TRANSACTIONS OP SECTION E. 489 



After the President's Address (see p. 478) the following Papers were read : — 



3. Geographical Considerations arising out of the Visit to Australia in 

 1914: a Comparison of Vegetation Maps. By 0. J. R. Howarth, 

 M.A. 



We have accustomed ourselves to think of the continent of Australia in 

 terms of maps on very small scales. The use of such maps is justified, in part, 

 by the fact that, speaking broadly, Australia is a continent singularly simple in 

 outline and form. But the employment of small scales makes it the more 

 important, on the one hand, that precision, where precision is possible, should 

 be adhered to, and, on the other hand, that where precise data do not exist, 

 I'recision should not be assumed. 



Three maps are shown : on all of them the same colouring is used to show 

 different types of natural vegetation, yet the appearance of each differs widely 

 from the others. The maps, chosen practically at hazard, are labelled A, B, and 

 0, and are enlarged from small maps to be found respectively in a standard 

 authoritative work on plant geography, a well-known school-book, and a well- 

 known atlas. In copying the representations of the various types of vegetation 

 were standardised. Some of the main differences between the maps are here 

 summarised : 



(1) Monsoon Forest of the North. — Map A shows a continuous belt of it east- 

 ward from York Sound ; B a continuous belt only east of the Victoria River ; 

 C shows the northern forest in isolated patches. A shows Cape York peninsula 

 as wholly forested, B nearly so, C along the east coast only. 



(2) Forests of the Eastern and Soutli-Eastern Highlands. — A and B show a 

 forest belt right along the eastern seaward slope; C interrupts this belt in the 

 vicinity of Townsville (Queensland), bringing the grassland of the interior down 

 to the coast. In New South Wales A gives the temperate forest and woodland 

 an extreme width of 200 miles ; C gives it fully 400 ; B gives an intermediate 

 width. A, distinguishing 'parklands, woods, and meadows' from 'forest,' 

 allows the State of Victoria no true forest area large enough to be shown. B 

 divides the sonth-eastern forest climatically (summer and winter rain forests) 

 approximately along the political frontier between New South Wales and 

 Victoria; C shows a division west of Melbourne — this distinction, however, is 

 made merely between 'broad-leaved forest' and 'evergreen trees.' 



B shows the scrub area, interrupting the forest in the lee of the Flinders 

 Range (Victoria, South Australia), which A and C ignore. A remarkable 

 divergence of opinion is seen respecting the forested area in the south-west of 

 Western Australia. 



Opinions may naturally vary as to what exactly constitutes forest, woodland, 

 parkland, steppe, grassland, scrub or semi-desert, and desert. But, as regards 

 qrassland, C allows a much smaller total area in Australia than A or B. B 

 recognises a strip of steppe bordering the Australian Bight, which A and C 

 ignore, though it is insisted upon in various official and other publications, 

 especially in connection with the trans-continental railway now building. B and 

 C distinguish desert from semi-desert or scrub, bnt their distinctions are at 

 variance, C showing a rather elaborate outline of the true desert areas, which 

 would suggest that complete data existed for laying them down. A (perhaps 

 wisely from the general educational standpoint) applies the term ' desert ' more 

 widely, to cover all these conditions. 



It is in the north of Western Australia that the three maps are perhaps 

 most completely at variance. 



Making every allowance for divergence of opinion as to the connotation 

 of terms and for erratic draughtsmanship, it would seem that the vegetation 

 divisions of Australia have been allowed to get rather out of hand. Probably 

 the same is true of vegetation maps of this sort, and of other distribution maps, 

 for other parts of the world. Though the present may not be the time to under- 

 take such work, it is suggested that the scratiny and criticissm of existing 

 material (whether original maps or verbal description) for the compilation of 

 certain classes of distribution maps might be entrusted to a small committee 



