364 PROCEEDINGS OF SECTION C. 
Having examined the glacial beds which rest directly on the 
silurian and granite in the Bacchus Marsh district, one of us, Mr, 
Brittlebank, has been unable to detect any beds in which the 
stratification could not be made out. There are, however, beds of 
varying thickness. These appear to be unstratified, but they are 
generally intercalated with beds of fine silt and sandstone, with 
which they form an even junction. 
IV. RE-ADVANCE OF ICE SHEET. 
From sections given herewith, it will be noted that boulder beds 
occur at certain horizons, intercalated with beds of sandstone, 
silt, and conglomerate. As these might represent a return of a cold 
period or possibly a re-advance of an ice sheet, careful search was 
made along the junction lines for ploughing, contortion, injection, 
striated pavements, and other evidence of the passage of ice over 
the underlying glacial beds. With the exception of slight local 
contortion, no evidence has yet been observed by one of us (Mr. 
Brittlebank) which would support the theory of the individual 
glacial beds being due to successive ice sheets, alternately advancing 
and retreating. As, however, it has been shown that glacial ice 
can creep over incoherent beds without contorting them, this 
evidence cannot be held to be conclusive. . 
V. CONTOURS OF OLDER ROCKS AT THE TIME OF THE GLACIAL 
PERIOD. 
As these naturally correspond with those of the present striated 
pavements observed in section in the Bacchus Marsh and surround- 
ing districts, we have been able to obtain a fairly accurate idea of 
the old surface at the time of the glacial period, perhaps towards 
its close. 
From the figures marked on the map some idea of the irregular 
surface may be gleaned [Plates XVII and XVIII]. We had 
hoped that from a careful study of this surface, and of the 
direction of striz, we should have been able to express an opinion 
as to whether land or floating ice had formed the striated pave- 
ments and overlying deposits. From the conflicting evidence it 
would, perhaps, be premature at present to arrive at any but 
provisional conclusions. 
The examination of the matrix of the Bacchus Marsh glacial beds, commenced by one 
of us, Professor David, has had to be postponed through his absence on the Funafuti Coral- 
boring Expedition. So far the examination for radiolaria and foraminifera has yielded 
negative results, though some obscure casts are slightly suggestive of the former having 
once been present, but this is very doubtful. Minute fragments of plants are, on the other 
hand, very numerous, and fairly well preserved. Minute chips of black shale, lithologically 
very like the local graptolite shales, are present in the matrix; and at Myrniong Creek, 
near Dunbar, the material forming the matrix of the glacial beds may very well have been 
derived from granite very finely pulverised. The comparative scarcity of large fragments 
of local rocks is nevertheless very remarkable, as already stated. No remains of diatoms 
have as yet been observed. 
