NATURAL SCIENCES OP PHILADELPHIA. 79 



ensis from true stolidus by the nearly white belly, and find it 

 absolutely inseparable from the Cuban sagrse. 



I cannot find where (if anywhere) Dr. Bryant has characterized 

 the Bahaman bird as " bahamensis ;" but on p. 90, in text under 

 " dominicensis" he says that the latter " differs from sagrae and 

 bahamensis in the distinct yellow of the abdomen," etc., showing 

 that his "bahamensis" is a white-bellied bird, and being from the 

 Bahamas, it must be the same as lucaysiensis. 



C. VAR. antillarum. 

 ? Myiarclms sp., Taylor, Ibis, 1864. Porto Rico. 

 Tyrannus {Myiarclms) antillarum, Bry., P. B. S. N. H. 1866, p. 2. 



Porto Rico. 

 Myiarclms antillarum, Sund., Of. Vet. Ak. Fork. 1869, 599. 

 Tyrannus (Tyrannus) antillarum, Gray, H. L. No. 5544. 



M. olivaceo-fuscus, pileo sensim obscuriore, gula et pectore 

 cinereo-albidis, ventre albo vix aut non flavo-tineto, remigibus 

 primariis minime rufo-marginatis, rectricibus ornnino immarginatis, 

 sed crebrissime macula rufa in apice pogonii interioris notatis. 



Hab. Porto Bico {Bryant, Sivift, Latimer, Mus. S. I.). An 

 Tobago (Jardine)? 



Obs. The Porto Bican form is almost a species. Local differen- 

 tiation is here at an extreme, the better marked examples looking 

 very little like the Jamaican stolidus, and not particularly resem- 

 bling even the whitish-bellied Cuban phcebe. In extreme cases the 

 tail feathers have no rufous edging at all, and the belly is pure 

 white. But we have already seen, in the Cuban and Bahaman 

 bird, th?t the belly fades away from the 3-ellow that is found in 

 the Jamaican, through every shade, till it is sometimes white ; 

 and we have likewise observed the reduction of the rufous to a 

 mere edging of the rectrices ; thence into antillarum is but a 

 step. Some specimens of antillarum have the inner webs mar- 

 gined with rufous part way down ; and the difference in this 

 respect between these and some Cuban phcebe is not so great as 

 may be found among different individuals of either of the other 

 varieties. These examples of antillarum also show the most 

 yellowish on the belly, this often exceeding the amount exhibited 

 by Cuban or Bahaman skins. In the purest antillarum, however, 

 the rufous is usually restricted to a mere trace at the end of the 

 inner webs, and it maj r disappear altogether. 



I could easily and plausibly describe antillarum as a species, but 

 1872.] 



