1882.] ANATOMY OF THE TODIES. 449 



special resemblance ; but, as already pointed out, though the arrange- 

 ment is similar, it is not identical, whilst, on the other hand, that 

 characterizing the Todies could easily be produced by a slight 

 modification of that found iu some of the other groups of Anomalo- 

 gonatse. 



In the face, then, of the many important differences that exist in 

 nil parts of the structure of the two forms, and iu the absence of 

 any special features common to them, I cannot agree to the proposi- 

 tion that the Todies are more closely related to the Motmots than 

 to any other group. 



In the possession of cseca and in the conformation of their 

 pectoral tract the Todies agree with all Garrod's " Passeriformes," 

 with one of the famihes of which indeed, the Galbulida, one of the 

 most acute ornithologists that has ever lived, the late Mr. Blyth, 

 associated them as a special group, " Angulirostres"'. On the other 

 hand, in possessing a well-developed tuft to the oil-gland, the Todies 

 differ altogether from the Passeriform series of Anomalogonatae. 

 Detailed comparison of the structure of the Todies with that of the 

 other families of this great group is unnecessary, none of them 

 possessing features indicating such affinities to the former as to render 

 probable any particular genetic connexion of the two. 



As Dr. Murie has already remarked, "Todus is inconsistent in several 

 respects,"^ a truth made more obvious by the facts above recorded. 

 In the possession of cceca combined with the tuft to the oil-gland, 

 7hdus presents an exception to Grarrod's definition of his group Ano- 

 malogonatae ^ though it agrees with all of them in the absence of both 

 the ambiens and accessory femoro-caudal muscles. Nevertheless it is 

 certain, from its characters generally, thaiTodits is an Anomalogouatous 

 bird, though its isolation from any other of the families of that group 

 seems to me to preclude its insertion in the Picif'ormes, Passeri- 

 formes, or Cypseliformes of Garrod-*. It is impossible, I think, to 

 say that Todus is more clearly related to any of the Piciformes than 

 it is to the Passeriformes ; and to include it' the definitions of either 

 of those groups would have to be altered. I propose, therefore, to 

 create a group of equivalent value to those just named, which may 

 be called " Todiformes," and of which Todus is the sole living 

 representative. 



Next, as to the meaning of these facts. I think few ornithologists 

 who have carefully considered the question can doubt that the 

 " Anomalogonatae " of Garrod are a natural group of birds ^ i. e. one 

 descended from a common ancestor. On this view this ancestor 

 must have possessed the sum of the characters — supposing, unless 

 there is reason for the contrary, that the latter have not been re- 

 developed, and excluding those that may reasonably be supposed to 



1 Charlesworth's Mag. Nat. Hist. ii. 1838, p. 361. 



2 P. Z. S. 1872, p. 678. 



3 P. Z. S. 1874, p. 118; Coll. Papers, p. 216. 

 * L. 0. p. 222. 



° It is neai-lj certain that the Cuculiclffi and Musophagidte, as also the Psit- 

 lacids, are in no way related to the other so-called Pieariai. 



