The coronoid process of the dentary has increased enormously, and now overlaps the rest of the lower jaw so greatly that it has nearly, but not quite, acquired an independent articulation on the squamosal.
The hinder part of the lower jaw, though still retaining all its constituent bones, has become so small that it seems inadequate to resist the very great stresses to which it might be subjected during feeding. The reflected lamina of the angular is still pres ent in the form of a slender downturned process.
The palate has changed greatly, the original roof of the median area is still present in part, but it is concealed from view by a sec ondary palate exactly like that of a mammal, which is formed by ingrowths from the maxillae and palatines. By this change the posterior nares are driven so far backward that they open be hind the cheek teeth, and the animal became capable of breathing whilst the mouth was full of food undergoing mastication. The ectopterygoid is greatly reduced, and the posterior part of the pterygoid, the quadrate ramus, has vanished altogether, its place being taken by a process which grows backward from the root of the epipterygoid.
The side walls of the brain case have become bony by a widen ing of the upper end of the epipterygoid, now recognizable as the homologue of the mammalian alisphenoid. This arrangement in volves the inclusion in the cranial cavity of a space, the cavum epiptericum, which in most other reptiles lies outside the cranium.
By a continuation of changes in the same direction as those which converted a Gorgonopsid such as Scymnognathus into the Cyno dent Cynognathus, a primitive mammalian structure is easily reached.
The face changes little, a disappearance of the internarial proc esses throws the bony nostrils into one, the prefrontal and post orbital disappear, and the orbit becomes confluent with the tern poral fossa.
Further growth of the dentary leads to the development of a new temporo-mandibular joint between that bone and the squa.
mosal, and the quadrate and hinder part of the jaw, freed from any function in connection with the jaw, become available as audi tory ossicles. The stapes persists, little changed, the quadrate, further reduced in size, becomes the incus. The articular is the
malleus, and the prearticular its processus folianus. The angular, to which the tympanic membrane has been attached for a very long period, becomes the tympanic, and the surangular disappears.
This account of the origin of the mammalian auditory ossicles is confirmed by the mode in which those bones develop in every mammal; indeed, all marsupials are still born in a stage in which the lower jaw still moves on the old reptilian joint between the incus and malleus, and the musculus tensor tympani still functions as a jaw muscle.
The palate of Cynognathus requires very few modifications to become typically mammalian. The already minute ectopterygoid vanishes, the great flanges of the pterygoids, which exist to ensure the accurate closure of the mouth, become unnecessary when the new temporo-mandibular joint is established, and vanish, and the posterior ramus of the alisphenoid becomes the tympanic process.
The skull of Scymnognathus is connected to the complex atlas by a single condyle composed of the basi and exoccipitals, that of Cynognathus has a mammal-like pair of exoccipital condyles.
More serious modifications have to be made in the ear region and brain case. The opisthotic and pro-otic of Cynognathus house only part of the inner ear, the summit of that organ lying in the supraoccipital. In mammals, the whole lies in a single bone, the periotic. The mammalian periotic is a much smaller bone than the pro- and opisthotics of a cynodont, and, unlike them, it is compara tively unimportant as a buttress for the squamosal. None the less, it is not impossible to homologize the different regions of the two sets of structures.
The post-cranial skeleton of Cynodonts shows a similar resem blance to that of mammals, some of the more important features of the evolution being discussed in the section Locomotion of this article. Thus we know in considerable detail the evolutionary stages which lie between the structure of an embolomerous amphibian and that of a mammal. Unfortunately, we can trace no such ancestry for the birds. We are certain that they sprang from a group of reptiles very remote from the mammal stock, but we are still faced by a considerable gap.