February 4, 1898.] 



SCIENCE. 



177 



dental and mandibular characters of the pro- 

 mammal, and from the investigations of Baur, 

 Howes, Hubrecht, Beddard, Albrechtand others 

 are derived, in the skeleton and soft parts, 

 other characters which are largely amphibian. 

 A comparison of the Dicynodontia, Therio- 

 dontia and Gomphodonfia (Seeley) shows that, 

 while widely separated in dental characters, 

 these reptiles are closely united in numerous 

 osteological characters, which, in turn, distin- 

 guish them from all other reptiles, the most strik- 

 ing regions being the palate, zygomatic arch, 

 expanded squamosal and correlated reduction 

 of the quadrate, and the complex structure of 

 the occipital condyle. Owen's definitions are 

 too narrow for this group, which appears to be 

 embraced only in Cope's larger definition of 

 the Theromora. As is well known, the Di- 

 cynodontia throughout the skeleton abound in 

 mammalian characters, and in the skull exhibit 

 a combination of special adaptations to the 

 greatly developed canine teeth with persistent 

 reptilian and promammaliau characters. The 

 term Theriodontia should be restricted, accord- 

 ing to Owen's original definition, to carnivorous 

 types, such as Cynognaihus, with triconodont 

 molar teeth and typical promammaliau dental 

 formula. The characters of the skull, teeth, 

 vertebrse, pectoral and pelvic arches and limbs 

 on the one side show the affinities of these ani- 

 mals to the Dicynodonts, and, upon the other, 

 make them appear prophetic of the Jurassic 

 Triconodonta. The Triassic Protodonta are 

 quite as primitive in dentition, but different so 

 far as known in that the jaw consists of a single 

 bone. The third group, or Gomphodontia, 

 Seeley, embraces herbivorous types with 

 grinding teeth of multi tubercular and rudely 

 tritubercular pattern. The latter fact is of 

 great significance in the support it apparently 

 lends to Osborn's hypothesis that the multi- 

 tuberculates are of trituberculate origin. In 

 cranial characters these animals are as similar 

 to the Theriodonts as they are dissimilar in 

 dental characters, and since they include the 

 genus Tritylodon, which was formerly placed 

 among the Multituberculata, it appears pos- 

 sible that we have here a phylum more or less 

 remotely related to the very ancient Mesozoic 

 phylum of Multituberculata. 



Summing up the Theriodont characters we 

 find promammalian resemblances both in the 

 form and formulse of the teeth ; in the terminal 

 position of the anterior nares and structure of 

 the palate ; in the posterior expansion of the 

 nasals ; in the main infratemporal or zygomatic 

 arch ; in the great development of the squa- 

 mosal and reduction of the quadrate ; in the 

 paired occipital condyles ; in the intercentra of 

 the cervical vertebrae ; in the suturally united 

 cervical ribs ; in the intervertebral anterior 

 dorsal ribs ; in the Monotreme type of scapular 

 arch (excepting, perhaps, the epicoracoid united 

 by suture withmetacoracoid); in the prescapular 

 spine; in the powerful deltoid crest, large entepi- 

 and ectepicondyles and entepicondylar foramina 

 of the humerus. The limb and pelvic structure 

 is evidence of a musculature similar to that of 

 the hypothetical Promammalia, and of a body 

 well raised above the ground and quadrupedal 

 in position. As persistent reptilian characters 

 may be cited the separate prefrontals, post- 

 frontals and postorbitals and separate quad- 

 rate, which, according to Albrecht, is a rever- 

 sional character in the Mammals; the separate 

 transversum ; distinct prevomer and complete 

 pterygoquadrate arcade ; the prominent basioc- 

 cipital element ; the separate elements of the 

 lower jaw, and finally as adaptive or specialized 

 characters are the several peculiar features in 

 the back, skull and other parts of the skeleton. 

 In conclusion, it appears that these true Ther- 

 mora have the geological age required for the 

 ancestors of the Mammalia. They are the only 

 class of reptiles which exhibit mammalian affini- 

 ties. They anticipate in the most surprising 

 manner the dental structure of the ancient Tri- 

 conodonla and Multituberculata. A most striking 

 difference is found in their size, which far ex- 

 ceeds that of the oldest undoubted Mammals. 

 This and certain specializations of structure 

 bar any of the known Therviora from the ances- 

 try of the earliest mammals, but do not pre- 

 clude the existence of very small, unspecialized 

 forms, which may have given rise to the oldest 

 mammalian types. The existence of Amphibian 

 structures, as observed by Hubrecht and others, 

 in the placentation and soft anatomy of the 

 mammals may be explained by the supposition 

 that these Thermora retained certain Amphib- 



