26 
May, 1927 
The Queensland Naturalist. 
in one movement, and without tilting or folding to some 
extent. More probably lias denudation gone on, while 
elevation or tilting has been slowly taking place. Prob- 
ably seldom has this so-called cycle of erosion been com- 
pleted without interference by renewed earth movement. 
Moreover, there was in the above story the quite unjustifi- 
able assumption that the land was of uniform structure, 
whereas we know that various rock types resist very 
differently the action of stream or weather, and that 
geological structure must modify the process consider- 
ably. Nevertheless the ideal stream gives us a ' very 
useful picture of the process taking place, with more or 
less modification in every valley. 
Particularly from it we must realise how permanent 
is the position of a stream owing to the rapidity with 
which, aided by its load of debris it can cut down its bed. 
Only a geologically rapid change of levels can seriously 
divert a stream, a change of levels more rapid than the 
stream’s power of erosion. 
If as a general rule streams maintain more or less 
their original position, what are the circumstances which 
determine that position? 
We can only surmise what these might be, but it is 
obvious that any chance development of a stream just 
by eroding back its head at the expense of the stream on 
the other side of the divide must yield an irregular 
arrangement of the main stream and its tributaries. 
Where there is a marked parallelism in neighbouring 
streams or a regular course, we may be reasonably certain 
that here there was no chance development of erosion, 
but that the position of the stream had been determined 
by some previously existing circumstance. For instance 
a uniform general tilting of a flat surface might give 
rise to roughly parallel streams, as also might the even 
surface of a lava flood, which would have a regular 
declivity away from the focus of eruption. A gentk 
warping of a flat surface into parallel ridges and hollows 
would also produce parallel streams. 
Apart from the original position of a stream it is 
generally recognised that occasionally, owing to some 
favourable circumstance the head of one stream may cut 
back and capture part of another stream, thus diverting 
the water into another channel. 
Having now referred to the essential principles of 
river development, in-so-far as they affect our enquiry, we 
are in a position to consider a few peculiarities of our 
