OF MT. KINABALU AND BRITISH NORTH BORNEO. 43. 
I saw no trace of the blood-red Rhododendron, which could not have been in 
flower without being noticed. Finally, over another granite wall, we struck 
the long inclined slope to the granite plateau of the actual summit, marked 
in the accompanying shaded map (Pl. 2) which Captain Learmonth has 
most kindly allowed me to reproduce, and in which he has embodied the very 
important results of his survey. 
The following plants were limited entirely to the cracks and were clipped 
like little box-edges, more, as I have explained before, by insolation and 
radiation from the rock surfaee than by wind, whieh only aecounts for the 
reduced height. The decumbent branches ran along the soil which filled 
the cracks, sending up little erect aerial shoots reduced to 0°25 m. in height. 
Styphelia Learmonthiana and S, suaveolens were both in flower, while 
Leptospermum recurvum, mostly still in bud, was very predominant, and I 
can well believe that it clothes the summit like a fall of snow, when out, 
as described by Low (2. 13) in March. 
Rhododendron ericoides was in flower and also very plentiful, with the 
tips of its acicular leaves converted into callous white points, while Kurya 
reticulata, in fruit only, showed a conspicuous callous white border round 
the leathery leaves; Diplycosia kinabaluensis was not modified in size or 
habit. The sight of Drapetes ericoides, in typical ericoid tufts about 1 dm. 
high, covered with minute white heads of flowers, and of Halorayis micrantha, 
forcibly recalled New Zealand mountain conditions, which were accentuated 
still more by Centrolepis philippinensis and a little Hrigeron-like composite, 
unfortunately only in bud and therefore indeterminable, but which Mr. Moore 
considers to bea Lagenophora. Knowing this would be an interesting record, 
I hunted all along the cracks for another specimen, but only found one more 
plant in the same condition. Other herbaceous plants were Didiscus sanicule- 
folius var. rupicola, Phyllocrater Gibbsie, now prostrate, the two Potentillas, 
previously cited, in rosettes with no stolons, and the minute P. parvula. 
Gentiana lycopodioides grew in erect tufts, as well as Muphrasia borneensis, 
while Haloragis micrantha and Pilea Johniana were both abundant, spreading 
under and round the other plants, but all again strictly limited to the soil- 
surface of the cracks. Scirpus Clarkei was very distinct, extending for 
some distance by itself ; Schwnus kinabaluensis, Carex hypsophila, C. rara, 
Deschampsia flexuosa var. rigida, and Poa epileuca, the latter as abundant 
as the Scirpus, competed for the little available space, making up the 
tule of the ghostly buffalo grasses, on which the phantom herds of the 
spirit dwellers browse. I was surprised to come across several plants of 
Blechnum capense, another new record for the summit. In passing up this 
incline, a soft rock with white crystals spreading over the surface of the 
granite, and so rotten that a piece broke off easily, is referred to by 
Dr. Cullis as a lava biotite-augite andesite. Farther on, under the lee of 
