16 Prof. J. Le Conte on the Discrepancy between the Computed 



one litre of pure dry air, at the same place and temperature, and 

 under a pressure of 0*760 metre of mercury at 0° Centigrade, 

 weighs 1*2932227 gramme. Hence it follows that at Paris, 

 60 metres above the sea, and under the physical conditions 

 above indicated, mercury is 10513*2163 times as heavy as 

 air. Consequently the height of a homogeneous atmosphere 

 = h = 0-760 metre x 10513-2163 = 7990*044388 metres, or 

 26214-52864 British standard feet. 



Secondly, in relation to the value of g (the velocity generated 

 by terrestrial gravity in a mean solar second of time), it is well 

 known that it depends on the accuracy of the determination of 

 the length of the seconds 3 pendulum. The suspicion of Newton, 

 confirmed by the long-neglected experiments of Du Buat, ex- 

 tended and verified as they have been by the more recent and 

 accurate investigations of Bessel, Poisson, and Baily*, have esta- 

 blished beyond a doubt that the old formula for the reduction to 

 a vacuum is very defective, and by no means expresses the whole 

 correction which ought to be applied — in short, that the coeffi- 

 cient of buoyancy heretofore used in these reductions was entirely 

 too small, in that it took no account of the influence of the mass 

 of adhesive air which a moving body drags along with it. The 

 absolute amount of correction required for this cause cannot be 

 deter mined by calculation, " but must, in every case, be ascer- 

 tained by actual experiment." The experimental investigation 

 has been undertaken both by Bessel and by Baily, from which 

 it appears that the amount of correction will " vary according to 

 the length, magnitude, weight, density, and figure of the pendulum." 

 According to BesseFs experiments, the old correction for this ele- 

 ment must be multiplied by the factor 1*956 to obtain the true 

 reduction. The more elaborate and direct experiments of Baily 

 indicate that the value of this factor is subject to such variations, 

 that no great degree of confidence in the exact number can be 

 secured, unless the length, magnitude, figure, construction, and 

 mode of suspension of the pendulums which are the subjects of 

 comparison are identical. Fortunately the platinum sphere 

 used by Mr. Baily in his experiments was sensibly identical in 

 all these particulars with that employed by MM. Biot, Mathieu, 

 and Bouvard for determining the length of the seconds 5 pendulum 

 at Paris. According to Baily's experiments with this sphere, the 

 old correction must be multiplied by the factor 1*881 in order to 

 obtain the true reduction. Inasmuch as M. BessePs experiments 



* Newton's Princlpia, Book 2. Prop. 27. Cor. 2. Du Buat's Principes 

 d'Hydraulique, 2nd ed. Paris, 1786. Bessel, in Memoirs of Royal Academy 

 of Sciences of Berlin for 1826. Poisson, in Memoires de VAcademie des 

 Sciences, vol. xi. p. 521 et seq. Pans, 1832; also Traite de Mecanique, 

 2nd ed. vol. i. p. 363. Paris, 1833. And Baily, in Phil. Trans, for 1832, 

 p. 399 et seq. 



