SEED PRODUCTION OF WESTERN WHITE PINE. 3 
sideration in determining the amount of seed production. Thus, 
in measuring the seed crop, three things must be determined: (1) 
The seed production for the stand per unit of area (not for individual 
trees), (2) the quantity of seed, and (3) viability of the seed. The 
weight of germinable seed per unit of area must be accepted as the 
standard for measuring seed crops. If a is the weight of clean air- 
; dried seed obtained from 1 acre, and p is their germination per cent, 
then the seed er op, or «, may be expressed by ae formula 7=ap. 
Since the aim is to déter mine the amount of seed produced per unit 
of area, the best method of studying seed production is by means of 
sample areas. These areas may be from one-quarter of an acre to 
one-half of an acre in extent, in accordance with the density of the 
stand. Each sample area, however, should include at least 100 trees 
of the principal species composing the stand. 
_ Jt would, of course, be more accurate to gather cones and obtain 
_ the seed from all of the trees on the sample area. This, however, 
- would necessitate the cutting down of the trees, which is not always 
_ practicable or possible. Moreover, this operation would require a 
great deal of time, which would make such an investigation difficult. 
For this reason it is preferable to collect the cones and extract the 
seed only from sample trees. 
It is a well-established fact that light is a necessary condition for 
seed production, and observations show that the greater the amount 
_ of light received by the tree the greater is its crown development and 
the amount of seed produced. It may be already accepted a priori 
that individual trees in a stand do not produce equal amounts of seed, 
but vary in accordance with their crown development. In the 
selection of the sample trees, therefore, one must be guided by the 
a form and development of the crown of the individual tree. In the 
different species the different parts of the crown have varying im- 
- portance; thus in Douglas fir the upper part of the crown is of the 
most importance, since it is there that the largest number of cones 
are developed; in other species it may be the extremities of the 
largest lower branches. 
In order that the amount of seed obtained from the sample trees 
_ should, when multiplied by the total number of trees on the sample 
1S plot, Wee represent the amount of seed produced on the plot, 
the sample trees must include representatives of all groups of trees 
_ which differ in any way in their crown development. With this end 
_ in view, all the trees on the sample plot are divided into groups in 
accordance with their crown development, and their diameters 
tallied. From these groups the sample trees are selected. Asa basis 
for dividing the trees into groups in accordance with their crown de- 
yelopment, the ordinary classification into dominant (1), codominant 
