RESEARCH METHODS IX STUDY OF FOREST ENVIRONMENT. 43 
definite and measurable. As M. R.adau (72) says: "The red rays and the yel- 
low rays in certain cases continue the work commenced by the violet rays, and 
in others undo what the last have accomplished. Thus, chloride of silver, 
slightly impressed by the violet rays, is then blackened under the action of all 
of the visible rays; and guaiacum, turned blue by the violet rays, is bleached 
by the more luminous rays. It follows that the chemical action of light is, in 
general, very complex, and that it can be used for measuring the energy of 
solar rays only with much circumspection." 
The inevitable conclusion is that direct photochemical methods can 
not be made to solve the problems of ecology, but this does not elimi- 
nate spectrophotochemical measurements, which may, in fact, give 
the best possible criteria as to the variations in the different spectra 1 
regions and the effect of such variations on plants. 
4. The fourth point to be considered in approaching the study 
of possible methods for radiation measurements, is the difficulty of 
securing complete absorption of sunlight. While lampblack is popu- 
larly conceived to absorb rays of all wave lengths and to transform 
them into measurable heat, recent investigations have proved that 
this is only approximately true, and have shown the existence of 
an infra-red spectrum of extreme wave length to which lampblack 
is partially transparent. Fortunately, this region is relatively unim- 
portant as a source of energy and may be, for biological purposes, 
almost wholly unimportant. 
A greater source of error than that arising from the failure of 
lampblack to absorb the radiation is undoubtedly the loss, as heat 
radiation, and by conduction and convection, before the heat can be 
properly measured. It must, of course, be borne in mind that tem- 
perature is not a measure of heat, and that the indications of a ther- 
mometer can not be used except as the radiation rate of the thermom- 
eter itself has been thoroughly studied. 
With this conception of the nature of sunlight and the difficulties 
in the way of its proper measurement, it is perfectly evident that 
the primitive methods that have been employed in measuring light 
in the forest do not give satisfactory results. Two distinct but sup- 
plemental lines of attack suggest themselves as being profitable : 
1. The growing of trees under controlled conditions of light, using 
both artificial lights of known composition and monochromatic and 
other screens wmich will transmit to the plants only certain wave 
lengths, as suggested in the work which has been started by Mac- 
Dougal (68) : The physiological action of each wave length must, of 
course, be studied. Through such study it is hoped that the require- 
ments for light may be determined, any actual deficiency in sun- 
light which exists in the forest recognized, and the effective supply 
measured. 
2. In studying either the light conditions as they exist in the 
forest or the effective supply suggested by the preceding line of 
