XXV] 



Introduction 



xliii 



Table XXV (p. 36) 



Value of the probability that the mean of a small sample of n, draiun at random 

 from a population following the normal law, will not exceed (in the algebraic sense) 

 the mean of that population by more than z times the standard deviation of the 

 sample. ("Student": Biometrika, Vol. VI. p. 19.) 



When n is greater than 10, it will be sufficient as a rule to use the approximate 

 result 



P= — ; I e ^ dx (xxxv) 



\/27r J-« ^ 



as a measure of the probability. This may be found from Table II. 



Illustration (i). Experiments of A. R. Cushney and A. R. Peebles on the 

 difference in effect of Dextro-hyoscyamine hydrubromide and Laevo-hyoscyamine 

 hydrobromide*. 





Additional Hours of 



Patient 



Sleep 





(Laevo - Dextro) 



1 



+ 1-2 



2 



+ 2-4 



3 



+ 1-3 



4 



+ 1-3 



5 







6 



+ 1-0 



7 



+ 1-8 



8 



+0-8 



9 



+4-6 



10 



+ 1-4 



Mean 



+ 1-58 



Standard Deviation 



1-17 



+ 1-58 ^^„. 



Table XXV shows that for ^ = 1 35 : 



P = -99854, 



or the odds are 666 to 1 that leavo- is a better soporific than dextro-hyoscyamine 

 hydrobromide. 



Illustration (ii). Difference in weight of crops of potatoes grown by Dr Voelcker 

 with (i) sulphate of potash and (ii) kainite as artificial manure. 



* Journal of Physiology, 1904. 



/2 



