122 F. E. Fritsch and E. J. Salisbury. 
It is striking that the Vaccium over most of the Hindhead heath is 
of very low growth, and this is probably to be related to the frequent 
fires which prevent any considerable accumulation of peat. 
Estimations of the humus-content have not been made on a 
sufficient scale to draw exact inferences as to the amount of organic 
matter in successive years after a fire. It is clear, however, that at 
first the humus-content is small (on an average 7-3% of the dry 
weight) and there does not seem to be any rapid increase until 
several years have elapsed. On the other hand, in the oldest parts 
of the heath (CU) the humus-content is from two to three times as 
great, reaching a maximum of about 30% where Vactinium is 
dominant. The regions with poor Calluna (p. 132) have a relatively 
low humus-content, and it is significant in this connection that the 
amount of organic matter on the slopes of recently burnt areas is 
lower than on the level. 
As a result of the increase of humus there is a marked increase 
of water-content; thus after a week’s heavy rain it was found that, 
whereas on burnt areas the water-content was 11-19% of the dry 
weight, in the tall Calluna it was 31-39%, and where Vactinium 
becomes dominant nearly 70%. 
B.— The Effects of Exposure. 
(I) Evaporation-data. 
Attention has often been called to the uniform level of the 
vegetation of heaths, a feature which is exhibited in a more or less 
marked degree during every phase of its development (cf. p. 129 etseq.) ) 
but most pronouncedly in the final stage which we have termed tall 
Calluna (CU facies). This is due to the dying away of most of the 
projecting tips which, owing to their isolation, are subjected to 
extreme exposure (3, p. 153). A measure of the extent to which 
mutual shelter is afforded by the vegetation, when growing up en 
masse, is furnished by estimations of the difference in the rate of 
evaporation as computed at succesive levels above the soil-surface. 
All our evaporation-data have been obtained in the following 
way. A number of porous pots of about 60 cc. capacity were filled 
with boiled water to a known level; their open ends were plugged 
with corks, previously impregnated with paraffin-wax and pierced by 
a short length of glass tubing whose outer end was sharply bent 
downwards as a fine capillary tube. The pots were fixed in vertical 
series to upright posts by means of wire attachments. The drop in 
level of the contained water was estimated at frequent intervals, the 
water being at the same time restored to its original level. 
