THE ORIGIN OF THE FAUNA AND FLORA OF NEW ZEALAND. 261 
and is no doubt a fragment of a set of beds which once filled all 
this part of the valley. At prescnt it is restricted to a patch 
occupying a sheltered side valley on the south side of the river, 
but it again appears at the river bed as an apparently detached 
fragment separated from the main mass in the valley by river 
gravels. This isolated portion is known as Castle Rock. Now 
the first thing to be noticed is that this oligocene limestone de- 
scends to below the present level of the river, proving con- 
clusively that the Rakaia is now running ata higher level than 
it did in the eocene period before the limestone was deposited. 
The second thing to be noticed is that the junction, up the side 
valley, between the limestone and,the palzozoic rocks on which it 
rests, must mark the limit of the Rakaia valiey at the time when 
the limestone was deposited. If therefore any great lateral 
denudation had taken place since that time, the line of junction 
ought to stand out as a prominence. But on the contrary it is 
in a valley, apparently much in the same position with regard 
to the other parts of the valley as when the limestone was 
formed. Consequently no great plateau on the south of the 
Rakaia can have been removed. 
Many other instances could be cited, but this one must 
suffice, for it alone is sufficient proof that the denudation which 
has taken place during the comparatively short time that has 
elapsed since the commencement of our last great glacier epoch 
cannot have affected the shape of the mountains to such an 
extent as to make it worth while to take this cause into con- 
sideration, even if it acted in the direction supposed. That the 
large river valleys were more or less filled to a height of 3000 or 
4000 feet above the present sea level by tertiary rocks, most of 
which have been since removed, is no doubt true. But as this 
is below the line of perpetual snow—which is estimated by Dr. 
von Haast and Mr. M‘Kerrow to be between 7000 and 8000 feet 
—this filling up of the valleys, if it affected the level’of the snow- 
line at all, would raise it by radiation in the same way that the 
plateau of Thibet raises the height of the snow-line on the 
northern slopes of the Himalaya. 
As therefore both the subsidence and the plateau hypotheses 
are quite untenable, we must fall back on elevation of the land 
as the main, if not the only cause of the former extension of our 
glaciers; and it is strongly confirmatory of this hypothesis that 
the two earlier glacier epochs each occurred at a time when we 
have independent proof that the land stood at a far greater 
height than at present. With regard to the last glacier epoch it 
has been estimated that an elevation of between 3000 and 4000 
feet would be quite sufficient to bring back the glaciers to their 
former dimensions.{ 
But if our last glacier epoch was caused by elevation of the 
land, it is easy to prove that it must be of an clder date than 
the glacial epoch of Europe, because while our islands are sepa- 
+ Trans, N. Z. Institute, XIII, p. 385. 
