that the first method would give the best results, but this is 
not necessarily the case. Growers do not always weigh or 
measure their crops, and in any case do not usually do so till they 
come to be sold. Thus a farmer may keep wheat in the rick and 
not sell off the whole of his crop for a year or sometimes longer, 
some crops, such as barley, oats, roots or hay may be consumed 
on the farm, and so far as this is the case many farmers keep no 
accurate records of quantities. In short, though farmers may 
make some estimates for their own information, either of their 
total yield or of the part they expect to sell, it is not possible to 
rely upon them in general for exact or complete information. If 
every producer were willing and able to supply a fair estimate of 
the output of his crops, the average result would, no doubt, be 
more accurate than the estimates of a limited number of expert 
estimators. The difficulty of obtaining a statement from the 
producer has, however, made it necessary in practice to adopt 
the method of estimating as a means of obtaining annual informa- 
tion as to the production of crops. In England and Wales the 
Ministry has the assistance of some 320 Crop Reporters, for the 
most part land agents, valuers, and others with considerable 
local experience, who estimate the yield per acre of the principal 
crops, parish by parish. Each parish estimate is multiplied by 
the acreage of the crop in the parish as returned by occupiers on 
the Agricultural Schedule and in this way the total production 
in the parish and hence in the country is obtained. As there are 
roughly 13,000 parishes, the final estimate of production is the 
result of practically the same number of separate estimates based 
on local knowledge and weighted in proportion to the acreage. 
These final estimates are made after the crop is harvested, when 
actual results are known in many cases. 
The Crop Reporter does not, of course, rely merely on his 
own observation in making these parish estimates, but supple- 
ments this by local enquiry from growers, and in the case of 
corn, from threshing machine proprietors. It is, however, the 
duty of the Crop Reporter to determine on the basis of the 
information thus obtained what can fairly be regarded as the 
average of the parish. It will be seen that, given reasonable 
skill on the part of the Reporters, the system should result in a 
high degree of statistical accuracy, and as regards the crops which 
are estimated annually this method has been adopted for the 
census of 1925. 
The annual estimates made in this way cover wheat, barley, 
oats, mixed corn, beans, peas, roots, potatoes, hay, sugar beet 
and hops. There remain, however, many important items, 
such as milk and dairy produce, meat, wool, poultry and eggs, 
fruit and vegetables, which present much greater difficulty. 
To supply these missing data is the main object of the 
Agricultural Census, and the method adopted so far as practicable 
has been to obtain from producers returns of output. The system 
varies with different products, but it has not been found possible 
in any case to obtain such complete information from producers