1922] THONE—STARVED ROCK 363 
even for this, the survival of relict communities, like the conifers 
on the edges of the cliffs. 
INSOLATION EFFECTS.—An examination of the radio-atmometric 
data serves to emphasize the atmometric phenomena already noted, 
in addition to its main purpose of getting some idea of the sunlight 
intensity as this affects evaporation. The evaporation from the 
black cups of course follows the same general seasonal curve as that 
from the white, the excess varying according to seasonal and local 
conditions. The radio-atmometric effect, as one might expect, 
was greatest during the season of greatest exposure, that is, during 
the month of May, before the leaves were on the trees. After the 
first of June the effect was much less marked. Even during the hot 
weather of July the evaporation rate from the black cups exceeded 
that from the white by but little. 
The general differentiation in the radio-atmometric effect with 
the development of the foliar screen was accentuated by local varia- 
tions. Thus, the tendency throughout the season was for a greater 
difference at station 7, the most exposed of the set, located in an 
open stand of second growth oak. Station 1, located on a treeless 
portion of the canyon bottom, in a stand of Impatiens pallida, 
showed a most notable radio-atmometric effect during May, before 
the surrounding herbage was well grown; after July 1 the effect 
fell almost to zero, and remained so throughout the rest of the season. 
Of course it is not possible to take the radio-atmometric effect as 
a measure of solar radiation in all its effects on plants. It is intended 
only as an approximate determination of direct solar radiation on 
the evaporation of water from a free surface. Taking it as a rough 
index of the total illumination received, however, we find that 
during the season when other ecological factors are favorable for 
the growth of seedlings on the forest floor it is at its maximum, and 
that it falls off notably after the leafing out of the trees. Pre- 
sumably then, sunlight conditions conspire with the ecological 
factors already discussed to make late spring the optimum season 
for seedling growth, and that at places like station 1, where other 
conditions are favorable even when they are most unfavorable 
elsewhere, sunlight intensity falls to an unfavorable level after 
about July 1, or possibly even earlier. It must be emphasized again, 
