THE HAUNTS OF THE CANON WKEN 165 



eyes brown. Length about 5^ inches; extent, 7|; wing, 2,-; tail, 2^; bill, 

 I ; tarsus, J. 



The numerous United States specimens of this bird I have examined diifer 

 decidedly from the Mexican form, as accurately pointed ont by Mr. Ridgway. 

 The Mexican bird is larger, with a different curve of the bill ; it is much 

 darker colored both above and below, with sharper distinction of the white 

 throat, and with the spots of the upper parts restricted to the back and 

 wings; with the black tail-bars much broader and more regular, and the 

 light markings on the outer webs of the quills mere indentations instead of 

 complete bars. 



POINTS, about the Canon Wren are its fondness for the 

 resorts the name indicates, and its wonderfully impressive 

 chant. More anon of the last of these two leading traits. I 

 will first speak of its haunts, which are no less characteristic 

 of the bird than its singular utterances. It is not very long 

 since the bird was unknown as an inhabitant of the United 

 States ; and no one could have surmised how large an area in 

 this country it really occupies from the hints of its distribution 

 which our literature of ten years ago afforded. It was supposed 

 to merely reach our border, with a little extension within our 

 limits up the Colorado Valley. The fact that I had never seen 

 it at Fort Whipple supported this notion of its limited distribu- 

 tion, and in my " Prodrome V of 1866 1 gave the bird as one gen- 

 erally distributed over the southern and western portions of 

 Arizona, up to S'ort Mojave at least. I now see that its absence 

 from that locality — at any rate, its rarity, so great that it never 

 came under my observation — was due to the topographical 

 features of the place, not its geographical position. There 

 were plenty of rocks about the fort (rocks, like reptiles and 

 cactuses, are nattwal products of Arizona), just suiting the 

 wants of the Salpinctes; but this immediate vicinity lacked 

 the singular walled chasms with which many portions of the 

 Territory are scored and seamed — those reproductions on a 

 smaller scale of the Grand Carton of the Colorado itself, most 

 wonderful crack of the ground in America — and such rifts of 

 solid rock alone are entirely to the liking of the Canon Wren. 

 So it fell out that it was left for the latest ornithologists of 

 the Southwest — for Allen, Aiken, Ridgway, and Henshaw — to 

 show that the range of the bird extends from Arizona and 

 New Mexico, and portions of Texas and Southern California, 

 into Colorado, Utah, and Nevada. How much further it may 

 actually reach we do not yet know ; but there is nothing in 

 the analogies of the case to forbid the supposition that the 



