Davis and Wood.] 
374 
[Nov. 20, 
no sympathy with the underlying rock* structure and must have 
been produced independently of it. 
The only way in which an even surface, that is discordant with 
the rock-structure, can be explained is to regard it as the form to 
which all land-masses are ultimately reduced by denudation, from 
whatever initial form they may have had, provided only that the 
denuding forces have had time enough to act ; and this is true 
whether the ultimate form is the subaerial baselevel plain of land 
denudation, or the subaqueous platform of marine denudation. 
Given sufficient time, the wasting away of the land under the air 
or at the edge of the sea will result in a low, even surface ; and in 
one or the other or both of these ways, we may hope to explain the 
peneplain of the Highlands. 
It is worth while to recognize clearly that this explanation rests 
on the principle of exclusion ; a principle that must be employed 
cautiously. Some process adequate for the production of a certain 
old plain is searched for ; several processes are suggested ; some of 
them are manifestly insufficient; then the true explanation must be, 
sought among the remaining processes. The postulate of this ar- 
gument is that all the processes by which plains can be produced 
have been considered ; otherwise the true explanation might escape 
us. But it may also be recognized that nearly' all geological argu- 
ment is of this character, and that it becomes safer as the progress 
of the science makes it unlikely that any adequate process has been 
omitted from the discussion. The reader must judge whether it 
is safe, in the present state of geology, to conclude absolutely that 
plains can be formed only in the four ways above named ; if he can 
suggest other ways, the conclusion of the preceding paragraph is 
invalid. 
But if it may be safely’ concluded, as we think it may, that the 
old peneplain of the Highlands was either a subaerial baselevel 
plain or a submarine platform, it follows that the land must have 
stood lower than it now does when the lost material was worn away. 
The baselevel of that time must have been B' L/, fig. 2. The present 
upland must then have been a lowland, and the opportunity for 
cutting the present valleys must be ascribed to a general elevation 
at a later date. This will be returned to farther on. 
8. Difference between plains of subaerial and submarine origin. 
It remains for us to inquire whether the old Highland peneplain 
was the product of subaerial or submarine processes. Thirty years 
