The Census of 1891 and Rural Depopulation. 53 
has been no actual diminution of the rural population as a 
whole, but a growth which is relatively less rapid. Viewing 
the matter, then, in this comprehensive manner, it is erroneous 
to speak of “ depopulation ” of the rural districts, although the 
phenomenon may be found in certain districts. Proceeding by 
one of the two rough standards of measurement indicated before, 
and adding to the population of the rural sanitary districts, 
first, that of urban districts of less than 10,000, we find that 
twelve English and eight Welsh counties exhibit a decrease 
between the two censuses of 1881 aud 1891. The eight Welsh 
counties coincide with those enumerated before by the Registrar- 
General (vide Table III., p. 56), and the largest decrease occurs 
in the cases of Montgomeryshire, Cardiganshire, Radnorshire, 
Flintshire, Merionethshire, and the border counties of Hereford- 
shire and Shropshire. Of the other English counties, Hunting- 
donshire and Rutlandshire are so small as to be by comparison 
insignificant ; the loss in others, such as Norfolk and Suffolk, 
has been vexy slight, amounting to less than 1 per cent., and the 
only counties of importance whichj have sustained a notable 
decrease ai'e Lincolnshire, the North and East Ridings of York- 
shire, Cornwall, Bedfordshire, and Wiltshire. 
It will be interesting to compai’e these l’esults with those 
obtained by Dr. Longstaff in a paper read before the Royal Sta- 
tistical Society in June last. 1 Dr. Longstaff adopted a different 
mode of distinguishing the rural population. He urged that 
counties were delusive as standards of measixrement. The 
growth of the town of Cambridge, for example, concealed the 
depopulation of the rural parts of the county, and the case was 
similar with Poole in Dorsetshire, with Yarmouth and Norwich 
in Norfolk, and with New Swindon and Salisbury in Wiltshire. 
The method, accordingly, which he adopted was to take 
all the registration districts in each registration county which 
had exhibited a decrease of population in either of the last two 
deceixnia, to exclude all the districts comprising towns of con- 
siderable size, and, lumping the population of the districts so 
selected for each county, to regard this as its rural population. 
While, therefore, the Registrar-General, in the figures we have 
already cited, proceeds by the methed of inclusion, Dr. Longstaff 
has followed a method of exclusion. The net loss of rural popu- 
lation thus ascei’tained was, he shows, pi’actically the same, both 
absolutely and relatively, between the census of 1871 and that of 
1881, and between this and that of 1891. It amounted to some 
160,000 persons, or about 3 per cent. 
* Journal of the Royal Statistical Society, Vol. LVI. Part III. p. 380. 
