1902.] OSTEOLOGY OF THE FALCOXI FORMES. 309 



deep in the remaining forms, whilst the ossified extensor bridge 

 is conspicuous by its absence. The absence of this bridge should 

 surely be as valuable a piece of evidence that Pandion is not 

 related to the Striges, as is the presence of an extensor bridge 

 on the tarso-metatai'sus to show that it is so related ! Further- 

 more, we may remai-k that this tarso-metatai-sal bridge is wanting 

 in the Barn- Owls, so that the value of this character is still 

 further weakened. 



The tibio-tarsus of Polyhoroides is, in its way, almost as re- 

 mai'kable as that of Pandion. It is long and almost of the same 

 thickness throughout. As in Pandion, the width of the shaft 

 immediately above the distal condyles is greater than across the 

 condyles themselves. This is pi-obably due to the fusion with 

 the shaft of the distal end of the fibula which extends down to 

 the tarsal elements — though, as just indicated, fused with the 

 shaft. The extensor- biidge is placed ti-ansvei-sely across the shaft. 

 The tibial condyles are but feebly developed, the ectocondyle 

 barely projecting beyond the shaft. The intercondylar goi-ge is 

 wide. 



The lower thii-d of the shaft of the larger Accipitres is dorso 

 ventrally depressed. There is often an additional and well-defined 

 fibular bridge. The second lies some distance below the first, and 

 afibrds attachment for the distal end of the fibula. The cnemial 

 ci-ests are relatively but slightly developed. In Aqu'ila, perhaps, 

 they i-each the maximum development, the entocnemial crest being 

 unusually sti'ong. In Thrasaetus the ecto- and entocnemial con- 

 dyles are conspicuously wide apai-t, causing the inner border of 

 the shaft to have a very marked curve. 



The fibula, as a rule, tapers to a fine point, and terminates near 

 the lower third of the tibial shaft, sometimes fusing therewith. 

 In Pandion, Polyhoroides, and Pernis only is it of almost uni- 

 form thickness and continued downwards as far as the tarsal 

 elements. 



The tarso-metatarsus vai'ies extremely, both in its relative 

 length and in the development of bony matter for the mechanical 

 requirements of the limb. 



In Serpentarius only is the tarso-metatarsus as long as the 

 tibial shaft. In Accipiter, howevei', it is veiy neai'ly so. Generally 

 it is shorter than the femur. 



The hypotarsus is simple in all save Pandion and Pe7^nis, 

 wherein it forms a tube. 



The distal trochlear, save in Serpentarius and Cathai^tse and in^^-l^t-b/A^ 

 Leptodon, are all on the same level. In the two first- mentioned 

 exceptions the middle trochlea is produced somewhat beyond the 

 level of the others. In Leptodon the inner trochlea is the longest, 

 the middle slightly shorter, and the outer shorter still, so that an 

 obliquely sloping series is foimed. The plane of the trochleas 

 forms a slight and regular curve, except in Pandion, in which this 

 curve is very strong ; and Polyhoroides, in which the 2nd and 3rd 

 trochleae lie close together, whilst the 1st is l)ent downwards so 



