1 14 MR. A. H. GARROD ON THE MUSCLES OF BIRDS. [Feb. 3, 



two latter families with the two former, or with one another. In 

 fact the Accipitres, as generally defined, are not a natural group at 

 all ; and the Cathartidce are not the least more nearly related to the 

 VulturidcB than to the Falconidce. 



Respecting families it may therefore be said that myological pecu- 

 liarities which do not involve more than a single structural change 

 from the typical arrangement of the family are frequent, and that 

 further differences indicate a more distinct relationship. 



The various opinions held by different ornithologists as to the 

 correct division of the Orders of the Class Aves are so numerous 

 that they make it impossible in this stage of the inquiry to discuss 

 the myological features which they present. An inspection of 

 Plate XVII. is sufficient to show that the formula of a bird is not 

 of direct value in estimating ordinal characters. 



Looking at the whole subject from another point of view, it may now 

 be asked, What does the arrangement in the muscles above described 

 teach as to the major divisions of the Class Aves 1 The remainder 

 of this communication will be an attempt to answer this question. 



A mere glance at Plate XVII. is sufficient to show that the facts dis- 

 closed by a study of the myology of birds do not, without extraneous 

 assistance, place the different families in their true relationship to 

 one another. Because the same muscles are present in two families 

 of birds, it cannot therefore be said that their kinship is extremely 

 close, or the reverse ; if such were the case we should have to put 

 the Ardeida with the Passeres, and to separate the Auks from the 

 Gulls, both of which results would be strongly in opposition to the 

 teaching of osteology. It is therefore necessary to look around to 

 find, if possible, myological characters which have some definite rela- 

 tions to equally well-marked pterylographic, visceral, or osteological 

 peculiarities. 



Before going further it will be necessary to clearly understand a 

 principle which is of much assistance in working out the details of 

 classification from a large number of unarranged facts. It is this : 

 when any certain structure is found to exist in an unmodified form 

 in several clearly separable members of any well-marked larger divi- 

 sion of the Animal Kingdom, that structure must be considered 

 typical of the division ; in other words, that structure, or the poten- 

 tiality for producing it, must have existed in the common ancestor 

 of the division under consideration ; and those of its members who 

 are wanting in the particular structure are so because they have lost 

 it in process of time, not because the others have separately acquired 

 it ; for the probability, if it were only a matter of probability, is very 

 little that several distinct and different species should separately 

 acquire a single identical structure ; whilst it is infinitely more likely 

 that several distinct species should all lose a common character. 

 That all Mammalia should acquire branched horns is improbable ; 

 but that many which possess branched horns should have them 

 broken off whilst rushing through a wood, whatever species they 

 belong to, is much more to be expected. 



Employing this argument with regard to the facts under discussion, 



