404 Report on the Experiments on Ensilage 
to be of the character generally termed " sour." Simultaneously 
14 loads of grass, weighing 11 tons 10 cwt. 3 qrs. 15 lbs., had 
been spread out in the meadow on July 2 for haying, and five 
loads more, weighing 3 tons 16 cwt. 2 qrs. 15 lbs., were added 
on July 3, making a total of 15 tons 7 cwt. 2 qrs. 2 lbs. of 
grass for haying. The weather, which had been fine through- 
out, continued so up to July 5, when the hay was made and 
weighed. There were then 5 tons 4 cwt. 0 qr. 1 lb. of hay, 
which were made into a rick. 
The details of the construction, size, &c.,of the silos have been 
given in a former paper,* and it will be sufficient here to say 
that the silo used was one of six formed out of an old barn 
by the building up of brick division walls, the sides of the silo 
and the floor being thoroughly well cemented. The silo was 
6 ft. 2^ in. wide, 20 ft. 6 in. in horizontal depth, and 16 feet 
high. The front was bricked to a height of about 7 feet, the 
material being filled in from carts over the top of this front. 
Above this could be fixed wooden boards when required. 
The pressure was given by stones in strong elm boxes, and 
amounted to 112 lbs. per square foot. The silo was unprovided 
with any drain. Temperatures were taken in different parts at 
depths of 1, 2, and 4 feet from the surface. The height of the 
grass immediately after the filling was completed was 12 ft. 6 in. 
On opening the silo on December 16, 1886, the contents were 
found to have sunk to 7 ft. 6 in., only reaching to the top of the 
first board above the brick frontage. 
The question of how to sample the silage was one of the 
greatest importance. Unless this sampling be done with the 
utmost care, altogether fallacious results are almost sure to be 
obtained. My experiences of the year before had shown me 
only too clearly that even by drawing a large number of samples 
from different portions of the mass of silage, an experimenter 
might readily fail to get accurate information as to the composi- 
tion of the entire contents of a silo, and of the loss consequent 
on ensiling. I therefore adopted this year the following plan 
for ensuring that the samples drawn should be really represen- 
tative ones of the whole bulk : — 
The silo was just over 20 feet long, and so I divided off the contents into 
four blocks, each about 5 feet in length ; as the middle of the block was being 
reached, a section, or rather " boring," of the whole block was cut out from top 
to bottom. This boring was collected in a bag and carefully but rapidly mixed 
together on a sheet, the quantity redut•^d by frequent division, and ultimately 
a representative sample drawn, which was taken for analj'sis. 
From July 3 until December 16 (when the silo was 
* ' Jomnal,' vol. xxii. p. 484. 
