The first glance at this Table shows that the observations are very discordant 

 but, for a proper calculation of their probable errors, the v's and (w)'s are put down 

 in Table XXIII. 



TABLE XXIII. 



From this Table, we find the probable error of a single observation to be- 

 es = = 12- 7 XlO- 7 sec. 

 e 7 ==24-6 

 e 21 ==11-9 

 e m = - 12-9 XlO- 7 sec. 



Also, for the probable error of the mean result : 



p b --= 5-1 XlO- 7 sec. 

 p, -9-8 



p m == 5-2 X 



sec. 



These values of e f and of p x are large in comparison with those obtained at Potsdam, 

 but hardly excessive in view of the difficulties inherent in work at such a station. 



As before, to exhibit the magnitude of the observational error, we form the following 

 Tables of the differences from the mean pendulum. 



TABLE XXIV. 



38 



