416 Mr. H. Seebolim on the 



suborders may or may not be natural groups, expressing tlie 

 mutual relationship of the birds contained in them ; but we 

 have every reason to believe that they are approximately so. 

 There can be little doubt that the osteological characters 

 of birds are a much safer guide to their natural affinities 

 than the external characters •which are to be discovered l)y 

 an examination of the bill and the feet, or the M'ings and the 

 tail, and which consequently are founded more upon modi- 

 fications of skin than modifications of bone. Other depart- 

 ments of anatomy may be more important than osteology, but 

 next to the skin and feathers the skeleton is the easiest to 

 obtain; so much so, indeed, that of a great many species 

 of birds the osteology is the only part of the anatomy of 

 which anything is known. 



Articulation of the Dorsal Vertebrae. 

 It has recently been shown (Parker, Proc. Roy. Soc. xliii. 

 p. 470) that other birds besides the Penguins and Auks are 

 opisthoccelous (or post-concave) in the articulation of their 

 dorsal vertebra3. We are assured that the Limicolse, Gavise, 

 Phalacrocorax, Plotus, Steatornis, and the Psittacidie are 

 opisthoccelous, and that it is the more common kind of 

 articulation in Archaic reptiles. It would, however, be a 

 mistake to assume that the Impennes and the Gavio-Limicolse, 

 because they have opisthoccelous dorsal vertebrae, are more 

 nearly related to each other than the latter are to some of 

 the groups which have heterocoelous (or saddle-shaped) dorsal 

 vertebrae. The fact is, that every intermediate form between 

 one and the other occurs in the Gavio-Limicolse — so much so, 

 indeed, that the dorsal vertebrae of Numenius arquata, for 

 example, more nearly resemble in their articulation the 

 dorsal vertebrae of the Gallinae than those of the Impennes. 

 The last free vertebra of Sphc7iiscus demersus is anteriorly 

 convex and posteriorly concave, both in its lateral and verti- 

 cal sections. That of Argus giganteus is the same in its 

 vertical section, but exactly the opposite in its lateral section. 

 That of Numenius arquata agrees with both in having the 

 posterior outline of the vertical section concave, is interme- 



