298 Yapp. — On Stratification in the Vegetation of a 
vegetation, but chiefly by the fact that in 1908 both these instruments were 
much nearer its upper level. 
In general, the results of the evaporation experiments show that the 
lower strata of the vegetation possess an atmosphere which is continually 
very much more humid than that of the upper strata ; and further, that the 
higher and denser the vegetation, the greater these differences are. 
Owing to evaporation of water from the soil, the frictional resistance of 
the latter to the wind, &c., the layers of air immediately above even bare 
ground are doubtless in general more humid than those at somewhat higher 
levels. But on dry days these differences are in all probability very much 
less than those recorded above for different vegetational layers. It was 
intended to make some experiments on this point during the summer of 
1908, using evaporimeters in the open. Want of time has so far prevented 
these from being carried out. Professor F. W. Oliver has, however, very 
kindly allowed me to make use of some observations made by him in April 
1908, on the salt marshes at Erquy in Brittany. Humidity readings were 
taken in five separate localities, while a wind was blowing. It was found 
that the average relative humidity at the ground-level was about 70 to 73 
per cent., while at five feet above the ground it was about 61 per cent. 
These results tend to support the suggestion made on p. 280 with respect to 
the part played by humidity in determining the proper position for the 
leaves of Hydrocotyle , &c. (cf. Text-fig. 2). 
(c) Air and Soil Temperatures. 
As previously mentioned, unscreened thermometers were used through- 
out. In the case of the air thermometers, readings were taken each time 
the evaporation was measured. The soil temperatures were taken at first 
twice, but subsequently only once, a day. The more important results are 
given (in degrees Centigrade) in the following tables. 
Table III is compiled from the whole of the 1907 observations (series 1 
to 4) ; but owing to defects in the B minimum thermometers during the 
first and last series, results involving the use of B minima are omitted from 
this table. 
Table IV, however, which refers only to the second and third series of 
the 1907 experiments, includes data for all positions. 
Table V gives the results obtained from the 1908 readings. 
The methods by which these results have been calculated will be found 
in the Appendix (p. 31b). Except in the case of the mean daily tempera- 
tures, these methods are much the same as those in general use by 
meteorologists. The positions (A, B, and C) occupied by the different sets 
of air thermometers are explained above (cf. Text-figs. 4 and 6, and PI. I, 
Fig. 2). The soil temperatures (D) were taken six inches below the surface. 
