242 GESTEKAL GEOLOGY. 



area occupied by the Middle coal-measures. That boundary 

 may be described approximately as follows : 



Draw a line from the northwest corner of Harrison county 

 almost directly to the middle of the northern boundary of 

 Madison county, thence to the northwest corner of Lucas 

 county, thence to the middle of the northern boundary of 

 Wayne county, thence to Centerville, and thence to the south 

 boundary of the State, along the west side of the valley 

 of Chariton river. 



The western and southern limits in Iowa of the Upper 

 coal-measures are the western and southern boundaries of the 

 State, but the formation extends without interruption far into 

 the States of Missouri, Nebraska, and Kansas. 



It has been previously shown that with the beginning of the 

 Lower coal-measure epo<fti, there was an abrupt change from 

 the deposition of calcareous rocks to those of sandy and shaly 

 composition, with comparatively little calcareous material; 

 which material continued to be deposited throughout the whole 

 epoch. Also, that although there was an accumulation of a 

 somewhat increased amount of calcareous material in the 

 Middle coal-measures, it is yet characterized principally by 

 argillaceous and sandy shales. But when we come to the 

 Upper coal-measures, we find the formation in Iowa to be as 

 fully characterized hy limestones as any of the formations of the 

 Sub-carboniferous group are. Indeed, the St. Louis limestone 

 which in Iow T a immediately underlies the Lower coal-measures, 

 and that of the Upper coal-measures are so similar in aspect 

 and lithological characters that they have not unfrequently 

 been confounded together. This mistake has not been made 

 by the ignorant alone, but even geologists of acknowledged 

 ability have regarded the two formations as one; or, at least, 

 have mistaken the Upper coal-measure limestone for Sub- 

 carboniferous limestone in spite of the well marked difference 

 in the palseontological characters of each. The error of the 

 former is doubtless due alone to the similarity of aspect and 

 lithological characters; but that of the latter seems to have 

 originated principally in the acknowledged similarity and 



