OF THE VIBRATIONS OF AN INVARIABLE PENDULUM. 219 



vibration in the air of full pressure and in rarefied air was not sensibly affected 

 by a small motion of the support. 



It was now considered, therefore, as established by the experiments, that the 

 true reduction to a vacuum is considerably greater than it had been customary 

 to suppose ; for the invariable pendulum, for example, nearly as 5 to 3. It was 

 also obvious, that all pendulums whatsoever, employed in air, and designed to 

 give results which should be independent of the variable retardation occasion- 

 ed by their vibration in aii", would require to have the influence of the air on 

 their respective vibrations, ascertained by experiment, since it is not attainable 

 by calculation. Now as the apparatus was suited by its construction, to furnish 

 this element with facility and accuracy, for any of the forms in which pen- 

 dulums have hitherto been made, either for determining absolute or relative 

 lengths, it was probable that it might eventually become more extensively 

 useful, than in its present office of furnishing the reduction for an invariable 

 pendulum. 



It was thought propei*, therefore, that the apparatus should be removed to the 

 Royal Observatoiy at Greenwich and established there, in order that it might 

 be hereafter at the command of persons to whom it might be useful, upon their 

 application to the Board of Longitude at whose expense it had been con- 

 structed. The iron suspension plate with the iron bars supporting it were now 

 removed, and the bell metal plate with the circular exterior ring, represented 

 in Plate VI, substituted, with the iron frame-work and screws, as represented 

 elsewhere in the plate, enabling the support of the pendulum to be fixed im- 

 moveably at pleasure in the manner already described, A clock by Dent, with 

 a mercurial pendulum carrying a disk, was placed in the angle behind the 

 apparatus ; and the telescope for observing coincidences in front, about 16 feet 

 distant from the detached pendulum when suspended. Arrangements were 

 made for observing the coincidences by artificial light, without interfering with 

 the temperature of the room, by directing the light of an Argand lamp, sta- 

 tioned in an adjoining apartment, on the disk of the clock pendulum, through 

 a tin tube, which prevented the diffusion of the light in the room. The dia- 

 phragm was placed between the glasses and the clock, and the arc within the 

 glasses close to the pendulum. The arc was graduated in inches and tenths, 

 and was read off to hundredths ; crossing the pendulum at 47.7 inches from 



2 F 2 



