TJ(€ Geolo<jii of Coimaidal. 191 



in water ; we are almost entirely in the dark as to whether this 

 was marine or fresh ; the little evidence furnished by the vegetable 

 remains found in the Bacchus Marsh sandstones, and the total 

 absence of any marine fossils point to a fresh-water lake. In this 

 water were laid down the stratified mudstones and sandstones. 

 We will first concern ourselves with tlie mudstones which form 

 the lowest beds of the series in our district. From all glaciers 

 terminating on the land sub-glacial sti'eams issue, the water of 

 these streams being derived from the melting of the glacier at its 

 contact with the subjacent rocky bed, and at its upper surface 

 exposed to the heat of the sun and atmosphere. That similar 

 streams issue from a glacier whose front terminates in water there 

 can be no doubt, as the same causes of melting are present in 

 both cases. There has been but little opportunity afforded of 

 studying the structure of beds formed from the sediment of 

 sub-glacial streams discharging into a sea or lake, but reference 

 will he made later on to tlie evidence brought to light in this 

 matter at the Malaspina glacier. If a sub-glacial stream enter a 

 sea or lake it will deposit the matter in suspension just as an 

 ordinary stream does, that is to say, in stratified beds ; this 

 simple view of the matter does not account for the included 

 stones in the mudstones. As stated before, we require that 

 tlie glacier from which the sub-glacial streams issue should 

 protrude into the lake or sea into which the streams empty. The 

 existence of icebergs proves the possibility of a glacier advancing 

 into the ocean, and the magnitude of icebergs shows that in cases 

 the distance advanced through by the ice front before the berg is 

 broken off is very considerable. Sir J. Ross records in his 

 " Voyage of Discovery and Research in the Southern and 

 Antartic Regions," that in many places the glaciers projected 

 several miles into the sea. In investigating the physical relations 

 of a glacier advancing into the sea, let us picture a mass of 

 glacier ice, more or less loaded with included material, moving 

 slowly down a moderate slope ; the ice front would be pierced in 

 places by the channels thi'ough which the sub-glacial streams 

 issue ; the lower part of the ice front would first reach the water, 

 and would therefore be first subjected to melting by its contact 

 with the water ; as the forward movement continues the process 

 of melting would I'esult in the formation of an overhanging cliff of 



