1SS5.] TROCHIMD.E, CAPRIMOLGIDiE, AND CYPSELID^. 911 



Returning to our ^yhite-throated Rock-Swift, I find that the 

 proximal eiid of \ts/ei»ur is so broad in its transverse diameter as to 

 obliterate the neck, making the head of this bone more than usually 

 sessile with the shaft. 



The trochanterian ridge does not rear above the plane of the 

 summit ; in fact, the top of the bone is entirely level. Its shaft is 

 cylindrical only at its middle, from which point it gradually enlarges 

 in the direction of the extremities. 



The intercondyloid fossa is shallower than is commonly the case, 

 and the proininences themselves not so sharply defined as they often 

 are in other birds. 



This Swift does not possess a patella, as we found to be the case 

 in Trochilus. 



So poorly developed is the fibula that in the specimens before 

 me I fail to find an instance where it is produced beyond the fibular 

 ridge on the side of the shaft of the tibio-tarsus. Above this, how- 

 ever, its condition is somewhat better, and it meets the femur in the 

 usual notch of the outer condyle. 



Tibio-tarsus differs very consideraljly from this bone as it occurs 

 in the vast majority of the class. The outline of its proximal ex- 

 tremity is nearly square, and the undulating articular surface it 

 encloses gradually slopes from the inner to the fibular side. There 

 is no trace whatever of the pro- and ectocnemial ridges on the ante- 

 rior aspect, where they occur in most birds. 



The shaft is straight though slightly compressed in the antero- 

 posterior direction. As we approach its distal end, it becomes curi- 

 ously twisted, giving the condyles a pecuhar cant not easy to de- 

 scribe. The inner condyle is fully as prominent behind as it is in 

 front (rare), while the intercondyloid notch is of equal depth all the 

 way round. The bony bridge for the extensor tendons is present 

 and situated low down, while on the inner and anterior border, 

 higher up on the shaft than common, is seen a strong tubercle for 

 ligamentous attachment, a ligament which we know fulfils a similar 

 purpose. 



The tarso-metatarsus is comparatively short, and it, too, is 

 quaintly fashioned. Its hypo-tarsus is perforated by one large and 

 open groove, which absorbs its entire central portion, to the very 

 base of this process, being continued down the shaft behind as a 

 shallow excavation. Anteriorly, the shaft is longitudinally and still 

 more decidedly grooved, being pierced above by a foramen that 

 passes directly through the bone. The usual foramen occurs at the 

 distal extremity for the passage of the arterj'. 



The anterior faces of the three trochleee are about in the same 

 plane in front, while posteriorly they develop prominent processes 

 for the attachment of strong ligaments, which confine the plantar 

 tendons as tliey pass to the toes. 



Of these three trochlear projections the innermost one is the 

 lowest, the middle one rather higher, and the outermost one the 

 highest of all. 



This Swilt forms no exception to the well-known rule which 



59* 



