A^"^-^! ORDOVICIAN TRILOBITES ULRICH 67 



that may be said to suggest that the Trhmcleus beds are of early 

 Richmond (i. e., post Maysville) age. 



Troedsson, in the stratigraphic part of his 1928 report on the 

 Middle and Upper Ordovician faunas of northern Greenland, de- 

 votes many pages to the discussion of this problem. But his conclu- 

 sions, at least in so far as they are concerned with the faunas of the 

 American Richmond and Mohawkian formations, are largely based 

 on erroneous or insufficiently digested data. In consequence the con- 

 clusions are usually at least open to question and in most cases defi- 

 nitely negatived by more competent modern evidence. Unfortu- 

 nately most of the latter evidence is as yet unpublished; and it is 

 impossible to settle the questions involved in the proper classification 

 of the Arctic Ordovician and early Silurian formations before the 

 old data have been either sub.stantiated or corrected and carefully 

 studied in the light of the new evidence. Although much of this 

 work has been done considerably more remains to do before I shall 

 feel ready to record final conclusions. 



Inadequacy of formerly prevailing methods of correlation. — So 

 long as we depended indiscriminatelj^ on predominance of trend of 

 evidence determined by matching entire faunas rather than on pre- 

 cise identification of particular species in both intra- and interpro- 

 vincial correlations, and so long as we followed Suess in explaining 

 the observed evidences of Paleozoic and later displacements of the 

 strandline as essentially eustatic, there really was but little or 

 no chance to achieve definiteness and verity in details in deter- 

 mining the age relations of disconnected formations, whether their 

 separation is ocean wide or relatively limited. I am referring par- 

 ticularly to formations in regions that as a rule are affected differen- 

 tially by the slowly but constantly proceeding undulatory movements 

 of the surface of the lithosphere. 



The rudiments of these revolutionary ideas entered my mind when, 

 nearly 30 years ago, I noted the rather unsatisfactory results attained 

 in the endeavor to correlate the Ordovician formations in America by 

 data obtained during the course of my paleontological work in 

 Minnesota.^* Casting about for some possibly more definite physical 

 means of checking the fossil evidence the rather obvious relation of 

 the processes of diastrophism to the then new Dutton theory of 

 isostasy seemed to offer a promising field for investigation. Soon this 

 promise gave way to greater and since then constantly growing con- 



'^No more conscientious or more thoroughly finished effort to correlate formations by 

 matching entire faunas and also no greater failure to achieve true solutions, particularly 

 as regards the lower faunal horizons, is to be found in geological literature than my 1890 

 attempt to correlate the Ordovician formations in Minnesota with those in Kentucky, 

 Tennessee, and New York. (See Minnesota Geol. Survey, Final Rept., vol. 3, pt. 2, pp. 

 Lxxxiii-cxxii.) What a help some of the ideas here briefly discussed would have 

 been. 



