May, 1914 A CHANG]{ IN FAUNA 123 ject are on a comparatively limited area. It is an oasis in the desert, for desert conditions still exist on all sides. Since the opening of the region the number of bird species has risen from fifteen to one hundred and thirty. A few of the most interesting changes are tabulated below. Boulder, Colorado, February 25, 1914. TItE RACES OF BRANTA CANADENSIS Suggested by Swarth's "Study of a Collection of Geese"* By ALLAN BROOKS O GROUP of North American birds so badly needed revision as the geese of the canadensis type, and Mr. Swarth has put not only all ornitholo- gists, but also the discriminating sportsman, under a deep debt of grati- tude for his excellent and carefully studied paper, which is 'practically in the nature of a monograph of the group. American ornithologists have always been rather prone to carefully study all small birds, and let the larger species severely alone. No better instance of this is needed than the case of these geese. Sportsmen were unable to cor- rectly identify the geese they shot by the aid of any of the available works, nor could their ornithological friends help them much, as all the given diagnoses were at fault. Could anything be more impossible than the breeding range of Branta c. occidentalis as given in the latest A. O. U. Check-List? This sub-species is there quoted as breeding in northeastern California and on the coast of west- ern Alaska---localities some fifteen hundred miles apart--though it was known that the whole intervoning country was occupied by typical canadensis. Many good ornithologists had worked the northern California and Oregon regions, but it was not until the last year or two that it was definitely ascer- tained that all records of occidentalis as a breeding bird were wrong, and that canadensis and only canadensis was the resident species. The wonder is that any credence could be given to the theory of an isolated breeding colony of occgdental?s. Mr. Swarth has been unable to find any reliable evidence of this sub- species in California, even as a migrant. In his treatise he first of all proves the fallibility of all the distinctions as based on pattern and color, and indicates that the most reliable distinctions between canadens?s, hutchinsi, and m?n?ma lie in the measurements of the cul- men and in the comparative lengths of tarsus and middle toe. The very adequate and ingenious system of symbolical measurement charts given by Mr. Swarth are a distinct feature of his work. From these it would seem that most reliance can be placed on measurements of the culmen to separate the three subspecies, i.e., canadens?s, hutcMnsi, and minima. In his table the two former do not coalesce at all in this respect, and the slight overlapping of the last two might
- A Study of a Collection of Geese of the Bravura canadensls Group from the San Joaquin
Valley, California. By }{arty S. Swarth. Univ. Calif. Publ. Zool., vol. 12, 1913, pp. 1-24, 2 pls., 8 text figs.
