336 Anthropometry of Scottish Insane 
Diagram XI. Eelationship between Q and Log P. — Hair Colour, Females. 
20 
-© 
© 
© 
© 
G 
©® 
( 
© e 
■01 02 03 04 OS 06 07 08 09 10 -11 -12 13 -14 IS -16 -17 -18 -19 -20 
Q 
with N) is not a symmetrical one, but can be accurately described by a skew 
curve of either Type I. or Type IV. Thus, he points out that the tables of the 
probability integral cannot accurately give the areas on either side of the ordinate 
which divides the curve at the abscissal value (y" — 'i/s')j'^mpq{N — m)j{N — l), 
and the probability of greater or lesser values occurring in future samples must be 
determined by other means. Since however the values of (m — l)/(-A^— 1) in the 
present series are small (although not quite negligible), an approximate estimate 
of the significance of each difference can be obtained by determining the values 
of the ratio {i/s —ys)l'^mpq, the distribution of these relative differences being 
assumed to follow the normal curve. The relative differences themselves are 
thus, on this basis, the abscissal values of the normal curve = 1 / V27r . e^^^l 
The latios {ys — ys')l^'mpq {N — m)l{N ~1) and (yj — ys')/\/mpq have both been 
