120 BROOKLYN BOTANIC GARDEN MEMOIRS 
ing soil temperature, air temperature, humidity, number of hours of 
sunshine, wind velocity, precipitation, evaporation, etc. 
The fundamental problem, however, presents itself after the analy- 
sis has been made of the elements which enter into the compound 
“weather.’’ It is the experimental determination of the effect of 
those elements, singly and collectively, as measured by the data 
compiled, upon the physiological activities of the plant under con- 
sideration. This effect can only be measured by means of accurately 
conducted experiments in which very expensive apparatus is used. 
One of our problems is so to outline the work and to set forth its 
fundamental importance that those in authority will be moved to 
purchase the ecological equipment without which these agricultural 
and silvicultural problems cannot be studied. 
In the attempt to solve the problem above outlined some workers 
have used a “‘plant instrument.’ A given kind of plant has been 
grown by the side of atmometers, etc., at stations established under 
different climatic conditions and an attempt has been made to interpret 
their effect as registered by the “plant instruments.’”’ As an illustra- 
tion of the attempt to interpret meteorological data in terms of plant 
development let us take the work dealing with temperature. One 
method contemplates the subtraction of a constant from the tempera- 
tures recorded and considers that thermometric degrees in excess of 
this constant are available for purposes of plant development. A 
second method seeks to express growth-rate in terms of the velocities 
of chemical reactions. A third—the physiological method—attempts 
to take into account the optimum and maximum temperatures as 
related to plant growth, and the attempt has been made to develop 
one formula which will express the combined effect of rainfall, evapora- 
tion and temperature on plant growth. This represents but little 
more than an attempt to show what might be done if we had sufficient 
experimental data on the reaction of plants to the complex conditions 
known as the weather. 
Much of this work has been based on averages—averages for a 
month, a year, or a number of years. We read that a large amount 
of data assists in ‘‘smoothing out the curve” or that the “spas- 
modically jerky graph may be smoothed.’ It is certainly true that 
in some cases the curve should not be smoothed out, because it is the 
spasmodic graph that shows sudden changes and the extremes. The 
burden of this paper is to show that, in some cases at least, averages 
for long periods are of little value as compared to the importance of 
the data obtained for certain critical periods in the conditions of the 
environment as shown by the ‘‘spasmodic graph.” Data collected 
for a short period in the summer may be very important, but are by 
