, where, as in New
Zealand, it has been quinquennial since 1875 and 1881 respectively. Up
to and including the census of 1901 each state conducted separately its
own inquiries. The scheme of enumeration is based on that of Great
Britain, modified to suit the conditions of a thin and widely scattered
population. The schedules are distributed by enumerators acting under
district supervisors; but it is found impossible to collect the whole
number in a single day, nor does the mobility of the population in the
rural tracts make such expedition necessary. In more than one state the
police are employed as enumerators, but elsewhere, a staff has to be
specially recruited for the purpose. The operations were improved and
facilitated by means of an interstatal conference held before the census
of 1891, at which a standard schedule was adopted and a series of
general tables agreed upon, to be supplemented in greater detail
according to the requirements of each state. The standard schedule, in
addition to the leading facts of sex, age, civil condition, birthplace,
occupation and house-room, includes education and sickness as well as
infirmities, and leaves the return of religious denomination optional
with the householder. Under the head of occupation, the bread-winner is
distinguished from his dependants and is returned as employer, employed,
or working on his own account, as is now the usual practice in
census-taking. Each state issues its own report, in which the returns
are worked up in the detail required for both local administrative
purposes, and for comparison with the corresponding returns for the
neighbouring territory. The reports for New South Wales and Victoria are
especially valuable in their statistical aspect from the analysis they
contain of the vital conditions of a comparatively young community under
modern conditions of progress.

_South Africa._--Almost from the date of their taking possession of the
Cape of Good Hope and its vicinity, the Netherlands East Indian Company
instituted annual returns of population, live-stock and agricultural
produce. The results from 1687 for nearly a century were recorded, but
do not appear to have been more accurate than those subsequently
obtained on the same method by the British government, by whom they were
discontinued in 1856. The information was collected by district
officials, unguided by any general instructions as to form or procedure.
The first synchronous census of