346 THE STRUCTURAL EVIDENCE OF EVOLUTION". 



the case in the Hyracoidea and many of the Quadrumana, but in 

 the anthropoid apes and man the second row is displaced inwards 

 so as to alternate with the first row, thus interrupting the series in 

 the longitudinal direction, and forming a stronger structure than 

 that of the Condylarthra. In the Bunotherian Rodent and Eden- 

 tate series, the tarsus continues to be without alternation, as in 

 the Condylarthra, and is generally identical in the Carnivora. In 

 the hoofed series proper it undergoes change. In the Proboscidia 

 the carpus continues linear, while the tarsus alternates. In the 

 Amblypoda the tarsus alternates in another fashion, and the car- 

 pal bones are on the inner side linear, and on the outer side alter- 

 nating. The complete interlocking by universal alternation of the 

 two carpal series is only found in the Diplarthra. (C) As to the 

 ankle-joint. In most of the Condylarthra it is a flat joint or not 

 tongued or grooved. In most of the Carnivora, in a few Rodentia, 

 and in all Diplarthra, it is deeply tongued and grooved, forming a 

 more perfect and stronger joint than in the other orders, where 

 the surfaces of the tibia and astragalus are flat. (D) In the high- 

 est forms of the Rodentia and Diplarthra the fibula and ulna be- 

 come more or less co-ossified with the tibia and radius, and their 

 middle portions become attenuated or disappear. 



Secondly, as regards the vertebras. The mutual articulations 

 (zygapophyses) in the Condylarthra have flat and nearly horizon- 

 tal surfaces. In higher forms, especially of the ungulate series, 

 they become curved, the posterior turning upward and outward, 

 and the anterior embracing them on the external side. In the 

 higher Diplartha this curvature is followed by another curvature 

 of the postzygapophysis ujDward and outward, so that the vertical 

 section of the face of this process is an S. Thus is formed a very 

 close and secure joint, such as is nowhere seen in any other 

 Vertebrata. 



Thirdly, as regards the dentition. Of the two tjipes of Mono- 

 tremata, the Tachyglossidse and the Platypodid£e, the known gen- 

 era of the former possess no teeth, and the known genus of the 

 latter possesses only a single corneous epidermic grinder in each 

 jaw. As the Theromorphous reptiles from which these are de- 

 scended have well-developed teeth, their condition is evidently one 

 of degeneration, and we can look for well-toothed forms of Mono- 

 tremata in the beds of the Triassic and Jurassic periods. Perhaps 

 some such are already known from jaws and teeth. In the mar- 

 supial order we have a great range of dental structure, which almost 



