364 BULLETIN 15 3, UNITED STATES NATIONAL MUSEUM 



The races may be identified as follows: The blue superciliary- 

 stripes are extended forward and form a continuous stripe along 

 the forehead only in cyanostictus and sharpei. In the former the 

 band is lighter blue and does not extend as far back on the dorsal 

 side of the black auriculars as in the latter. In sharpei the band is 

 darker, almost purplish or violet blue in some individual^, and ex- 

 tends posteriorly at least to the end of the black auriculars. Caution 

 should be exercised in using the last character as the " make " of the 

 skin may distort the results. Of the other three forms, pusiUus alone 

 completely lacks the bluish superciliaries, and ocularis differs from 

 meridionalis in having narrower black bands on the rectrices, these 

 marks being about 10-12 millimeters wide in the former, and 14—16 

 millimeters wide in the latter race. 



The size variations of sharpei are as follows : 



Males: Wing, 74-82 (77.7); tail, 62-67 (63.6); culmen, 22.5-29 

 (26.1 millimeters). Females: Wing, 76-80 (78); tail, 62.5-70 

 (66); culmen, 24-28 (25.9 millimeters). The size differences be- 

 tween races is slight and onl}^ an average one at that, tlie over- 

 lapping being practically as extensive as the limits of variation. For 

 example, the wing length in males of cyanostictus varies from 76 to 85 

 with an average of 79.9 millimeters. The other measurements are 

 similarly only slightly different from those of sharpei. In ocularis 

 the wing length of adult males ranges from 76.5 to 79.5 with an aver- 

 age of 77.3 millimeters. 



The immature plumage of sharpei is quite like that of cyanostic- 

 tus, but the superciliaries and forehead are darker, more violaceous 

 blue in the former. The adults of both these races vary in the color of 

 the middle rectrices and also of the edges of the secondaries. Six birds 

 of the 20 listed above have these parts distinctly bluish green, the 

 rest have them greenish without any trace of blue. Granvik ^' first 

 noted this for cyanostictus and observed that, " * * * this * * * 

 blue tint is a color appearing at about the time of — or somewhat 

 earlier than — the moult." This is borne out by the series examined 

 by me. The bluish color appears to be due to wear and fading, but 

 not entirely to these factors, as some fresh plumaged " green " birds 

 have bluish edges to some of the inner secondaries and a very little 

 bluish tinge on the margins of the middle tail feathers. Of course, 

 it may be that the edges being more exposed than the rest of the 

 feather, the blue would show there first. Tlie " blue," worn-plumaged 

 birds were all taken in March, April, and May. 



According to Erlanger -° the breeding season is in March and April, 

 which observation is in keeping with the worn ])lumage of the speci- 

 mens taken during that period. On April 9 at Gato River, Mearns 



"Journ. f. Ornitli., Sonderheft, 1923, pp. 108-109. 

 =»Idem, 1905, p. 455. 



