FIELD INVESTIGATIONS IN 1 92 1 839 



numerous. 20 The Pentremites also are of small forms and few in num- 

 ber. Lyropora has not been observed. 



It is in the oolitic layers of the Upper Gasper that the recurrent Sper- 

 gen fauna is mainly to be found. A long list of these species is given in 

 my 1917 report. Careful study of these Gasper specimens shows that 

 most of them had changed very little from their ancestors in the Lower 

 Sainte Genevieve and Spergen oolites. They are therefore of only lim- 

 ited value in determining precisely the age of the rock in which they arc 

 found. As a rule, however, the several occurrences are accompanied by 

 more decisively characteristic fossils. Among the latter the calyces of 

 species of Pentremites have proved the most serviceable and reliable in 

 my work. The long-based forms with flat interumbulacral areas, which 

 commonly occur in the Gasper oolites, are never found in the older oolites. 



The lithologic character and faunal peculiarities of the locally dis- 

 tributed upper shaly beds of the Gasper in western Kentucky have already 

 been mentioned a few pages back in discussing the Paint Creek formation 

 of Illinois, which it is quite evident corresponds at least essentially to 

 the Upper Gasper of Kentucky. 



The sections in southeastern Kentucky and northern Alabama. — Two 

 other sections shown on the chart remain to be considered. These are 

 much alike in essential features, although the one gives the facts as 

 known in southeastern Kentucky and the other shows the sequence of 

 beds found 200 miles to the southwest, in the vicinity of Huntsville, 

 Alabama. In both sections the Cypress sandstone and the succeeding 

 Golconda shale, which overlie the Gasper limestone, are thin, and either 

 one or both may be wholly wanting or at least unrecognized in the com- 

 monly narrow interval between the Gasper and the unmistakably charac- 

 terized overlying Glen Dean limestone. Here and there in the interven- 

 ing area of Tennessee a similarly thin representation of the Hardinsburg 

 sandstone, which belongs between the Golconda and the Glen Dean, is 

 recognized. 



The main reasons for bringing these two sections into the present dis- 

 cussion pertain to the contents of the Sainte Genevieve part and its rela- 

 tions to the overlying Gasper limestone. Regarding the Gasper itself we 

 need to say only that in both sections the Gasper contains no median 

 shale or sandstone member, so that the formation constitutes an unbroken 

 sequence of limestone ledges. In general, however, there is no difficulty 

 in distinguishing the lower half or third from the remaining upper part 



20 Mr. Butts reports finding north of Cerulean a layer of limestone in the Lower Gas- 

 per in which specimens of this genus are abundant. As none were collected, we have 

 no means now of determining the specific relations of these remains. 



