'280 T. S. Hart: 



the process of denudation. But to form such a series of parallel 

 valleys, so close together, merely iis consequent strc.-ims, se<'ms to 

 require a very regular original surface. Small irregularities can 

 scarcely be negligible with valleys s(» cU)se to one another. Therefore 

 it is probable that the valleys have been from the first guided by lines 

 of easy erosion. Actual inequalities of surface to guide them need not 

 here l)e required. 



Further north, however, we begin to liave evidence of a more 

 important south-west fall of the country. The Notting Hill-Chel- 

 tenhani axis itself falls to the south-west j so does the high land 

 from East Camberwell to Malvern. The tributaries from the noi-th 

 are necessarily more important. Yet they have not been aVjle to 

 continue south-west. Either Elsternwick Creek and Gardiner's 

 Creek are on lines of weakness which gave them an inunediate 

 advantage, or there was originally an irregularity in the south- 

 west fall. There is actually no icason why the original elevated 

 surface should not possess i-egularly arranged inequalities. 

 Ordinarily streams have been able to do so much work that 

 minor features have been lost, but heie we are dealing with the 

 weakest of streams. It must be remembered, however, that all 

 these creeks had to start as weak stieams in the upper l)eds of the 

 tertiaries. It may have needed no more to determine the courses 

 of Gardiner's Creek and the Elstei-nwick Ci-eek than those of the 

 Brighton valleys. It may bi' noticed, liowever, that .Malvern Hill 

 is a little higher than we would expect if the i-idge is falling 

 uniformly to the south-west. ;iml that the cori-esi)ondence between 

 Gaidiner's Creek and the Sjiringvale valley on the opposite fall, 

 favours the idea of an original stiuctural character or inetpiality. 

 The I'llsternwick Creek system is a further develojiment of a system 

 like tiie Brighton valleys with the addition of the inHuence of the 

 soutli-west fall as soon as tlie main axis is left. In (iardiner's 

 Creek and its tributaries these cliaradt-rs are fuitlier devel(q)ed. 

 If the tertiaries north of Gardiner's Creek are mostly terrestrial, 

 and those to the south mostly marine, this would involve an 

 original slope of deposition in the terrestrial part and original 

 streams before the emergence of tlie southern area, but these streams 

 would all be in shallow valleys. 



Area south of Carneyie and M n rrumbeena. — If .Main (^reek ever 

 continued to the south-west it must have been across this locality. 

 Tlie comparatively low levels of the ridges, somewhat lower than 

 the hills near Caulfield stittion. and also than the high land to the 

 .south-east, at first suggest a broad valley. But the detail of the 

 surface is not easily e.xplained ori this supposition. By e.xaiuiuing 



