49 
The Census o/1891 and Bural Depopulation. 
an influence on correctness of description. But it is probably 
safe to assume that the amount of error in tbe last and the present 
census is not materially different, and that the figures, whatever 
intrinsic error they contain, are reliable for purposes of com- 
parison. 
It is also possible, as the Registrar-General shows, to coun- 
teract the causes of error to a certain extent by selecting for 
special examination certain counties which are purely agricul- 
tural, and where all the labour is, with exceptions which may be 
neglected, of an agricultural character. Taking, for instance, as 
examples of such counties, the East Riding of Yorkshire (exclud- 
ing Hull), Lincolnshire, Norfolk (excluding Norwich), Suffolk, 
Cambridgeshire, Bedfordshire, Hertfordshire, Oxfordshire, Wilt- 
shire, Dorsetshire, Devonshire (excluding Plymouth), Hereford- 
shire, Brecknockshire, and Cardiganshire, the Registrar-General 
finds that “ the total number of men and women returned in these 
counties, either as agricultural or general labourers, together 
with the shepherds, and the carters, carriers, and hauliers of all 
kinds, amounted in 1881 to 380,161, while in 1891 it was only 
354,972." The decline was therefore a “ fraction less than 7 
per cent.” 
The Registrar-General mentions three main causes to which 
this decline appears to be due. The first is the attraction of 
the towns, with the prospect of higher wages and the certainty 
of a more varied life; the second is the natural effort of the 
farmer to diminish his labour bill ; and the third and last is the 
conversion of arable land into pasture. If we consult the Agri- 
cultural Returns, this last influence would certainly seem to be a 
vera causa) for in 1891 those returns showed a diminution of 
1,074,077 acres of arable land in England and Wales as com- 
pared with 1881, while the increase of pasture beyond this 
difference, although reaching a total of some 552,234 acres, 
seems to have been largely nominal rather than real. 1 And 
yet, in spite of the reality of the cause, it is not easy to trace 
the precise connexion between it and its effect in the figures of 
the census. Mr. Druce, in an article 2 in this Journal on the 
census of 1881, failed to discover any definite relation between the 
decrease of labourers and the substitution of permanent pasture 
for arable land, although such a relation was discoverable between 
1 A part may be “ traced to an amended classification of certain areas 
reckoned for the first time as coming within the category of permanent 
pasture rather than that of unenclosed mountain and heath land." 
2 The Alteration in the Distribution of the Agricultural Population of 
England and Wales between the Returns of the Census of 1871 and 1881, by 
S. B. L. Druce. — Journal R.A.S.E., Second Series, Vol. XXI., 1885, pp. 96 et seg. 
VOL. V. T. S. — 17 JE 
