94 
president’s ADDRESS — SECTION C. 
(7.) The biological evidence points to the probability of a general 
refrigeration of the land and sea in Permo-Carboniferous time, as 
suggested by (a) the great break between the floras of the Carbon- 
iferous and Permo-Carboniferous periods in Australia; (b) the entire 
absence of beds of pure limestone from the marine beds of the Permo- 
Carboniferous strata south of lat. 31° 8. 
(8.) Nature of the Permo-Carboniferous Glaciation . — As regards 
the question as to whether all the glacial phenomena in the Permo- 
Carboniferous rocks are referable to (a) floating ice alone ; ( b ) to land 
ice alone ; or ( c ) to the combined action of land ice and floating ice, the 
(a) and (c) hypotheses will be discussed first, and ( b ) last. As regards 
(a) the following arguments may be adduced in f favour of all the 
glacial phenomena being referable to floating ice alone : — • 
(i.) Although the grooves are deeply cut in places in the pave- 
ment at Bacchus Marsh, there is a slight want of uniformity in their 
trend, some sets of grooves crossing others so as to enclose a small 
angle at their intersection. It also appeared to me that the grooves 
were not quite so firmly nor sharply cut as those made by some land 
glaciers. 
(ii.) No evidence has as yet been observed of any true moraines, 
either terminal, medial, or lateral. 
(iii.) All the included boulders observed by me were either 
rounded or subangular, whereas angular fragments would, perhaps, 
have been plentiful if the material had been deposited by land ice. 
(iv.) While the groundmass of the rock is largely of local origin, 
many of the erratics are quite foreign to the district, whereas had they 
been brought by land ice the large boulders would have been of local 
origin, unless the glaciers were over 100 miles or so in length. 
(v.) The perfect stratification of the glacial beds implies aqueous 
deposition. This is most noticeable at ilallett’s Cove, where the 
glaciated surface of Pre-Cambrian rock is immediately capped by finely 
laminated sandy shales belonging to the glacial beds. The tough 
mudstones containing the boulders are not distinctly stratified in places, 
but occur in well-marked beds, from 10 to about 30 feet thick, with 
great regularity, being separated from one another by widely 
distributed beds of conglomerate and fine sandstone. 
(vi.) The intensely worn characters of the sand grains in the 
groundmass of the boulder beds at Ilallett’s Cove is very suggestive 
of the sands being of marine origin. 
(vii.) The indenting of the beds on which the large erratics 
repose is proof of the latter having been dropped from some height, 
and having fallen, probably through water, until they struck the 
bottom. 
(viii.) The manner in which the conglomerates have been 
squeezed down in the form of pockets, in places, into the boulder beds 
is suggestive of the action of floating ice, but might, of course, have 
been accomplished by land ice. 
(ix.) During the great submergence which affected a large portion 
of East Australia and Tasmania, in Permo-Carboniferous time, the area 
now occupied by the glaciated pavements would also probably have 
been submerged. It is singular that in Australia, South Africa, and 
India the pavements are less than 1,000 feet above the sea, and at 
