REPORT or THE KEW COMMITTEE. 



Ivii 



of the Standard and the Barograph Barometer -will he comparatively great, 

 but if he both observe his Standard and measure his curves well, the mean 

 difference -will be small. 



The foUo-nang Table exhibits the results of monthly comparisons bet'sveen 

 simultaneous Barograph and Standard readings for the year 1868 for all the 

 observatories. 



Mean Diiferences bet-ween Barograph and Standard Keadings. 



1868. 



January ... 

 February 

 March ... 



April 



May 



June 



July 



August ... 

 September 

 October ... 

 November 

 December 



It is imagined that the mean differences shown by this Table have for all 

 the observatories by the end of the year reached a minimum value not much 

 larger than would be obtained by two observers reading the same Standard, 

 or by the same observer reading it twice. 



But while the simultaneous readings of the Standard and Barograph Baro- 

 meter afford us one means of testing the correctness of the observation mea- 

 surements, they do not yet do quite enough ; for, in the first place, these 

 simultaneous differences may be caused in part by an instrumental error or 

 by some local peculiarity, such as rapid heaving of the barometer, and in 

 the next place, an observer may unconsciously bestow a greater amount of 

 pains upon these measurements, which are simultaneous with Standard read- 

 ings, than he does upon his other measurements, and the above differences 

 may not therefore be a true representative of his general correctness. A 

 certain number of remeasurements of the curves of each observatory should 

 therefore be made at the Central Observatory, and the monthly mean differ- 

 ence between these and the corresponding measurements by the local observer 

 be recorded*. 



* It was not until tlie various observatories bad been supplied with their improved 

 tabulating in.strument that the final metliod of making these measurements vi-as decided 

 on. Since the beginning of 1809 the plan has been to make for each month for each 

 ob.servatory forty remeasurements of the curve at Kew, obtaining also independently the 

 residual correction. These final values are then compared with the corresponding values 

 obtained at the outlying observatories, and the result of this comparison for the first 

 three months of 18C'J has beeu as follows : — 



Mean Difference between 1st and 2nd Measurements. 



