STRATIGRAPHIC TAXONOMY 637 



parently indistinguishable from C. proliferum, of the Saratogan in JsTew 

 York, beiDg common only in the lower and middle parts. A third form — 

 ledgelike, loosely lamellar in structure, and with mammelons about 2 

 inches in diameter — is most frequently seen about the middle and may 

 be confined to this part. 



Upper division of the Knox. — In east Tennessee and southwestern Vir- 

 ginia to the west of the Athens shale trough, the Copper Eidge chert is 

 usually overlaiQ by a third member of the Knox. Along the lines of the 

 principal barriers it is often absent, presumably having been removed by 

 erosion during the Canadian period. Usually this upper division begins 

 with 100 feet or so of fine-grained, dove-colored, nearly pure or but 

 slightly magnesian limestone, the lower part of which sometimes contains 

 a considerable amount of chert. Above this, from 400 to nearly 1,200 

 feet of fine-grained, mostly yellowish or very light bluish gray magnesian 

 limestone may be found before reaching the base of the Ordovician. 

 Very little of it could with any justice be called a dolomite, much of it 

 is argillaceous and laminar, and many layers fairly pure limestone. Some 

 of the beds are massive, others weather shaly. The latter not infre- 

 quently are highly colored or mottled with purple and green. As a rule, 

 chert is not a conspicuous feature of areas underlain by these rocks. 



Although the lower boundary may not always be drawn at precisely theT 

 same horizon, there is certainly little difficulty in distinguishing this 

 lithologically variable upper formation as a whole from the more uni- 

 formly dolomitic and much more cherty Copper Eidge formation. There 

 are, however, two grave difficulties in the way of determining the exact 

 age of the upper Knox. In the first place, aside from a few Cryptozoon 

 minnesotense ? observed in the faulted Knox band west of Dry Branch, 

 Virginia, the upper member has so far proven entirely unfossiliferoils. 

 The second difficulty is that there are two evidently distinct formations 

 with either of which it may finally be correlated. The first of these is 

 the Eoubidoux-Jeiferson City division of the Ozarkian in Missouri. The 

 second is the 2,000 feet of slightly magnesian and sparingly cherty lime- 

 stone which is fairly well displayed at Jonesboro. Tennessee, and extends 

 northeasterly into Virginia. The latter rests on unquestionable Noli- 

 chucky and has always been described as a lithologically modified repre- 

 sentative of the Knox dolomite. However, in a recent visit to Jonesboro, 

 fossils were procured at several horizons down to within 400 feet of the 

 base of the limestone, which make it reasonably certain that the whole 

 mass is younger than the Knox and probably of Canadian age. Prior to 

 this discovery I was inclined to correlate the Jonesboro limestone with 

 the upper Knox, but according to the evidence now in hand it is thought 



