182 PROCEEDrNGS OF THE NATIONAL MUf^EVM. vol. 41. 



perplexed by the reference to the Keokuk horizon of many forms, 

 inchiding several genera which had never been seen in that position 

 in the typical region, but which would have seemed perfectly at home 

 in the Lower Burlington. These reputed occurrences seriously inter- 

 fered with the logical sequence of several specialized genera, whose 

 position was otherwise perfectly clear. After a while I came to sus- 

 pect that the real trouble lay in the correlation of the Knobstone beds, 

 and I concluded to undertake a reexamination of that region, with a 

 view of determining the actual horizon of its fossils. In this I was 

 encouraged by Doctor Weller of the University of Chicago, who 

 expressed the opinion, based upon his own extensive studies of the 

 Kmderhook, that the Knobstone fauna was more likely to prove 

 Kinderhook than Keokuk. 



Accordingly, in 1909, 1 sent the veteran collector, Frederick Braun, 

 into the field to make a careful and systematic collection from the so- 

 called Knobstone beds from the northern lunit of the knobs in Indi- 

 ana, to and including the famous Button-mould Knob region of Jeffer- 

 son and Bullit counties, Kentucky, and thence as far south, along the 

 Devono-Carboniferous outcrop, as time would permit. He spent the 

 entire season at this work, covering the ground thoroughly from Wash- 

 ington and Clark counties, Indiana, to Marion County, in central 

 Kentucky. Good specimens of crinoids have always been rare at the 

 Knobs, but Braun was instructed to collect minutely all fragments 

 and detached plates. In this way he accumulated a large quantity 

 of material from the weathered slopes of the talus — the so-called 

 Button-mould washes — and from ravines and fields formed by 

 their erosion. The crinoids are chiefly derived from the disintegra- 

 tion of thin limestone bands interbedded in the argillaceous shales 

 and marls, and are usually highly sihcified, often pure flint. It is 

 rarely possible to find them in place, and therefore we can not desig- 

 nate the exact layers from which they came; but it is usually practi- 

 cable to fix their approxunate limits, and to say whether they were 

 derived from the main body of the Knob, or only from siliceous lime- 

 stones at the upper part of the section. 



This material, when assembled, showed clearly — 



1 . That there were no characteristically Keokuk forms in the debris 

 of the Knobs in southern Indiana, or those of the Button-mould Knob 

 region; but farther south m Marion County, where well-defined heavy 

 beds of siliceous limestone cap the hills, true Keokuk species, such as 

 Dorycrinus gouldi, Adinocrinus jugosus, Agaricocrinus americanus, 

 began to appear, more or less intermingled with the others, but their 

 origin was evident. 



2. That the whole fauna, except the last-mentioned occurrences, 

 is of unmistakably Lower Burlington type. 



