52 PAEALLELISM IN DEVELOPMENT. 



ruminant and a two-clawed horse. To our thinking, this 

 is indeed one of the most curious phases of development 

 yet discovered. 



In addition to the j)arallelism in the evolution of their 

 molar teeth and limbs, the hoofed mammals likewise 

 exhibit the same feature in regard to the second vertebra 

 of the neck. As many of our readers are probably awai-e, 

 the first, or atlas vertebra of the neck, turns with the 

 head when the latter is moved sideways, the axis of 

 rotation being formed by a process — the odontoid process 

 — arising from the second or axis vertebra, and projecting 

 into the central hollow of the atlas. Now in pigs and 

 likewise all the jirimitive hoofed mammals, the so-called 

 odontoid process is (as in ourselves) in the form of a 

 flattened peg. On the other hand, in the ruminants, the 

 modern horses, and the camels, which, as we have seen, 

 represent three distinct 'phyla of the order, this peg has 

 become modified into a spout-like half- cylinder, which 

 must clearly have been separately evolved in each of these 

 three groups. It is true that such a half-cylinder affords 

 a far better liasis of support for a heavy skull than does a 

 mere peg ; but the curious ].)art of the matter is why these 

 half- cylinders should be so exactly alike in the different 

 groups, seeing that, as it would not be difficult to design 

 some other structural modification by which the same end 

 might have been attained, there is no necessity for their 

 similarity. 



Our last examjile of parallelism will be drawn from two 

 groups of extinct JSTorth American hoofed mammals, to 

 which brief allusion has already been made in the chapter 

 on "Tusks and their Uses" ; the one group being known as 

 uintatheres, while the second is represented by a single 

 species to which the name of Protoceras has been applied. 

 Now, although the uintatheres have five-toed feet approxi- 

 mating in structure to those of elephants, while in Proto- 

 ceras each foot ajiproximated to the ruminant type, in both 

 groups the skull, as shown in the accompanying figure, 

 was armed with several pairs of large, irregular, bony 

 processes, which during life may have been sheathed in 

 iiorn ; while in each case a pair of long tusks projected 



