Proceedings of the Royal Society, 301 



2dly. With the invariable pendulum the disc subtended a some- 

 what less angle than the tail-piece, so that the inferred number 

 of vibrations in 24 hours was diminished about two-tenths of a 

 second. But experiments with the invariable pendulum being in- 

 tended to be in the strictest sense of the word comparative, this 

 constant difference will no way affect the ultimate result. But, 

 as the most direct way to remove any doubts which may be en- 

 tertained on the subject, the author has computed from the whole 

 of Captain Sabine's observations, the successive differences in the 

 vibrations at the various stations visited by him, by the two me- 

 thods, viz., that of employing the disappearances and re- appear- 

 ances, and the disappearances alone. The results only in one in- 

 stance differ so much as a tenth of a vibration, they are indiffer- 

 ently in excess and defect, and the mean of their discrepancies is 

 exactly nothing. From this he concludes that if the observations 

 be made as nearly as possible under similar circumstances, the 

 method of observing by disappearances alone, is productive of no 

 perceptible error in practice, in experiments with the invariable 

 pendulum ; while in those with the convertible pendulum, the 

 equal apparent sizes of the disc and tail-piece, preclude the possi- 

 bility of any, either in practice or theory, from this cause. 



The standard of Sir G. Shuckburgh having been found identi- 

 cal with that by Bird, in the custody of the Clerk of the House of 

 Commons, adopted as the imperial standard unit of extension, the 

 length of the pendulum already determined is fixed with the 

 same degree of precision in parts of the imperial standard yard. 



A repetition of Sir G. Shuckburgh's experiments on the weight 

 of given volumes of distilled water, and a re-measurement of the 

 cube, sphere, and cylinder used by him, were found to give no 

 material variation from his results, these being stated in terms 

 of the mean of several standard weights kept at the House of 

 Commons. The troy pound nearest the mean has been adopted, 

 and declared by the legislature to be the original unit of weight 

 under the denomination of the imperial standard Troy pound. 



The relation between this pound and the cubic inch of distilled 

 water at 62° Fahr., bar. 30 in., has been ascertained by the com- 



Vol. XX. Y 



