78 MR. F. E. BEDDARD ON THE [Feb. 19, 



trinae of the family Falconida?), and by G. R. Gray ', is, as Milne- 

 Edwards acknowledges, in the main justified by the osteological 

 characters ; Milne-Edwards, however, considers that its peculiarities 

 necessitate the creation of a separate subfamily for its reception. 

 This view is accepted by J. H. Gurney". 



The Accipitres have been divided by Prof. Huxley ' into three 

 groups — (1) Cathartidse, (2) Gypaetidse, (3) Gypogeranidse — on the 

 characters of the skeleton. Prof. Garrod's investigations* emphasized 

 the naturalness of this grouping ; he showed that these three divisions 

 could be defined by the presence or absence of certain muscles in 

 the leg. 



In the Cathartidse the ambiens, semitendinosus and accessory semi- 

 tendinosus, and femoro-caudal are present, the formula being on 

 Garrod's system AXY-f-. 



In the Gypaetidee (termed Falconidae) the muscles present can be 

 indicated by the formula A-(-. 



In the Gypogeranidse (Serpenfarius) the formula is BXY-|-. 



These muscular divergences led Prof. Garrod to remove Serpen- 

 tarius and the Cathartidse from the Accipitres and to associate them 

 with other birds. Without following Prof. Garrod in this latter 

 alteration of existing arrangements, it must certainly be admitted 

 that his results entirely justify the breaking up of the Accipitres 

 into the three groups already indicated. 



I do not, however, tind myself able to agree with Prof. Garrod in 

 believing that the absence of the semitendinosus muscle is absolutely 

 distinctive of all the Gypaetidse. 



I have found this muscle in Falco suhbuteo, where it was rather 

 feeble and apparently fused at its origin with the semimembranosus, 

 but it ended in a separate and perfectly distinct tendon and was 

 present on both legs ; in the Merlin {Falco cesalon), where it was a 

 little better developed ; and finally in Circus mavrus. 



Apart from these exceptions, which do not affect the classification 

 of the group, the formula of Gypaetidse is, as stated by Garrod, A -}-. 



Poli/boroides ti/picus possesses the ambiens and femoro-caudal 

 alone of the leg-muscles, upon the variations of which Garrod's 

 system was based ; it therefore agrees with Accipiter, Circvs, &c., and 

 should be referred to the Gypaetidse and not to the Gypogeranidse. 



In examining the muscles of the wing I have compared Polybor- 

 oides with Serpenfarius, Cathartes, and with Gypohierax as a type of 

 the Falconidae. 



The tensor pat agii brevis' of each wing is a stoutish muscle which 

 divides into two tendons, inserted as shown in fig. 1 (p. 79) ; each 

 tendon is shght and thin and of equal diameter throughout. 



^ Hand-list, i. p. 38. 



■^ ' A List of the Diurnal Birds of Prey,' &c. (London, Van Voorst), p. 18 

 3 " On the Classification of Birds," P. Z. S. 18(i7, p. 4fi2. 

 '' "On certain Muscles in the Thigh of Birds, and on their Value in Classifi- 

 cation," P. Z. S. 1873, p. (534. 



5 These muscles were dissected in another specimen. 



