HATCHER : OLIGOCENE CANID^ 81 



longer than is usual in the former, but absolutely and proportionately shorter than 

 in the creodonts, Patriofelis and Hysenodon. 



The most anterior vertebra of our series of caudals I regard as the first. It is 

 short and stout, with widely expanded anterior zygapophyses bearing well developed 

 nietapophyses. The transverse processes are single, widely expanded both trans- 

 versely and antero-posteriorly. They are directed outward and backward and 

 support a considerable median prominence on the anterior edge. The top of the 

 neural spine is wanting, but the base indicates that this process was fairly well 

 developed. The above mentioned characters are all well shown in Plate XVII., 

 Figs. 6 and 7. For an anterior caudal it very closely resembles that vertebra in 

 Hoplophoneus, one of the Oligocene saber-toothed cats. But among recent carnivores 

 it is decidedly more canine than feline in character. 



I will next describe that caudal in our series which I regard as the fifth. It is 

 shown in Plate XVIL, Fig. 8. The posterior zygapophyses extend well beyond the 

 posterior end of the centrum. The posterior transverse processes point strongly 

 backward and outward and bear a rather prominent rugosity on the superior surface 

 and near the external border. This vertebra is just commencing to develop an 

 anterior transverse process and articulations for the chevrons. 



The next vertebra, Plate XVIL, Fig. 9, to be described from our series, appears to 

 be the seventh caudal. In this vertebra the anterior transverse processes are well 

 developed, being equal in expanse although more slender than the posterior. The 

 metapophyses continue as well developed expansions of the anterior zygapophyses. 

 There is a large foramen at the base and just in front of the posterior zygapophyses 

 not shown in the figure. 



The next vertebra to be described in our series I regard as the ninth caudal. 

 It is shown in PI. XVIL, Fig. 10. In this vertebra the anterior of the transverse 

 processes have become the stronger and are directed forward and outward instead 

 of directly outward. The development of the anterior transverse process of the 

 caudals of this region is much more marked in Daphwnus than in most modern 

 dogs and cats, and is quite unlike that which obtains in Hoplophoneus. This ver- 

 tebra bears on the inferior surface at its anterior extremity very prominent rugosi- 

 ties for the attachment of chevrons. 



Between the supposed ninth and the next vertebra represented in oui- series a 

 number of caudals are missing. I have estimated the number at three. This 

 would make the vertebra represented in Fig. 11, PI. XVIL, the thirteenth. In 

 this vertebra the neural canal has already become obsolete, the anterior and 

 posterior zygapophyses are reduced to mere prominences, as are also the ante- 



