In most cases the median and quartiles and sometimes the dispersions have been calculated. They may be explained as follows. Suppose the wages of, say, sixty persons to be arranged in ascending order, _e.g._, 5_s._, 5_s._ 3_d._, 6_s._, 6_s._ 1_d._ ... 11_s._ 9_d._, 12_s._, 12_s._ 6_d._, then the wage _halfway_ up the list is the _median_ wage; thus, there are as many individuals above the median as below it. The wages halfway from the ends to the median (_i.e._, fifteenth and forty-fifth from bottom), are the _quartiles_, so that between the quartiles half the wages are grouped. Thus, if the median and quartiles in the above list were 7_s._ 6_d._, 10_s._ 6_d._, 12_s._ 6_d._, there would be fifteen earning less than 7_s._ 6_d._, fifteen more than 12_s._ 6_d._, thirty between 7_s._ 6_d._ and 12_s._ 6_d._, thirty below and thirty above 10_s._ 6_d._ For a single measurement of the grouping of the wages about their median, the distance between it and the quartiles is significant: in this example 3_s._ and 2_s._ are these distances. The more convenient way of stating this is to express half the distance between the quartiles (2_s._ 6_d._) as a fraction of their average 10_s._, which is generally very nearly the median. This fraction (¼ or ·25) we call the _dispersion_, and it enables us to study the changing character of a group in a very simple and efficient manner. I.--STATISTICAL VIEW OF THE VARIOUS FIRMS. FIRM A. _Information obtained._--Wages of thirty-six hands tabulated week by week through 1899. Total amount paid in wages and total number employed each week, 1885-1899. The whole wages sheet for one week in July and one week in November for each of these fifteen years. A. is a firm employing from fifty to one hundred and ten women and girls as folders, stitchers and sewers. The number employed has changed gradually; in 1885-7 there were about a hundred: from 1888 to 1894 the number continually diminished to sixty, and after a brief spurt in the autumn of 1894 to one hundred and fifteen and a rapid fall, has from 1895 to 1899 gradually risen from fifty to ninety. Through the fifteen years which the statistics cover, 1885-1899, the _annual_ average (roughly calculated) has fluctuated within the narrow limits of 8_s._ 9_d._ and 10_s._ 6_d._; it was above 10_s._ in 1888, 1889, 1897, 1898, 1899; below 9_s._ in 1886 and 1894. This average includes the learners. But when examined more minutely it is seen that the fluctuations week by week and month by month are very rapid. Briefly, there is a change of about 4_s._ in four-weekly cycles. Thus in November, 1899, the averages for the five weeks were 10_s._, 11_s._ 2_d._, 13_s._ 9_d._, 13_s._ 10_d._, 12_s._