lii 



REPORT — 1869. 



repose, and the adliesion of the inercuiy to the glass is not counteracted by 

 tapping or moving the tube, it is desirable to test the results obtained in order 

 to see if the influence of adhesion causes a perceptible sluggishness of the 

 mercury. The Standard Barometers, to which in aU cases the Barographs 

 are referred, are, on the other hand, subject to motion, and are probably 

 sufficiently moved in the operation of reading to counteract any sluggishness 

 of the mercury. 



Now, four or five times each day, while the light is cut off from the re- 

 cording cylinder of the Barograph by the clock an-augment, the Standard 

 Barometer is read. We can thus compare these standard readings with the 

 simultaneous measurements of the Barograms, these latter being of course 

 properly tabulated, converted into true inches, and the residual correction 

 applied as described in the Eeport of the Meteorological Committee for 1867. 



Should there be any sluggishness in the mercury of the Barograph we 

 might expect to discover it by means of this comparison, for in such a case 

 the Barograph would lag behind, and thus read too low with a rising and too 

 high with a falling barometer. 



If therefore we presume that the Standard Barometer is free from slug- 

 gishness, and denote its readings by S, and those of the Barograph by B, 

 theu S — B ought in the case of sluggishness of the Barograph to be positive 

 for a risincf and negative for a falling barometer. 



Several months' observations have been discussed in this manner for each 

 of the observatories, and the result is exhibited in the following Table : — 



From this Table we see how inappreciable in all the observatoi-ies is the 

 retardation of the Barograph Barometer as compared with the Standard, 

 while in Stonyhurst the Standard even appears to be a trifle more retarded 

 than the Barograph Barometer. 



Errors of recording nml tabulating. — Under this head we maj^ include 

 (A) errors of adjustment and attachment of paper, (B) errors of time and 

 date, (C) errors in tabulating from the traces. To begin with the first of 

 these : — 



(A) Errors of adjustment and attacJiment of j^ajicr. 



Want of definition arising from an improper adjustment of the lens ought 

 to be noticed, but it is bcUeved that the definition is good in the case of all 

 the observatories. As the instrumental constants for all the various Baro- 

 graphs have now been determined, it would hardly seem expedient to alter the 

 position of the lens, which Avould alter these constants, for the purpose of 

 procuring greater perfection in defiintion. 



The photographic sheet which is attached to the cylinder of the Barograph 



