ORIGIN OF FLIGHT IN BIRDS. 471 



specialisation. Whenever in nearly-related Mammalia terrestrial 

 and arboreal forms occur, it is in the latter that the clavicles 

 are certainly present. According to Weber (19), the clavicles are 

 missing in the terrestrial genus Hystrix, but present in the 

 arboreal Erethizon. 



Since in cursorial forms in which the humerus moves only in 

 a plane that runs parallel to the body, no lateral strain is brought 

 to bear on the glenoidal fossa, of course in such forms clavicles 

 are useless, but they must be retained in the forms that extend 

 the arms strongly sidewards, for here such lateral strains occur. 



If birds are derived from terrestrial forms in which the arms 

 and the shoulder-girdle wei'e temporarily reduced, it is in the 

 shoulder-girdle that this must show. 



In the Thecodontia, in which, as shown by the relative length 

 of the fourth digit, the elbows stood yet somew^hat oft' fi'om the 

 body, an interclavicle and clavicles are present ; in the Orocodilia, 

 only the interclavicle remains, but also the fourth toe is abbrevi- 

 ated, showing that the lateral strain on the leg was already 

 relatively small ; finally in all the Dinosaurs, as we know, clavicles 

 and interclavicula are absent. In Dinosaurs, moreover, a well- 

 ossified sternum is present, arising from two centres of ossification. 



If we turn now from these reptiles to modern and extinct birds, 

 we find always an osseous sternum and further on a bone that 

 has been considered as corresponding to the clavicle of reptiles. 

 This bone has been called the furcula, and it is this name tliat 

 I adopt, for it has no phylogenetical meaning. In some birds 

 this bone touches the sternum, in others, as ArchcBopteryx, it 

 does not. 



Not even the slightest trace of an interclavicle has ever been 

 found in birds (6), although this bone has been eagerly sought 

 for. The scapula and coracoid in Arcliceopteryx recall these 

 elements in the primitive bipedal Dinosaurs and Struthious 

 birds (text-fig. 7). 



If the present arboreal birds are the direct descendants of a 

 clavicle -bearing quadrupedal lizard-like reptile that took to 

 climbing trees, surely a clavicle ought to be present, and perhaps 

 even an interclavicle might be expected, for it is exaxitly the 

 interclavicle that is retained both in Crocodiles and Mosasaui's 

 when the osseous clavicles are already gone. If features exist 

 that show that the furcula of birds is a newly-acquired bone and 

 that both clavicles and interclavicle are gone, then this is an 

 important argument in favour of a primitive running Proavis. 



As can be seen, a great part of the question turns on the nature 

 of the furcula of birds. 



While the clavicles in all animals are always paired bones 

 attached to the scapula with one end and to the interclavicle 

 with the other, the furcula of birds adheres to the coracoid with 

 one end and fuses, even in Arcliceopteryx. in the middle with the 

 furcula from the other side. 



31* 



