1070 BR. p. CHALMERS MITCHELL ON THE 



of the different groups with regard to the central avian condition. 

 It is to be noted that nearly every group seems to have made 

 experiments in both directions, but as we i-ead from above down- 

 wards in the diagram, we pass from movement towai-ds the left to 

 movement towards the right. And it may also be noticed that, 

 in a general way, what are usually regarded as the groups ex- 

 hibiting the lower modifications of avian structure are towards 

 the top of the diagram, those representing the higher types 

 towards the bottom. If I had felt justified in expanding this 

 diagram, by placing the names of the minor groups in their 

 proper orientation with regard to the peroneals, it would have 

 been found in the same fashion, that inside each group, on the 

 whole, those which are generally regarded as the higher types 

 were towards the right hand side of the spaces, those representing 

 the lower types towards the left hand. In other words, I think 

 I may say that the higher types of avian modification are associated 

 with a tendency for the degeneration of the Peroneus longus and 

 an increase of the Peroneus brevis. 



Adaptation, direct or associated, may be the fundamental 

 explanation of the facts that I have ti'ied to set out, but it is 

 difficult to follow, and it is easy to see that kinship appears to be 

 a more important factor. The Eagles and Vultui'cs have many 

 adaptive resemblances with the Owls, but the former contain 

 members directly linking them with the primitive condition and 

 have never moved far from it, and the latter show the extreme 

 modification of the Coraciiform group. Swifts and Swallows have 

 many points in common, but the former, in the condition of the 

 peroneals, are extreme Coraciiform birds, the latter very slightly 

 modified from the true Passerine condition. So also Humming- 

 birds are extreme Coraciiformes in this respect, and Sun-birds 

 are true Passerines. The family tradition appears even in many 

 of the minor changes; all the Storks have lost the brevis and 

 the Hei'ons have retained it. 



There remains to say a word as to the few birds which do not 

 lie comfortably, so far as the peroneals are concerned, in the 

 positions usually assigned them. In a memoir dealing with a 

 similar large series of facts in bird anatomy (" On the Intestinal 

 Tract of Birds," Trans. Linn. Soc, Zool. viii. p. 173), I pointed out 

 what seems a logical necessity (frequently, however, overlooked 

 by those who use anatomical chaxacters for systematic purposes), 

 that if we have reason to believe a particular character to have 

 been ancestral, we ca.nnot assume that animals now without it, 

 are more nearly allied than those that have retained it. There 

 is no a j)riori ground for assuming that it may not have been lost 

 twice or several times independently. It follows therefore that 

 the loss of the longus muscle, or of any important part of it, or 

 of the brevis muscle, is no valid clue to systematic position. 



It is equally clear that the common retention of the ancestral 

 condition is no gi-ound for placing the descendants of a particular 

 aiicesti'al stock together, if relative affinit}^ and not convenience 

 is to be the basis of classification. 



