84 INDUCTION SHOCKS 
which the ratios of to Z in successive experiments are 
plotted. The horizontal line represents the average 
ratio of j8 to Z as determined in these experiments; the 
variations from this line of the different actual ratios 
are, as is seen, relatively inconsiderable, the greatest 
being 18.5 per cent, the average of all slightly under 
ii per cent. 
Assuming the data cited in Table IX to be fairly rep- 
resentative of the relations between and Z that are 
Exp. 
FIG. 17. Illustrating the relatively slight departures of individual 
ratios of /3 to Z from the average. Ordinates represent successive ex- 
periments; abscissae represent ratios of ft to Z. The horizontal line 
is drawn at the level of the average ratio. 
likely to occur in experiments of the sort under consid- 
eration, to what extent are we justified in such experi- 
ments in making use of the values of Z for expressing 
quantitative relationships? 
The figures show clearly, I think, that all except the 
finest relationships are revealed with sufficient exactness 
by the values of Z. While one cannot always know cer- 
tainly, in cases in which several nearly equal values of 
Z are under comparison, which will give smaller and 
