8< 'i L - IMPLING. 
11 
pling frame used for the first depth is usually 144 square inches (12 by 12 inches) , 
a smaller frame (6 by <> inches) being used for the succeeding depths. * * * 
The iron frame has a stout rim along its upper edge to increase its strength. 
The best sampling frame is made of cast steel: this form of frame needs no rim. 
( Models of both the larger and smaller steel frames used at Rothamsted were on 
this occasion presented to the United States Department of Agriculture.) 
When the soil sampling is carried below the first depth, care must be taken 
when digging around the frame that each depth of soil removed is placed by 
itself, so that when the pit is filled in the soil may be returned to its proper posi- 
tion. A record is kept of the place where the sampling was conducted, as a soil 
can not be accurately sampled twice in the same place. 
Each sample of soil is weighed as soon as it is removed from the frame and is 
put into a bag by itself. When the soil reaches the laboratory it is at once broken 
up by hand into small pieces and laid on paper trays, which are placed on the 
shelves of a storeroom kept at a temperature of About B5 0, till thoroughly dry; 
each sample is then returned to its bag. This immediate drying of the soil at a low 
temperature is essential if changes in the organic matter, and especially nitrifica- 
tion, are to l>e stopped. This practice dates at Rothamsted from 1877. After 
drying the soil it may be stored till leisure is found for further work. Each hag 
is then weighe 1. The soil is crushed and passed through a one- fourth inch sieve; 
the stones that do not pass through this si.-ve are weighed and subse uentiy des- 
cribed) as stones. All that passes through the sieve is thoroughly mixed and a 
sufficient quantity is finely powdered for analysis. Mixed samples are prepared 
after the soil has passed through the one-fourth inch sieve or after it has reached 
the stage of fine powder. 
This method of taking; soil samples, with the exception of the adop- 
tion of the definite temperature above mentioned for drying, lias been 
(as has already been said) in use at Rothamsted since L856, and it 
is interesting here to record thai between -1,000 and 5 9 000 individual 
samples of the Kothamsted soils have been collected in the fashion 
above described. 
The depth of '.» inches fixed for sampling the surface soil is some- 
what greater t ban would now be adopted in the light of the experi- 
ence thai has been gained; although il would now be obviously inex- 
pedient to make any change at Rothamsted, since in that ease future 
samples would not be comparable with earlier ones. But in new soil 
work begun elsewhere, both Sir John I. awes and Sir Henry Gilbert 
came to the conclusion that about s inches would be a fairer depth at 
which to mark off the surface soil from the subsoil. And, since the 
metric system is so widely used in scientific work, the recommendation 
made by them to those initiating experimental work in soil is to lake 
the nearest round number metric approximation, and adopt a cube 
with Ihe linear dimension of 0.2 under as the unit for soil sampling. 
This would give to each layer of soil a depth of 7.874 inches, or 
practically 8 inches. 
1 Representing each plat and each depth. 
