876 



SCIENCE. 



[N. S. Vol. IV. No. 102. 



mation in the Cretaceous. The numher of 

 persons to whom the local question of correla- 

 tion is important may not be large, but the 

 whole body of geologists and paleontologists 

 are concerned with the methods and principles 

 of correlation, and an excellent opportunity 

 seems to be here afforded for the comparison of 

 vertebrate with botanic evidence. I therefore 

 write to express the hope that when Prof. 

 Marsh continues the subject, as he has promised 

 to do, he set forth the grounds for the conclu- 

 sion he has announced with so much confidence. 

 His article states, in effect, that through a 

 comparison of vertebrates from the Potomac 

 formation with vertebrates from other forma- 

 tions he has inferred the Jurassic age of the 

 Potomac ; but he gives no hint of the character 

 of his evidence or the course of his reasoning, 

 so that the conclusion has at present only the 

 authority of his statement, without opportunity 

 for verification. Unless I am mistaken, the 

 conclusion that the Atlantosaurus and other 

 horizons of the Rocky Mountain region are 

 Jurassic was announced in the same way, with- 

 out citation of evidence ; and in that case it is 

 important to establish the correlation of the 

 Potomac beds, not merely with these Western 

 horizons referred to the Jurassic, but with 

 European beds whose age admits of no ques- 

 tion. 



My own desire to learn Prof. Marsh's method 

 of correlation is stimulated by certain considera- 

 tions which seem to show that it must differ in 

 an important way from the method ordinarily 

 used by students of invertebrate fossils and 

 fossil plants. As he has pointed out, land ver- 

 tebrates are peculiarly sensitive to climatic and 

 other physical conditions, and the evolution of 

 new forms is consequently rapid. The life of a 

 species is short, and its value for purposes of 

 correlation is correspondingly high because its 

 chronometric indication is precise ; but it ap- 

 pears to me a priori that this quality of rapid 

 evolution is a two-edged sword ; while it facili- 

 tates correlation within the same faunal prov- 

 ince, it introduces a difficulty when remote 

 provinces are compared. In remote provinces 

 the progress of evolution must follow different 

 lines, so that there can usually be no common 

 species for comparison. Therefore, as correlation 



by means of other organisms depends chiefly on 

 the comparison of faunas through their identical 

 species, correlation by means of land vertebrates 

 must have a different basis. This point seems 

 to be illustrated by the general problem of Jur- 

 assic correlation. In Prof. Marsh's recent classi- 

 fication and synopsis of the Dinosauria (Six- 

 teenth Annual Report, U. S. Greological Survey) 

 it appears that only the higher categories of 

 classification include representatives from both 

 Europe and North America. There is no com- 

 mon species ; there is not even a common 

 genus. Of nineteen families referred to the 

 Jurassic, six are European only, eleven are 

 North American, and but two span the Atlantic. 

 One of these last is not peculiar to the Jurassic 

 and is therefore of minor value for the correla- 

 tion of American horizons ; so that the closest 

 affinity of the European and American forma- 

 tions seems to be expressed by the statement 

 that there is one American genus which falls in 

 the same family with a European genus. 



There is yet another reason why inference in 

 this particular case needs to be fully supported 

 by evidence, and that is that the physical re- 

 lations of the beds afford a presumption in 

 favor of their Cretaceous age. Prof. Marsh 

 mentions that the Potomac formation in New 

 Jersey passes by insensible gradation into 

 marine Cretaceous above and is separated by 

 unconformity from rocks of supposed Triassic 

 age below ; but he apparently sees in this re- 

 lation merely the fact that the Potomac lies be- 

 tween formations of Cretaceous and Triassic 

 age. The geologist, however, infers that the 

 unconformity beneath the Potomac represents 

 a time interval, and consideration of the ex- 

 tensive dislocation and deformation of the 

 Newark beds and of the enormous degradation 

 they suffered before the deposition of the Po- 

 tomac gives the impression that that time in- 

 terval was very long as compared to the time 

 represented by the Potomac beds themselves. 

 When it is further considered that the lowest 

 marine horizon determined above the Potomac 

 is correlated by its invertebrate fossils with a 

 horizon somewhat above the middle of the 

 European Cretaceous, it seems easier to assign 

 the Potomac to the lower Cretaceous of Europe 

 and correlate the time-break with the Jurassic 



