44 BULLETIN 1059, U. S. DEPARTMENT OF AGRICULTURE. 
investigation it is obviously necessary to use spectroscopic methods. 
Since the growth of a tree requires many years, and even complete 
suppression in the densest forest is seldom accomplished in less than 
two or three years, it is evident that in the forest minute examination 
of every variation in sunlight is unnecessary. An examination cov- 
ering the entire period of the activity which is being studied must, 
however, be obtained. The nearly ideal and still practicable arrange- 
ment would seem to be provision for continuous observation of the 
total energy available as radiation throughout the period of plant 
activity, with sufficiently frequent spectroscopic observation of the 
composition of this energy to establish not only an average quality 
analysis for the whole period but also to show the variations which 
occur from season to season and year to year and their relations to 
the functioning of the plant. 
Unfortunately, spectrum analysis by present common methods does 
not permit an examination of either the ultra-violet or infra-red 
spectra. For this reason it has been suggested that all spectrum 
analyses might be better conducted by means of energy measure- 
ments (e. g. thermal effect) than by optical comparisons. This is an 
almost unexplored field and presents infinite possibilities for the in- 
vestigator who will devise a satisfactory method of measuring the 
energy of all parts of the spectrum under both laboratory and field 
conditions. 
HORIZONTAL AND VERTICAL ExPOSVEES. 
It is perhaps well to point out at this stage that, particularly in 
forest studies, light measurements of whatever kind may be on two 
distinct bases. In forestry the growth of an individual tree is rarely 
spoken of, or even if it is, no practical significance is attached to it, 
because the individual can rarely be separated from the influence of 
other individuals. Forest growth, in any practical sense, must be 
growth (volume increment) per unit of land area. Similarly, if an 
attempt is made to find a relationship between growth and available 
light, it is certain that the energy must be expressed in terms of a unit 
of area inclined at the same slope as the ground. The total energy 
available to the crowns of trees on a northerly exposure of given 
gradient, for example, can not be more than that which would be inci- 
dent upon a plane of exactly the same aspect and gradient. Land 
areas, however, are always measured in terms of their horizontal 
projections. It therefore follows that the measurements must be re- 
duced to horizontal areas, and the simplest means for reduction is to 
expose a given area horizontally for the original determinations. 
Determinations of total energy available for growth, however, 
will rarely be made in ecological studies, which are much more likely 
