Karl Pearson and Adelaide G. Davin 
389 
and somewhere between that branchpoint and that of the Marsupials the fibular 
process had to be replaced by a parafibula, and lastly somewhere between the 
branchpoint of the Marsupials and the early placental mammals a cyamella and a 
lateral fabella had to replace the parafibula. We can of course take the view that 
a process will disappear without vestiges, if it ceases to be useful, and that a sesamoid 
will appear whenever and wherever it is likely to be useful*. But on philosophical 
grounds there is need of a connected account of the evolutionary origin and 
subsequent history of even the sesamoids of the knee-joint. 
(22) Reptiles. 
The chief feature of the knee-joint sesamoids of the Reptiles consists in the 
presence usually of a considerable number of lunulae. Parker and Haswellt speaking 
of the young condition of certain Marsupials state that there is an accessory element 
situated outside the fibula at its proximal end ; " this apparently corresponds to a 
bone known as the parafibula which occurs in some Lacertilia." When the cycwhella 
has left the popliteal groove and articulates with the head of the fibula, it becomes 
very difficult to say whether we are speaking of a cyamella or a parafibula. In our 
opinion the parafibula ultimately is divided into two elements ; the upper portion 
becomes associated with the lateral head gastrocnemius and is the lateral /'a6e^/a 
of the placentals, the lower portion associated with popliteus is the cyamella proper. 
Hence when there is no fabella as in the Reptiles and Marsupials it seems a more 
reasonable course to term the single large sesamoid, which articulates with the head 
of the fibula and usually at the same time with the under surface of the lateral 
condyle, the parafibula. At the same time popliteus in most of these cases arises 
from the fibula, it may also be united with the semilunar, and it is not always easy 
to distinguish this form of the cyamella, a small parafibula, from a lateral anterior 
lunula. Our drawings, Plate XXIV, Figs. 64 a and 64 h, show a typical reptilian knee- 
joint, that of Varanus niloticus (Nile monitor). The anterior aspect shows the 
cyamella as a parafibula. The popliteus has a tendon in the popliteal groove of the 
femoral condyle, it is also attached to the superior posterior corner of the fibula, 
just internal to attachment of the external lateral ligament shown in the drawing. 
There are two anterior lunulae. The posterior aspect shows two posterior lunulae, 
a small mesial and a much larger lateral lunulaX- In Varanus the cyamella has 
become entirely anterior ; it thus diff'ers widely in position from the cyamella in 
higher mammalian forms. Cf. the Orang in our Plate XXVI, Figs. 68 and 69. 
This position of the cyamella is not universal in reptiles. Thus in Trachysaurus 
rugosus, the Australian stump-tailed lizard, the cyamelhi (see our Plate XXIV, 
Figs. 63 a and 63 h) is seen from both posterior and anterior aspects, and it is more 
easily confused with a lunula of the lateral semilunar. The knee-joint of this lizard 
has in addition two anterior lunulae and two posterior lunulae, of which the internal 
is larger than the external — exactly the opposite to what occurs in Varanus. 
* We have already indicated the difficulty of arguing as Gruber does the value of the knee-joint 
sesamoids to man in face of the fact that 90*^/0 of mankind do just as well without tbem. 
t Te.xt-hook of Zoology, Vol. ii. p. 524. 
t In the specimen of Varanus niloticus mounted in the British Museum (Natural History) only the 
parafibula and the two anterior lunuhu: are visible. 
Biometrika xni 25 
