OF MT. KINABALU AND BRITISH NORTH BORNEO. 47 
&. China sea, but in this case (B. N. Borneo) their direction follows that of 
the coast-line, and generally speaking the wind is from the N.E., November- 
April, and from the S.W., May—October.” The former period covers the 
autumn fruiting-season in Australia and the S. Pacific. On the higher 
parts of Kinabalu a southern type of vegetation no doubt predominates, and 
though we cannot estimate the stages in which it may have reached the 
mountain, such a constant factor as seasonal winds is a matter for considera- 
tion in relation to the occurrence of heterogeneous elements in isolated 
situations. 
Of summit-plants not known from lower altitudes on the mountain itself, 
the Australian elements Drapetes ericoides and Lagenophora occur in New 
Guinea at 8000—10,000', Didiscus saniculejolius is common to New Guinea 
and the Philippines, and Centrolepis philippinensis is so far known from the 
latter only, while Haloragis micrantha spreads from New Zealand to the 
Himalaya. The three Potentillas are a northern hemisphere element 
occurring in the Himalaya and the mountains of New Guinea ; Deschampsia 
flexuosa and Agrostis canina are temperate cosmopolitan types, whilst Blechnum 
capense is widely spread in the tropics and subtropies. 
The only endemic species limited to the extreme summit are, so far as we 
know at present, Pilea Johniana, Symplocos buxifolia and Coprosma Iookert. 
Should the incidence of mossy forest on Kinabalu be found limited to the 
sandstone, then the question of the development of this peculiar botanical 
formation, which previous observations in Fiji had already led me to suppose 
must be edaphic in character, would be easy to explain. In the sandstone 
some form of * ortstein " or pan may possibly have formed at a certain depth, 
which by impeding the free passage of water would bring about acid 
conditions of the surface-soil, inhibiting the action of animals and fungi, 
and consequently all metamorphie processes. An invasion of moss-growth 
would result, whieh by absorbing and holding water would aet like a 
sponge, checking surface aeration, analogous to the development of humus 
in northern beech-forests, as demonstrated by Müller (Studien über die 
natürlichen Humus-Formen und deren Einwirkung auf Vegetations-Béden, 
Berlin, 1887). On the strength of this hypothesis, the mossy forest must 
be secondary in character. 
In my own limited experience I have never come across any fungoid 
growth in this formation, and I should imagine that earthworms were also 
out of the question. Epiphytes and undergrowth are generally limited, 
producing an impression of sterility, but occasionally both may be more 
pronounced, as is the case in the association above Kamburangau (Pl. 6. 
fig. 4). The older the formation, the more prostrate logs encumber the 
ground. Sometimes the trees may be dwarfed, but as a rule the forest 
consists of a thick growth of slender unb 'anched trees, as described from 
