1922] THONE—STARVED ROCK 361 
An examination of the differences between the curves of the sepa- 
rate stations brings to light certain topographical correlations. In 
the first place, there was a general correlation between the exposure 
of a station and its evaporation rate. Thus, station 7, the most 
exposed of the set, shows consistently the highest evaporation 
rates, while station 1, the least exposed, shows consistently the 
lowest. Even between these extremes the correlation holds, for 
of the three talus slope stations, no. 5, the most exposed, shows 
higher rates than nos. 3 and 2, which were partly sheltered within 
the canyon. 
Again, the more exposed stations showed much more marked 
variation, both seasonal and diurnal, than did the more sheltered 
stations. The highest mean daily rate for a ten-day period (26.9 cc. 
per day, July 21-30) exceeded the lowest mean daily rate (8.1 cc. per 
day, September 9-20) by 18.8 cc. This may be compared with the 
excess of the maximum mean daily rate at station 1 (9.2 cc., July 
21-30) over the minimum mean daily rate (1.3 cc., September 9-20), 
which is 7.9 cc. The excess at station 7 is 10.9 cc. greater than the 
excess at station 1, or a ratio of 2.4. Another notable thing is the 
relative stability of the daily rates at the more sheltered stations as 
compared with the greater fluctuation at the more exposed stations. 
Thus, at the two stations and for the two periods noted, the greatest 
mean daily rate at station 7 (26.9 cc.) exceeds the greatest mean 
daily rate at station 1 (9.2 cc.) by 17.7 cc., whereas the least mean 
daily rate at station 7 (8.1 cc.) exceeds the least mean daily rate at 
station 1 (1.3 cc.) by only 6.8cc. The difference during the 
period of highest evaporation rate is 2.6 times as great as the differ- 
ence during the period of lowest evaporation rate. The general 
effect of this factor is strikingly brought out by the “bunching” 
of the curves during June and September, the low evaporation 
periods. This greater variability displays itself even more mark- 
edly, of course, in the diurnal variations than in the variations of 
mean daily rates. During the period of high evaporation just 
cited, for example, the readings at station 7 on three successive 
days (July 14, 15, and 16) were 31.8 cc., 8.9 cc., and 19.1 Cc. respec- 
tively, while on the same days readings at station 1 were 9.8 cc., 
