212 The American Geologist. April, i89'.t 
This does not strike me as a serious or a novel difficulty. 
(Geologists are often compelled to work on fragmentary evi- 
dence; they are satisfied if the fragments can be logically built 
up into the complete structure. In most parts of the world, 
rock outcrops occupy less than ten per cent — often less than 
one per cent — of the land surface; yet no field geologist hesi- 
tates to "color in" a formation over an area where scattered 
outcrops give reasonable proof of its occurrence. The sur- 
face area thus colored in is often but a small part of the en- 
tire body of the original formation, which may be largely- 
covered by later deposits or destroyed by erosion; but the 
covered and eroded portions are reasonably inferred, and a 
formation thus established is a stock subject in historical ge- 
ology. It is therefore not so nmch a high percentage of di- 
rect observation, as a logical method of reconstructing the un- 
seen whole from the observed parts that is necessary. Here* 
the dissected peneplain seems to me to stand on a par with 
many other things. Its fragmental condition is most natural: 
its discovered parts are connected and the lost peneplain is re- 
stored by a line of argument that is perfectly reasonable in 
itself, and that is objected to only because it runs counter to 
certain views that are held by Professor Tarr to be estab- 
lished principles in the science of geology. These views will 
be considered below (B i, 2). 
A 4. Certain so-called poic plains are now inclined. Pro- 
fessor Tarr says: — Uplift or tilting "is an assumption ren- 
dered necessary to explain the difference in elevation of the 
supposed peneplain: but i fail to find that there is any evi- 
dence to prove it, unless the peneplain be previously accepted 
as a condition covering thi:; entire region" (p. 358). Here we 
fully agree. I have repeatedly insisted that it was only by 
recognizing the existence of a peneplain that uplift or defor- 
mation could be determined in certain cases; and that only in 
this way could certain stages of geological history be discov- 
ered, in the absence of what might be called orthodox geo- 
logical evidence in the iorm of marine deposits. For ex- 
ample, it is by the remnants of an uplifted, inclined, and 
warped peneplain in the even crest lines of the Pennsylvania 
Appalachians that the post-Cretaceous uplift of the moun- 
tain belt has been determined: it was fvirmerly supposed that 
