Individuals agreeing in the length of the bill, however, naturally exhibit the maximum difference in the length of the wing, while those agreeing in the wing can be distinguished by the length of the bill. In other words the ratio of bill to wing length in the two species _S. obsoletus_ and _S. guadeloupensis_ is constantly different and furnishes a diagnostic character by which the species may always be distinguished. In the former the wing is more than three and a half times the length of the bill, in the latter less than three and a half. In addition there is a well-marked difference in color.
It seems reasonable to consider such differentiation in proportions when developed to the point where there is constant difference in ratio as of specific value. Measurements appear to indicate that this point has been reached in the Rock Wrens, and that the dark, long-billed forms should therefore be regarded as specifically distinct from the paler, shorter billed races. The same conclusion was arrived at by Swarth in 1914 (Condor, XVI, p. 216).
The Guadalupe Junco (_Junco insularis_) easily fulfills the above requirements of a species. Indeed as it averages 10 mm. less in length of wing than its nearest relative _J. townsendi_, and its bill is nearly 2 mm. longer, there is small likelihood even of ordinary intergradation. There are also well-defined color characters.
In Dr. Dwight’s recent paper on the Juncos (Bull. Amer. Mus. Nat. Hist., XXXVIII, 1918, p. 269) he has reduced this Junco as well as _Junco townsendi_ to subspecies, on the grounds that their characters are _quantitative_ rather than _qualitative_. But _are_ their peculiarities merely _quantitative_, and do not the differences exhibited by these forms more nearly approach the characters commonly regarded as of generic value than do the “_qualitative_” color differences between the forms regarded by Dr. Dwight as species?—W. DEW. MILLER, _American Museum of Natural History, New York City_.
=“Off” Flavors of Wildfowl.=—Following is an extract from a letter on this subject by Dr. L. C. Jones of Falmouth, Mass., who has been quoted in a previous article[34] on this subject. It will be noted that one of Dr. Jones’ theories is much the same as that advanced by the writer in the last sentence of his first communication on fishy flavor.[35]