494 E. O. ULRICH REVISION OP THE PALEOZOIC SYSTEMS 



recognizably maintained through long periods. The idea, which must 

 have been suggested already by preceding remarks on the Spergen fauna, 

 is well illustrated by the Eussian Baltic Ordovician section, which em- 

 braces between the Glauconite sandstone and the middle part of the 

 Lyckholm the results of six or seven distinct invasions of the Arctic Sea. 

 With the exception of a few evident migrants, chiefly from the north 

 middle Atlantic, this Eussian succession of Ordovician faunas strongly 

 suggests periodic exhibits of the faunal evolution in a single faunal 

 province. 



The Ordovician formations in central Kentucky, considering only 

 those containing middle Atlantic and Gulf of Mexico faunas, afford a 

 similar illustration, and the contrast between these Kentucky and Eus- 

 sian Ordovician faunas, which have less than 2 per cent of species in 

 common, offers a very striking example of the independence in constit- 

 uents and development that distinguishes the several centers of faunal 

 origin and distribution. The contrast between the faunas of the Decorah 

 shale, Prosser limestone, and Stewartville dolomite in Iowa and Minne- 

 sota, which are believed to have invaded from the Arctic basin, and the 

 Gulf of Mexico faunas in Kentucky, that correspond most nearly in age, 

 is perhaps even more remarkable. Of over 600 species in the former, 

 only about 40, or less than 7 per cent, occur in the latter, and nearly half 

 of these 40 species are recognized as long and wide ranged, perhaps 

 strictly cosmopolitan forms, whose geographic origin is unknown. 



The continuous genetic relationship of the successive faunas in sections 

 like the Eussian and the Kentucky Ordovician has commonly been sup- 

 posed to indicate practically uninterrupted sedimentation, even when, as 

 in the case of the Baltic section, the volume of deposits is greatly inferior 

 to that laid down elsewhere in the same period. In fact, however, it im- 

 plies nothing of the kind. On the contrary, the almost total absence of 

 gradation between the specific stages in the evolution of the genera proves 

 that the transition from one species to the next in its line usually took 

 place in times not represented by deposits in the local sedimentary record. 

 As it is unreasonable to assume that, with continuous opportunity, only 

 the finished specific or varietal stage entered the areas of accessible depo- 

 sitional record, and as the process of mutation, while not uniform in rate, 

 is yet continuous, the conclusion is inevitable that long periods intervened 

 in which deposition was interrupted by sea withdrawal and during which 

 mutation progressed to some later distinguishable stage. Probably mu- 

 tation was accelerated during these intervening stages of sea withdrawal 

 and consequent change in habitat, but this fact can not greatly reduce 

 the duration of the periods of withdrawal because the suggestion offered 



