18 BULLETIN 1056, U. S. DEPARTMENT OF AGRICULTURE. 
The best position for the stenciled number is across the top of the 
bag. In this position it will be visible when the bags are standing 
on end or when stacked or piled. All bags bearing the same lot 
number should be kept together and placed or stacked in the same 
section of the warehouse. 
STACKING. 
In stacking, it will be found advisable to alternate the position of 
the bags in each laj T er by placing a double row end to end in the first 
layer, and in the second or succeeding layer placing them parallel to 
each other and crosswise or at right angles to those in the first or 
preceding layer. This provides a stack or pile the width of the 
length of a bag and any desired height and length that the dimen- 
sions of the warehouse will permit. A space of at least 6 inches 
should be left between each stack or pile and between the bags and 
the walls. Such arrangement permits a free circulation of air and 
reduces greatly the possibility of heating. 
GERMINATION. 
All other factors being equal, the value of a given quantity of seed 
is in direct proportion to the percentage that will grow. Eegardless 
of this fact, a widespread belief exists that the making of germina- 
tion tests of planting cotton seed is of minor or secondary importance. 
This belief is based largely on two factors: (1) That seed from 
mature cotton harvested under favorable conditions and stored prop- 
erly will germinate satisfactorily, and (2) that the heavy rate of 
seeding employed bj^ most farmers will give a good stand of plants 
although a large percentage of the seed may not grow. These fac- 
tors are true to a limited extent. However, two lots of seed may be 
handled and stored in exactly the same manner and yet there may 
be a wide difference in the percentage of germination. Also the 
farmer who buys planting cotton seed is entitled to know, so far as 
practicable, what percentage may be expected to grow in order that 
he may effect a corresponding saving in the cost of seeding by ad- 
justing accordingly the quantity to be sown per acre. 
The results of germination tests made by the Seed Laboratory, 
Bureau of Plant Industry, of samples submitted to the Bureau of 
Markets and Crop Estimates show that there is a wide variation in 
the viability of cotton seed sold for planting purposes. This varia- 
tion occurs not only between the samples submitted by different 
dealers but also between the percentage of germinable seed in each 
of two or more samples submitted by the same dealer. For example, 
of the 1919 crop, one of five samples of as many varieties submitted 
by a dealer tested onl} T 56 per cent, while the other four ranged from 
80 to 93 per cent: and of the 1920 crop, one of four samples sub- 
mitted by another dealer tested only 59 per cent, the other three 88, 
88, and 94.5 per cent. 
With further reference to the 1919 crop, the percentage of germi- 
nation of the 70 samples tested ranged from 45 to 96 per cent, aver- 
