420 BOTANICAL GAZETTE [DECEMBER 
experiments are supposed to be open, in the day condition, through- 
out. I shall also limit my considerations to short periods of time, 
at least to short periods in strong light, thus aiming at an avoid- 
ance of the problem of incipient wilting as above set forth. The 
former of these problems has been touched upon already (see Science 
N.S. 29:269-270. 1909), and the full consideration of it should 
make another title. The second problem, of incipient wilting, 
cannot be experimentally considered at the present time. 
The specific problem which now holds our attention is, then, 
Is it possible by any simple means to estimate quantitatively the 
various light intensities to which plants in the open are subjected 
and so to sum the effects of these as to be able approximately to 
calculate the variations in transpiration thus brought about, and 
the total transpiration for the longer period in which these varia- 
tions occur? It is obvious that the solution of such problems as 
this is of the utmost importance in establishing a basis for a scien- 
tific agriculture. Also, such problems lie at the bottom of con- 
siderations of the factors determining plant distribution, and it 
must be through their solution that ecology may at length emerge 
from the descriptive and classificational stage in which, for the 
most part, it now finds itself. The importance of our present 
inquiry is exceeded only by its difficulty. 
Apparatus and methods 
To attack the problem outlined above it was necessary to meas- 
ure the water loss from experimental plants under different light 
intensities, and to compare the various rates thus obtained with 
readings taken, for the same periods and exposures, upon whatever 
physical instruments were available for the estimation of light 
intensity. For the plants, the ordinary method of weighing potted 
and sealed specimens was resorted to. A number of different 
instruments for determining light intensity were tested. I shall 
proceed first to a discussion of these instruments. : 
Since the intensity of solar radiation varies from time to time, 
even for short periods, as on a partly cloudy day, our great desider- 
atum in the present connection is an instrument or method for 
automatically obtaining an integration of this factor for a given 
