314 PROCEEDINGS OF THE NATIONAL MUSEUM vol.84 



the Museum of Comparative Zoology, Cambridge, Mass.; the Car- 

 negie Museum, Pittsburgh, Pa. ; the Field Museum of Natural History, 

 Chicago, 111.; and the private collections of Dr. Jonathan Dwight, 

 Dr. Louis B. Bishop, Dr. Witmer Stone, J. E. Thayer, and J. H. 

 Fleming. To the authorities of these museums and to the mdividuals 

 mentioned the writer is much obligated for the use of material. To 

 Dr. Charles W. Richmond, Charles B. Cory, Outram Bangs, Dr. 

 Alexander Wetmore, W. E. Clyde Todd, and particularly to Dr. 

 Herbert Friedmann, the writer is further indebted for other courtesies. 



In North, Middle, and South America there are approximately 27 

 forms of large rails belonging to the genus Rallus Linnaeus. These 

 may, in general, be considered to belong to two groups, the king rails 

 (Rallus elegans) and the clapper rails (Rallus longirostris) . They 

 have, however, variously been treated as representing from two to six 

 distinct species, some of these with several subspecies. 



The first problem of this investigation was presented in the necessity 

 for determining the number of species involved. Examination shows 

 that Rallus longirostris and the other South American forms inter- 

 grade completely with Rallus crepitans of the Eastern United States, 

 through the forms inhabiting the West Indies and the Eastern 

 United States, when individual variation of the island forms is taken 

 into consideration. Therefore, the latter must be treated as a sub- 

 species of Rallus longirostris, along with all the forms inhabiting the 

 West Indies. The same is obviously true of the additional races in 

 South America. Furthermore, when all the races represented by a 

 sufficient series are compared with one another it becomes increasingly 

 evident that none of the forms of the Pacific coast can be trenchantly 

 separated, including Rallus obsoletus, Rallus levipes, Rallus beldingi 

 of Lower California, and the other recently described subspecies from 

 northwestern Mexico and southeastern California. Even the isolated 

 Rallus tenuirostris of the Valley of Mexico presents no characters that 

 are not bridged over b.y individual variation when all the forms are 

 considered. There thus seems to be no alternative to regarding all 

 these as races of a single species. It might be mentioned also that 

 none of these overlap in their breeding distribution. 



This accomphshed, it remains yet to determine the status of the 

 king rail, Rallus elegans, of the Eastern United States, and its single 

 subspecies, Rallus elegans ramsdeni, of Cuba. This is an unusually 

 difficult matter to decide, and one concerning which there may well 

 be differences of opinion. The chief external characters separating 

 the Idng rails from the clapper rails consist in the much more reddish 

 bend of the wing, and in the rich rufescent-ohve tinge of the upper 

 parts of the former bu'ds, this mvolving both the centers and margins 

 of the feathers. There is little or no trenchant difference m behavior, 

 voice, nest building, or other habits betw^een these two species. 



