28 The American Geologist. January, 1901. 
fifteen miles wide, trending east to west, bowed slightly along 
a central line, but otherwise remarkably even in surface. The 
erosion of valleys 500 to 1,000 feet in depth has pretty thor- 
oughly cut up this plateau into flat-topped ridges, although 
there are undissected tracts of 300 or 400 acres as level as any 
plain. The origin of these flats is difficult of explanation 
except on the theory that they are remnants of an ancient 
baselevel of erosion, a peneplain. Nearly all the ridges reach 
this peneplain level. 
The Boston mountain is monadnocked upon the Tertiary 
peneplain to the extent of about 500 feet vertical. Hence, the 
dissected peneplain to its summit is an oldr one. From the 
general correspondence in bight between the Ouachita moun- 
tains and the Boston mountain, it appears evident that the 
same peneplain may be represented in both. Hence, T feel 
safe in identifying the Cretaceous peneplain north of the Ar- 
kansas river, at a maximum altitude of about 2,250 feet near 
Winslow and 500 feet above the main Tertiary peneplain. 
Northward from the Boston mountain, the Cretaceous 
peneplain is represented by isolated outliers of the main 
plateau — flat-topped peaks, sometimes elongated into ridges — - 
in other words, by a series of monadnocks standing on the 
Tertiary peneplain. These are mainly of Coal Measure shales 
and sandstones and some might consider them as due to struct- 
ural rather than physiographic conditions. However, I am 
confident that many of them (especially those whose summits 
are truncated) are remnants of the Cretaceous peneplain, so 
well represented on the Boston mountain. This gradually 
descends toward the north and approaches the main Tertiary 
peneplain ; near Hindsville, in Madison county, Ark., there is 
an interval of only a few hundred feet between them, and near 
Eureka Springs, several monadnocks forming small groups 
widely separated from each other and far distant from the 
main system near Boston mountain, seem to indicate that here 
the Cretaceous peneplain has descended to within 100 feet of 
the lower baselevel. 
Over the Ozark plateau region of southern ^Missouri, it 
is doubtful if any hill can be positively identified as a remnant 
of the Cretaceous peneplain. There are a few low monad- 
nocks in Stone and Barry counties, which seem to belong to 
