OF .MTKIIilT!-;.-. IN >TAT1STI(S 3 17 



of publiiMtiun or arrangement of its results, and the directions in which future work 

 might |M.sv;ililv IM- undertaken. 



To deal with the questions of arrangement, &c., first. I suggest that the notation 

 might with great advantage be altered to the notation of JEVONS, such as I have used. 

 The shortcomings of a notation which has to represent the simple second-order 

 frequency (A/8) hy (A+) (AB+) are obvious. Next, I have noticed that the 

 arrangement of frequencies is very irregular in the report. Where the ultimate 

 fourth-order frequencies (Dr. WARNER'S " Primary " groups) alone are given it is 

 quite clear, but where more are given the data are seldom complete, and groups of the 

 same order are not kept together. As an example of the way in which I think 

 frequencies ought to be given I append a table giving the general results of the 

 1892-94 investigation. If so much space could not be spared, I think the statement 

 of the fourth-order frequencies is quite sufficient, as the others can be so readily 

 calculated from them.* 



As it L;MI.]S future work, I find myself unable to follow at all the remarks made by 

 the Committee on p. 5 of the " Report " (the italics are mine) : 



" A very valuable addendum to vital statistics might be obtained by following 

 up the history of certain cases recorded, by subsequent periodical inspections, but 

 as this is beyond the power of the present Committee, it can only be suggested as 

 one among many other directions in which enquiry may be pushed in the hands of 

 official Commissions." 



The Committee or Childhood Society is, as I gather from its Reports, continuing 

 the work of inspecting children in schools, and why it should be " beyond their 

 power " to reinspect a few large schools year by year is by no means obvious. They 

 are at present (in the last Reports,! for 1898-99) issuing statistics of the frequencies 

 of different groups of defects for different ages. As I have already pointed out 

 (note on p. 304) these statistics are rendered almost worthless by the omission of the 

 frequency (/8yS) the number of vndefective children at each age, only those who 

 were defective having apparently been noted. Even if the material were, however, 

 complete, it would not enable us to distinguish between changes due to selection and 

 changes due to growth, nor consequently to state what are the effects of these agencies 

 each by itself. Nothing but observance of one group of individuals year by year 

 can do this, and there seems little more difficulty in this than in the work already 

 being carried out. It is surely futile to expect a Royal Commission on the subject. 

 It would, in fact, be more appropriate for a body, specially created for the " Scientific 

 Study " of childhood, to take up an investigation of the greatest scientific interest 

 but probably of little immediate practical use. I would most strongly urge the 



* There appear to be a good many misprints in the Report, many of the frequencies being in disagree- 

 ment with those given for the ultimate (" Primary ") groups. I have generally assumed the " primary " 

 frequencies to be correct. 



t British Association Reports, 1898, p. 691 ; 1899, p. 489. 



