 cent tinge on the upper parts, which combination is not found in any 

 form of clapper rail even though each one of these characters may not 

 be separately trenchant. Careful studj^ and comparison are necessary 

 sometimes fully to appreciate these differences, but they are present 

 and are characteristic. In view of the facts above presented it is 

 apparent that we have here what might be regarded as a biological 

 species, and one to which the criterion of intergradation as a test of 

 subspecific relationship is inapplicable. There is a similar case in 

 Europe in the gulls called Larusfuscus and Larus argentatus; and others, 

 in America, as for instance in some species of flycatchers of the genus 

 Empidonax. Under these circumstances it seems best, at least for the 

 present, to consider these birds as representing two species, Rallus 

 elegans consisting of two subspecies, and Rallus longirostris, made up 

 altogether of 25 races. It is with only the latter that we have to do 

 in the present connection. 



In a group as variable racially and individually as the clapper rails, 

 examination and study of the whole group reveal the significance of 

 differences in characters, which sporadic investigations wholly fail 

 to do, and this light enables one to predicate distinctions and separa- 

 tions of subspecies much more safely than would otherwise be the 

 case. The group shows a great tendency to break up into local races, 

 many of which have very limited distribution, which accounts in con- 

 siderable measure for the recognition of so many forms. The most 



