Physiography of Werrihee Arm. 287 



may flow right across the lava tongue, as has happened in the case 

 of Turritable and Willi mi gongong Creeks at Mt. Macedon. The 

 ultimate result must of course be the total destruction of the lava 

 tongue. Avith the possible formation of one stream again, but such 

 a stream should itself have a characteristic shape that would indi- 

 cate its mode of origin. Possibly none of the post-newer basaltic 

 streams in Victoria have yet reached that stage. 



Goodman's Creek has a fairly st^ep and even grade; it 

 flows mainly in Ordovician, parallel to and below the northern con- 

 tinuation of the Rowsley fault. A long, high scarp of tree-covered 

 Ordovician hills bounds the western side of Goodman's Creek, while 

 the Bullengarook basalt limits the eastern side of its valley. A 

 large number of important tributaries come down from the western 

 ranges; a few very steep gullies enter on the east. The lowest of tlie- 

 three western tributaries shown in Plate XII. B flows, in its W.-E. por- 

 tion, along the Coimadai fault, between Ordovician and glaciaL 

 For the remaining two miles of its course, to its junction with the 

 Lerderderg, Goodman's Creek lies in glacial rocks, in a wide val- 

 ley, with gentle grades. 



(h) Pyrete Creek. — This is also known as Coimadai Creek. It 

 rises in the deeply ravined block of high Ordovician east of Mt. 

 Bullengarook. The tributaries that lie in these ranges are pre- 

 newer basaltic in age, with the possible exception of subsequent 

 minor captures, and no doubt belonged to the ancient Bullenga- 

 rook Creek. The main part of the river is, however, post-newer 

 basaltic, as already described. A vigorous attack is being made 

 on the lava tongue where it bends westward, and the old river 

 gravels are here exposed in three places; where seen these gravels 

 are iron-cemented and very coarse. The extensive *' bight " in 

 the Bullengarook basalt tongue (Fig. 22) does not represent the 

 original outline of the flow at that point, but is mainly due to 

 the erosive work of the Pyrete Creek. Tlie river is in Ordovician 

 as far soutli as the village of Coimadai ; here it crosses a small 

 fault (see Fig. 5), and enters on an area of glacial sandstones and 

 conglomerates, and tertiary limestones a,nd clays. It occasionally 

 cuts through these to the Ordovician bedrock beneath, exposing in 

 places the beautifully striated permo-carlx)niferous glacial pave- 

 ments descriljed by Officer and Hogg (ref. 4*^), etc. Foi- the last 

 few miles the river has cut through the newei- basalt and tertiaries,. 

 exposing low-level Ordovician slates. It finally enters the Bacchus 

 Marsh basin and flows across the alluvial flats to the Werribee. 

 The grade of this creek, as plotted in Plate XII A., show.s tiaces of 



