66 Queensland's plant associations 



in the favour of one of them, there results a strong struggle 

 between them. They do not enter into a special association 

 with the representatives of both types (as it is usual in 

 other countries with forest associations), but each one tries 

 to get the upper hand. Therefore we cannot expect a 

 mixture of types and slow transitions from one forest type 

 into the other. 



2. But all these reasons Avould not be sufficient, if 

 one factor were not here, and that is the regular hush fire, 

 which kills all the scrub plants springing up on the border 

 of the forest. 



3. Last we must deal with so-called " forest pockets,^'' 

 in which the phenomenon mentioned above attains its 

 highest evolution. In some few cases there is no difficulty 

 in explaining these forest pockets. On the slopes of Grant 

 Hill, near Cape Grafton, in the North of Queensland, is 

 for instance a system of creeks running mostly through 

 open forest, but accompanied by a dense narrow belt of 

 vine-scrub. In some places, however, the creeks are close 

 together, and besides they often change their course, as is 

 usual on the granite hills and mountains in this country. 

 The result of it is that a bigger area is sufficiently watered, 

 and this enables the vine-scrub to occupy the whole ground. 

 But in one case the bordering creek was too distant from 

 the next current. In the middle of this part, which was 

 evidently drier, was a forest pocket of nearly circular shape 

 enclosed on all sides by dense vine-scrub. 



But there are in the Atherton scrubs forest pockets 

 of a circular shape, some of them not broader than one mile ; 

 the line of demarcation is here most distinct. It is very 

 hard to explain this most extraordinary contact between 

 forest and scrub. My opinion is that there are again two 

 different reasons :■ — 



a. Sometimes the stratum of basalt is in some places 

 very thin, and the soil consequently poor, insufficient for 

 a vine-scrub. 



b. In other cases I regard the forest pockets as a re- 

 mainder of a former bigger open forest, which has been 

 obliged to give way to the scrub. But on very small areas, 

 where the scrub land had not so great possibilities, there 

 remained forest pockets, which, I think, would disappear 



