214 Annals of the Carnegie Museum. 



Glaucidium brasiUanum [subsp.] Ridgway, Bull. U. S. Nat. Mus., No. 50, VI, 

 1914, 801, footnote (Bonda; meas.). 



Glaucidium brasilianum medianum Todd, Proc. Biol. Soc. Washington, XXIX, 

 1916, 98 (Bonda; orig. descr. ; type in coll. Carnegie Mus.). — Cory, Field 

 Mus. Z06I. Series, XIII, 1918, 46 (ref. orig. descr.; range). 



Glaucidium brasilianum phalmnoides (not Strix phalcenoides Daudin) Chap- 

 man, Bull. Am. Mus. Nat. Hist, XXXVI, 1917, 255 (Bonda; crit.). 



Seventeen specimens : Bonda, La Tigrera, Santa Marta, Mamatoco, 

 Fundacion, and Gaira. 



The series shows a certain amount of variation, paralleling that ex- 

 hibited by the Central American and Mexican form. No. 8,956, al- 

 though in the brown phase, has the tail decidedly tinged with rufous. 

 No. 42,630, also in the brown phase, has the pileum plain, except in 

 front. No. 41,964 is as gray above as the average Central American 

 bird, and differs further in having the pileum and back conspicuously 

 spotted with white. No. 41,847, May i, is a young bird in the rufescent 

 phase, with the tail only about half grown, but coming in wholly ru- 

 fous, unbarred; the striping on the under surface is less distinct. 



Sharp'e referred the four Santa Marta specimens of this owl which 

 came under his notice to the Brazilian form, called by him G. ferox, 

 but this would seem to be a mistake, judging from the material ex- 

 amined in this connection. G. brasilianum brasilianum, as this south- 

 ern form must be called, is characterized by its relatively larger size, 

 richer, browner coloration above, more decidedly punctate pileum, and 

 heavier, more " solid " streaking below in the brown, phase. The last- 

 named character may also serve to distinguish it in the rufescent phase 

 also, but with only one specimen in this phase one cannot be sure. 

 Comparison of the present series from the Santa Marta region with 

 another from Trinidad and northern Venezuela on the one hand, and 

 with one from Central America, Mexico, and Texas on the other, 

 shows that they represent a form distinct from either, as already sur- 

 mised by Mr. Ridgway. In the brown phase the new form differs from 

 G. brasilianum phalcenoides by its paler, more buffy brown color above, 

 less heavily streaked under parts, and more decidedly and more regu- 

 larly barred tail, the bars usually running across the feather, or at 

 least touching the shafts, while in phalcenoides they seldom are more 

 than a' pair of spots on the outer webs. In the rufescent phase there 

 is apparently no difference in color, but the streaking of the under 

 ■ parts averages less broad in medianum than in phalcenoides. From G. 



