No. 387.] THE OSSICULA AUDITUS. 229 
dentition is not extremely rare in the non-mammalian verte- 
brates, which are certainly far removed from the mammalian 
line. On the other hand, the theromorpha have single-rooted 
teeth throughout, while certain of the dinosaurs have them with 
two roots. Heterodont dentition may easily be explained by par- 
allel development from similar conditions. 
In the theromorphs, as in the higher mammals, the coracoid 
is united to the scapula. This point is as favorable to the 
amphibian as to the theromorphous ancestry,, since a similar 
state of affairs occurs in certain existing urodeles. 
In theromorphs, as in mammals, there are bicipital ribs, one 
head articulating with the neural arch (diapophysis), the other 
with centrum or intercentrum. Here again the urodeles will 
fill the bill. 
In the theromorphs, as in many mammals, there is an ento- 
picondylar foramen in the humerus. This seems a feature of 
minor importance, since it is lacking in many mammals, while 
it is developed in some forms (e.g., Hatteria) which cannot have 
had a theromorphous ancestry. 
In the theromorpha, ischium and pubis fuse to form an in- 
nominate bone. This feature also occurs in certain urodeles 
as well as in some other reptiles. 
In certain theromorphs (Clepsydrops) there is a differentia- 
tion of calcaneum and astragalus, recalling the relations in the 
mammals. Unfortunately, almost nothing is known of the foot 
structure in other theromorphs, and the resemblances pointed 
out in Clepsydrops are not conclusive. In all other sauropsids 
there is a strong tendency towards the development of an in- 
tratarsal ankle joint. Certainly the mammalian tarsus could 
have been derived directly from that of the Amphibia instead of 
indirectly through the theromorphs. 
There is one serious objection to the theromorphous ancestry 
of the mammals upon which sufficient weight has not been 
placed. This view assumes that the suspension of the lower 
jaw in the theromorphs is, at least in part, homologous with 
that in the mammals. It assumes that the glenoid fossa of the 
latter has arisen by the squamosal of the reptile usurping the 
functions of the quadrate. But here lies a difficulty other than 
