508 J. T. JaUon: 



in diameter to those 20 or 30 yards long. Like the rapids 

 they are rare below Templestowe, and this fact immediately 

 suggests a connection between these small islands and the 

 nature of the valley in Avhich they occur. In addition, they 

 are found to be almost constantly associated with the rapids, 

 and observation proves that this is not merely a coincidence, 

 but that the cause of the rapid is also the cause of the island. 



The hard rocks which form the rapids tend to project above 

 the softer rocks, and also, at low water, above the level of the 

 stream. These projections on their up-stream side arrest por- 

 tions of the gravel sand and silt carried along by the stream. 

 Boughs and trunks of trees floating down the I'iver are also 

 at times caught by the same obstacles. 



The rock projections, assisted by the stranded trunks and 

 boughs, when present, cause silting up in a small way, and a 

 miniature island may be formed. ^ Vegetation of some kind will 

 soon spnng up, and if a strong plant such as a young shrub or 

 tussocky grass takes root, it will increase the resisting power 

 of the embryo island. Should the conditions continue favour- 

 able, the island increases in size and vegetation becomes 

 stronger, until trees of moderate height, shrubs, and dense 

 undergrowth often cover the soil and bind it firmly together. 



The island thus becomes able to withstand the river, even 

 in the winter, and becomes a definite feature of the stream. 

 Islands in all stages of growth, from the single tussock of 

 coarse wiry grass growing in the river, to the large island with 

 tall trees and dense undergrowth, can be seen in the portion 

 of the Yarra now referred to. 



The formation of the island of course divides the stream into 

 two parts, with the result that when passing the island its 

 volume, in proportion to its space, is increased, and conse- 

 quently its velocity also. The vertical erosive power of the 

 stream is thereby accelerated, and the exposed parts of the 

 rapids are more cpiickly planed down than they would be 

 without the island. At length a temporary or permanent 

 grade is reached, and vertical erosion ceases. Lateral erosion 



1 Small islands can of course be formed by obstacles in places other than at rapids, hut 

 the latter would appear generally the most favourable position and certainly are so in the 

 part of the Yarra here dealt with. 



