Marsh , and its Relations to Evaporation , &c. 291 
Of course the ratios varied somewhat on different days, e. g. on July 11 
they were ioo; 34*1: 7-7, and on July 21, 100 : 37: 14*1. But on the 
whole, the differences for the three positions are not only very considerable, 
but also moderately constant. 
Treating the third and fourth series (the total time being again between 
seven and eight days ; in August and September) in the same way, position 
A being now clear of the vegetation, the totals are 
A. 409-1 c.c. 
B. 114-65 c.c. 
C. 12-9 c.c. 
And the corresponding ratios : — 
A : B : C = 100 : 28 : 3-2. 
Thus the differences of evaporation for the three positions are, on the 
whole, even greater than in the first two series. This is no doubt due in 
part to the raising of the evaporimeter A. In addition to this, however, it 
must be noted that while the actual evaporation from A is higher, that from 
C is lower than in the first two series. B is almost exactly the same in 
both cases. 
Combining the whole of the 1907 observations, the ratios work out at : 
A : B : C = 100 : 32-8 : 6-6 
That is, for the vegetation in question, the promotion of transpiration 
by the atmosphere, as regards shoots which project above the general 
vegetation level, is something like three times as great as the effect pro- 
duced in the middle layer of the vegetation. Further, the average efifect on 
these projecting shoots is some fifteen times, and on the shoots of the 
middle layer about five times, as great as that experienced by plants 
occupying the lower layers of the vegetation. 
August 15, 1907, is of especial interest, on account of the high evapora- 
tion from the exposed evaporimeter A. The water-loss from this instrument 
was, on this occasion, the greatest recorded (for twenty-four hours) during 
either year. 1 That from B and C was, however, not exceptional (cf. Table I). 
Further, the wind-velocity was greater on this day than on any other during 
the field observations. Readings were taken at intervals of two hours 
through most of the day. Text-fig. 5 shows the curves of evaporation and 
temperature on this occasion. July 20 and 21 are added for comparison, and 
the soil temperatures are also included. 
It may be noted that during the four hours from 10.30 a.m. to 2.30 p.m. 
the loss of water from A averaged nearly 9 c.c. per hour, which again was 
the highest rate for either year (cf. curves in Text-figs. 5, 7 and 8). For 
the same four hours the total evaporation from C was only o-i.c.c. During the 
1 This was 86-5 c.c. (see Table I), or, in linear evaporation, 0-169 inch; which is almost 
exactly one-third of the highest evaporation (for 24 hours) recorded by Livingston during his desert 
experiments (see Appendix, p. 316). 
X 
