1910.] Tarsu.s of Oudenodon, etc. 453 



the dominant factors and to have as it were, greater growth vigor to ci-owd 

 and modify their fellows. As shown by Weithofer (1888) and Osborn, (1889, 

 p. 539) in the Proboscidea the ulna and the lunar are in the ascendant, the 

 scaphoid is literally crowded out; in the Amblypoda the unciform and 

 digit III broaden, in the Hippoids it is digit III and the magnum ; in Macmu- 

 chenia and the Litopterns the cuneiform acquires a broad contact with the 

 magnum, widely disrupting the former lunar-unciform contact. 



Conceivably the active or dominant features may have been "unit char- 

 acters" or orthogenetic characters which were not divided in crossing, while 

 the elements which appear to be passive may have been more subject to 

 fluctuating variation and to division. The evolution of the carpals furnishes 

 a good illustration of the very obvious principle that thei'e is a strong analogy 

 bet\\een the evolutionary relations of parts to each other and of parts to 

 environment. 



Tlie Hind foot. 

 1. The astragalus. 



Oudenodon. The pes of Oudenodon figured by Broom (Fig. 28, no. 2) 

 indicates that the true "tibiale tarsi" is not the sesamoid so named by 

 Baur (1885) in Rodents, GaleopitJtecus, etc., but the astragalus itself, the 

 intermedium being a greatly reduced element. 



The astragalus and calcaneum of Oudenodon are prototypal in form to 

 those of mammals, because they are very simple in shape and the astragalus 

 is not differentiated into trochlea, neck and head, nor the calcaneum into 

 tuber, sustentaculum, etc. The other elements of the pes are homologous 

 with those in mammals. Most of the elements of the pes correspond in 

 position with those in the manus, but the lunar (intermedium) of the manus, 

 which was destined to increase in size, corresponds with the intermedium 

 in the pes which was destined to disappear as a separate element; hence in 

 the mammals the correspondence between the manus and pes is never com- 

 plete (see below, page 455). 



Monotremes. The pes is described on page 154. 



Marsupials. The astragalus and calcaneum of the ancestral Marsupials 

 (as described below) are adapted to a divergent hallux and more or less 

 arboreal habits. As observed in the smaller Marsupials these bones are 

 especially interesting because they are morphologically intermediate between 

 the Therapsid and Monotreme types on the one hand and the Placental types 

 on the other. F'or example, in the astragalus of Didclphis (Fig. 30, .1) the 

 trochlea is very broad, with a very indistinct groove while the neck is feeble 

 and not differentiated from the head, the sustentacular facet is in the middle 



