Jiivei- Vnlhnis of the Ozark rUitcmi. — // < rs/ici/. 355 
As farther north, the streams have done hut little more work 
since the Columbia epoch than to clear out their ancient 
channels (which they have not 3'et (juite accomplished), and 
to begin an attack on the solid rock in ])laces, though they 
have as yet made comparatively little impression on it. 
The i)resent altitude of the Ozark region is, I believe,, 
rather above, than below the normal level. Many of the 
streams abound in rapids, and even some low falls occur that 
show the streams to be still at work cutting down the l)ottonis 
of their beds. In dry weather some small streams disapi)ear 
entirely and flow below the surface in some cases through 
crevices and caves in the bed rock of the valley. It is also to 
be noted that many of the springs flowing from caves have, 
within a comparatively recent time, cut rapidly to a lower 
level, and emerge at other places than formerly. Of course, 
it is the habit of springs to change their place of emergence,, 
but the recent change in the Ozarks is of such a nature as to 
indicate, I believe, a slight elevation of the region. All the 
larger caves are Tertiary in age, and are generally quite dry. 
At a lower level, and much less in size, occur the C^uarternary 
caves. Still lower, and generally well filled by the streams 
flowing in them, are the caves which uom- are in tin' ])rocess 
of fonjuition. 
ChKONOLOGY indicated I'.V TIIK OzAHK \ Al.l-KYS. 
studies of the changes, and of the time recjuired for each, 
in the drift-covered regions, have given a probable mean for 
post-Columbia time of 50,000 years. Those who have studied 
the question from phenomena occurring in the coastal plain 
region want a longer time; but in the Ozarks this time seems 
sufficiently long, and even more than is necessary, to account 
for the work which has been done. In the drift regions of 
the upper Mississippi valley, the time above given is the mean 
of all estimates made by the writer; and 1 shall consider it as 
the most probable approximate length ol' ])ost-('olunibia tini(\ 
It is sonu'what difficult to compare the amount oferosionMn 
the Ozark valleys since tlie Lafayette period with that since 
the Columbia epoch; but, making due allowaiu'e for a tliU'er- 
ence in hardness of nuiterial excavated, and also for a prolta- 
bly greater erosive power liefore than since the ('olunil>ia 
epocli. due partly to dill'erence in grailient. climate and a 
