GEOGRAPHIC AND GEOLOGIC DISTRIBUTION 381 



very imperfect material; this, in the case of dinosaurs particularly, ren- 

 ders comparative study a matter of extreme difficulty and is apt to give 

 rise to complications of synonymy. The best comparative study of Old 

 and New World Sauropoda which I have seen is that by Professor Marsh, 

 a summary* of which was first published in 1889, and later in "Dinosaurs 

 of North America," in 1896 (page 185). In studying the European 

 types. Marsh was impressed Ity throe prominent features in the specimens 

 investigated : 



"(1) The apparent absence [in Europe] of any characteristic remains of 

 the Atlantosauridne. which embrace the most gigantic of the American forms. 



"(2) The comparative abundance of anotlier family, Cardlodontld;ie (Cetio- 

 saurkhB) nearly allied to the Morosauridai, but as a rule less specialized. 



"i'i) The absence, appai'ently, of all remains of the Diplodocid*. 



"The most .striking difference between the Cardiodontidae and the allied 

 American forms is that in the former the fore and hind limbs appear to be 

 more nearly of the same length, indicating a more primitive or generalized 

 type. Nearly all the American Sauropoda, indeed, show a higher degree of 

 specialization than those of Europe, both in this feature and in some other 

 respects." 



At least two American genera of Sauropoda have been recognized in 

 Europe, but the identity in each instance seems open to question. These 

 genera are Morosaurus and I'lcuvocoelus, the former a typical Morrison 

 form, having been reported I'rum the Oxfordian of Ourem, Portugal, and 

 from the Wealden of England. Pleuroccelus, a genus which includes the 

 smallest ol' the Sauropoda, has hut a single rare Morrison species referred 

 to it, while two forms, PleuroraduH nanus and P. alius, are known from 

 the Potomac formation of Maryland. There is reason to l)clieve that in 

 tlic (irst of the Maryland species at least the fore limbs were nearly equal 

 to tlic hiiiil ill Iciigtli. aiul in other respects, as well as in its relatively 

 siiuill size, tlu' creature was of primitive type. Bofliriospondylus, de- 

 sci'ilicd l)v Sir liicliard r)\v('n. from the Kimmeridgian of England, was 

 lliought Ity Marsh to I'cpi'cscid a \ery young, if not foetal, individual, 

 wiiich miglit he nearly allied to Pleuroccelus, while species from the 

 Wealden of Caen, France, and the Aptian of Portugal are- referred to the 

 'American genus itself. 



There are, so far as 1 know, no generic comparisons to he made be- 

 tween the p]uropean Sauropo(hi and those of Tendaguru, although the 

 name Giganlosaurus a|)pears in each faunal list, but apparently applied 

 to forms wliicli ;ire not congeneric. On the other lumd, it is difficult to 

 make comparisons with the ^Torrison ty|ics A\liich wouM he of strati- 



<Amer. Jour. Sci. (3), vol. xxxvll, pp. 323-331. 



