TARSAL ENVELOPE OF CAMPYLORHYNCHUS, ETC. 155 



the other two western genera, Catherpes and Salpinctes, in com- 

 parison with the more typical Troglodytes. In the latter, the 

 tail is tliin, that is, the individual feathers are narrow ; in the 

 western forms, these feathers are broad and rounded, and the 

 tail as a whole is consequently fan-shaped. As already stated, 

 the species are of great size for this family, and their habits are 

 in some respects peculiar. 



Impressed with certain differences observable between typical Wrens and 

 the three western genera, Campylorkynchus, Salpinctes, and Catherpes, gen- 

 erally assigned to the Troglodytidw, I have been led to look into the techni- 

 cal aspects of the case, with the result of becoming dissatisfied with the 

 alleged position of these forms among the Wrens. In establishing the genus 

 Catherpes as distinct from Salpinctes, Professor Baird noted certain discrepan- 

 cies in the structure of the feet ; and in 1864 (Review, p. 109), he enlarges 

 upon the remarkable structure of the tarsus of Salpinctes, which he charac- 

 terizes as " especially peculiar among all its cognate genera by having the 

 usual two continuous plates along the posterior half of the inner and outer 

 faces of the tarsus divided transversely into seven or more smaller plates, 

 with a naked interval between them and the anterior scutellce ". This is 

 certainly a remarkable feature for a presumed thoroughly Oscine bird to 

 exhibit, since it is highly characteristic of Oscines to have the postero -lateral 

 tarsal plates continuous, meeting in a sharp ridge behind. I verify the state 

 of the case in Salpinctes as given by Professor B iird, but I find, to my sur- 

 prise, that in Campylorhynchus the lateral plates, but especially the outer one, 

 are broken up into a series of conspicuous scutella ; and that Catherpes shows 

 a tendency, not so fully expressed,. to similar division of the tarsal envelope. 

 If this structure really possesses the significance attributed to it by many 

 of the best writers, the question whether these birds are Wrens at all is 

 re-opened. That they possess decidedly Wren-like habits is no strong argu- 

 ment, for nothing is more fallacious than such teleological bending of 

 diverse structures to similar ends. It will be remembered that Lafresnaye, 

 and other writers of repute, have placed species of Campylorhynchits in the 

 genus Picolaptes, which is a member of the large family Dendrocolaptidcv ; 

 some of these birds have rigid acuminate Certhia-like tail-feathers, and 

 Creeper-like habits; in others, however, the tail is soft, and among them is 

 witnessed the greatest diversity of habits. On comparing our Campylorhyn- 

 clius with a typical Dendrocolaptine (Dendrornis erythropygia), I find that the 

 bills of the two are extremely similar, and that the tarsal envelope of Den- 

 drornis is broken up posteriorly into a number of plates, of which those on 

 the inner aspect are continuous with those in front, while the postero- 

 exterior ones are a series of rounded and isolated scales. Again, in the case 

 of Salpinctes, it will be recollected that Bonaparte placed it in the genus 

 Myiothera, and considered it an Ant-thrush (Formicarlidce). On examining 

 the tarsus of a species of Thamnophilus, a typical Formicarian, I find that the 

 plates are divided behind, and the general structure is substantially the 

 same as in Salpinctes. The case of Catherpes is less clear, but it would proba- 

 bly go with Salpinctes. These points may not suffice for the summary dis- 

 missal of the genera under consideration from the Troglodytidce, but they go 

 to show that their position in that family is not assured. 



