118 
IOWA ACADEMY OF SCIENCE 
lower gray (post-Kansan) loess extends far beyond the limits of the Iowan 
border, and the upper yellow member (chiefly post-Iowan) also clearly ex- 
tends into the territory designated as forming the “southern loess” area. 
McGee ascribed the formation of the paha and river-ridge, together with its 
cap or covering of loess, to melting ice,* and his explanation may be all sufficient 
so far as concerns the drift core, but the inclusion of the loess is unfortunate, 
as the ice-theory postulates conditions under which the mollusks, which are 
more or less abundantly represented in the loess, could not have existed. 
Since the publication of McGee’s great paper, the paha has received frequent 
mention, especially by Calvin, Norton, and Savage, in the Reports of the Iowa 
Geological Survey. 
The most detailed of these references are those of Norton,** who, notably in 
the Bremer county report, attempts to And support for the fluvio-lacustrine 
hypothesis of loess-formation in the loess of the paha. 
In general Norton’s observations and conclusions, so far as they concern the 
structure of the paha, and the position and origin of its loeSs-covering, are sub- 
stantially the same as those recorded by McGee, though they are somewhat 
more extended and detailed. 
In the Bremer county report he observes that the problem of the paha “re- 
solves itself into two parts, that of the origin of the nucleus of till and of similar 
pahoid drift hills and ridges on which loess is absent, and second, the origin of 
the loess cap.” 
With the first part of this problem this paper has little to do, excepting as it 
bears on the second part. However, attention should be called to the fact that 
the drift core of the paha, when present, is Kansan, and in nowise different 
from the mass of Kansan drift found elsewhere, and that it represents only the 
higher parts of the Kansan sheet, portions of which have been covered with thin 
Iowan, leaving these paha ridges mere islands near the border of the Iowan 
area.*** 
But the identification of the paha as detached ridges or knobs has not always 
been accurate. It must be borne in mind that large islands of Kansan drift are 
found within the Iowan border in a number of counties. McGee reported! 
such Kansan islands from Floyd, Mitchell, Chickasaw, Jones and Bremer coun- 
ties; Calvin from Chickasaw! and Buchanan§ counties; Savage from Benton 
county; 1 1 and Arey found a similar area in Black HawM and another in Butler 
county.® 
All these areas, now for the most part well known to the students of surface 
geology in Iowa, are bordered by outlying isolated knobs and ridges which are 
true paha. ^ 
"^Ibid., pp. 254-5, andp. 572, et seq. 
** Especially those in the county reports of the Iowa Geological Survey, on Linn county, 
vol. VI, 1895; Cedar county, vol. XI, 1901, and Bremer county, vol. XVI, 1906. 
^^^This fact has been recognized by Norton himself, among others, in the Rep. la. Geol. 
Sur. , vol. XI, 1901, p. 356, who says that “the paha are so closely associated with the Kansan 
on this border that they may be treated as detached portions of it, even when entirely sur- 
rounded by the younger drift. ’ ’ 
tibid. , pp. 402 and 450. 
fla. Geol. Sur., vol. XIII, 1902, p. 326. 
§Ibid., vol. VIII, 1898, p. 246. 
II Iowa Geol. Sur., vol. XV, 1905, pp. 139-145. 
Tllowa Geol. Sur., vol. XVI, 1906, p. 445. 
° Not yet published . 
