By reference to the figures given above, a comparison of the individual unions with one another can be readily made. Thus it will be seen that if the cost for salaries, etc., in Whitechapel, since 1870, has risen seriously, in Islington the advance has been still more considerable. The comparison in this particular stands thus: the Whitechapel union since 1870 shows an increase of 256 per cent., while Islington, for the same period, shows a rise of 356 per cent.,-the correction on the ground of alteration in population having been made in each case.

So far for the light thrown upon the question by the statistics of the Metropolitan unions; let us now see whether the evidence furnished by the comparison with one another of some provincial unions points in the like direction.

1 I omit Brixworth from among these for the reason that, since 1890, its system has been altered, and it is no longer representative of the Class A. Also Atcham, because since 1870 its area has been extended and its population so materially increased as to make a comparison of the two periods impossible.

Result of the comparison of the two periods:

Cost of in-maintenance diminished

14.9 per cent. 83.1

I take for comparison with the above, seven Oxfordshire unions of the normal type, in all of which out-relief is freely given.

I will conclude this portion of the examination of the subject

with a comparison of two groups of large provincial urban unions:

The net result of the examination of poor-law statistics which has been made seems irresistibly to point to the conclusion that the policy which restricts outdoor relief does not lead to a necessary increase of indoor pauperism or of expenditure, whether upon salaries, establishment charges, or otherwise. Whatever the causes may be which lead to such an increase, it is not the cause alleged, for the figures show conclusively that in both particulars, in by far the larger number of cases, a greater increase is shown in unions in which the opposite system prevails.

One more statistical table I will append. Earlier in this article, referring to the cost per head of population as a basis for comparing one union with another in point of expenditure, I noticed that such a test might be interesting for the comparison of large groups of unions, as suggesting the existence