SecoJids Pendulum at Port Bomoen. 



171 



of this experiment and the comparison of it with my formula, 

 Phil. Maff. for Oct. 1826, are as follows: 



Observed 

 pendulum. 

 39-20347 



Computed 

 pendulum. 

 39-20265 



Excess of cal- 

 culation. 

 — •00082 



Latitude. 

 73° 13' 39"-4. 



The error is not great: and this is the 29th experiment re- 

 presented by my formula with small discrepancies. 



The latitude of Port Bowen being little more than a de- 

 gree short of that of Captain Sabine's station at Greenland, 

 we may compare the two experiments. 



Latitude. Observed pendulum. 

 Sabine .... 74° 32' 19" 39-20335 

 Foster .... 73 13 39 39-20347 



Here the pendulum has shortened for an increase of latitude 

 equal to 1° 18' 40". But it ought to have lengthened at least 

 •00250. Thus there is a discrepancy between the experiments 

 of the two observers, greater than between my formula and 

 Captain Sabine's result. 



Port Bowen is in the middle of Captain Sabine's northern 

 stations. We may therefore employ Mr. Foster^s experiment 

 to compute the pendulums at those stations in different hypo- 

 theses of ellipticity, in order to compare them with the experi- 

 mental determinations of Captain Sabine. Put I and A for 

 the length of the pendulum and the latitude, at Port Bowen ; 

 and let /' and A' denote the same things for any of Captam 

 Sabine's stations ; then, 



V = Z-/(sinU-sin-x'). 



According to my formula, /= 0-20835 ; and according to 

 Captain Sabine's calculations,/ = 0-20227; and these values 

 may be considered as nearly the greatest and least that can be 

 assigned with any probability. Calculating, now, with these 

 data, we get : 



Station. 



Drontheim . 

 I lammerfest 

 Greenland . 

 Spit/bergen 



Observed 

 pendulum. 



39-17456 

 39^19519 

 39^20335 

 39-21469 



Computed 



pendulum 



/= 0-20835 



39-17920 

 39-19804 

 39^20605 

 39-21436 



Computed 



pendulum 



/ = 0-202.'27 



39-17990 

 39-19820 

 39-20597 

 39-21404 



The computed quantities are very consistent with my for- 

 mula; but they do not agree well widi the observed pendu- 

 lums. In particular tlic discrepancy at Drontheim, comijuted 

 from Mr. Foster's experiment on one side, is nearly c(]ual to 



Z 2 what 



