388 
Oti the Sesamoids of tlte Knee- Joint 
patella of Mergus servator, to the flake-like patella of Colymhiis, up to the huge 
quadrate patella of Aptenodytes pennantii. But he does not seem to lay any stress 
on the fact that where the cnemial crest is large as in Colymhus and Fulmarus 
rodgersii the patella is small, and where the patella is large as in Podiceps the 
cnemial crest is relatively small. Yet this result seems all important and easily 
interpretable on the assumption of the patella arising from a tibial crest. Lastly 
Shufeldt gives no explanation of the condition in the cormorant Phalocrocoraw 
hicristatus of which he writes that it is almost unique ; that on the face of it 
it almost looks as if a patella had developed of a size equal to that of the tibial 
crest and subsequently the two became thoroughly united and formed one large 
patella. This suggestion appears to us as an inversion of the real state of affairs. 
It also seems to be D'Arcy W. Thompson's view*. He treats the large patella of 
Hesperornis as homologous with the tibial crest of Colymhus. He actually speaks of 
this crest as a " co-ossified patella." He writes : " The existence of a small additional 
sesamoid in the knee-joint of Colymhus (Owen, Camp. Anat. ii. p. 83) does not 
invalidate the homology here adopted of the long ' rotular process of the tibia ' with 
the patella." 
If the cnemial crest be identified with the patella, then it seems to us more 
reasonable to look upon the patella as a released tibial crest, than the tibial crest 
as a co-ossified patellaf. It has not yet been suggested that the olecranon is a 
co-ossified sesamoid. Perhaps one of the chief difiiculties in the idea of the tibial 
crest as a " co-ossified " patella is the large free patella of the cormorant ; here the 
pro- and ecto-cnemial ridges of the tibia are carried right up the anterior face of the 
patella itself! (See Plate XXXVI, Fig. 93 h.) Why should an independent sesamoid 
have developed these ridges ? Unless we accept the view that a sesamoid can 
always arise where it is a convenience, this leaves us still without an explanation 
of the origin of either the sesamoid or the patella. We have at any rate to bear in 
mind the very close association of patella and cnemial crest where they co-exist J. 
We note also that whereas a true patella, an orthosesamoid, exists in all 
placental mammals, it is absent in many Marsupials, it is not invariable in Birds 
and has disappeared when we get down to the Reptiles. Somewhere between the 
Reptiles and the branchpoint of the Birds a patella had to arise, somewhere between 
the Reptiles and the branchpoint of the Monotremes a fibular process had to arise §, 
* "On the systematic position of Hespcrornis" Studies from tJie Museum of Zoolopy hi University 
College, Dundee, Vol. i. p. 108, Dundee, 1890. 
t In both Hesperornis and Ichthyornis the tibia has a considerable epicnemial crest, in the former as 
well as the massive patella, in the latter no patella is figured or referred to; whether none has been 
found or it is assumed not to exist, we cannot say. See Marsh, "Odontornithes," U. S. Geological Ex- 
ploration of the iOth Parallel, Vol. vii. Plate XXXIV, 1880. 
% In the fossil Hesperornis re galis ythich we have examined the patella so far accords with Professor 
D'Arcy Thompson's view, that it has at least a superficial resemblance to a liberated cnemial crest. See 
also Plate XXXVI, Fig. 92. 
§ An approach to a fibular crest occurs in some birds ; thus in the Cereopsis Goose (see Plate XXI, 
Figs. 48 and 49), the head of the fibula rises above the proximal articulating surface of the tibia, and 
works in a groove of marked character on the posterior face of the lateral condyle. There is also a sub- 
stantial cnemial crest. 
