DETERMINATION OF TRANSATLANTIC 



45 



A weight of £ is given to each of the values marked with an asterisk in obtaining the means of the grou, 

 For these values were obtained from signals in one direction only, with assumed values of the transit *ion 

 time, derived from other sets of the same group. On February 10, the Brest I 



Duxbury. 



(*) 



gnals 



It 



mean value of x derived from all the other signals of Class II. An assumed value of (j-) was applied to th| 

 signals of Class IV., viz. the mean value of x obtained from all the signals of Classes II. and III. 



DIFFERENCE OF LONGITUDE BETWEEN BREST AND DUXBIKV. 



Date. 



x 



1870. 



Class II. 



Jan. 5 



8 



ii 



a 



9 

 22 

 26 

 28 

 Feb. 10 



(< 



« 

 « 



h m 



s 



4 24 42.893 



42.895 



42.853 

 43.101 

 42.152 

 42.645 



I 



Class III. 



h 



4 



m 



24 



42.746 

 42.879 

 42.858 

 43.016 

 43.188 

 43.276 



1 



i 



Class IV. 



h m 



4 24 42.663 



Mr.ANS. 



h 



4 



24 42746 

 42 S6 



48 76 

 42.98 



4 4 



4.5.214 

 42.654 



Mean of Division A 4 24 42.92 



± .05 1 



h m § 



4 24 42.507 



42.084 



42.935 



44.098 



12 .«> 



> 



Mean of Division B (Rejecting January 14) 



4 24 42.905 



±.098 



L includes all the nights on which the local time was obtained at both stations by transits observed 

 « -~ u^ of the signals. Division B contains all the nights on which the Duxbury tune was co mi ,u„d 

 from transits observed on the preceding or following nights. The value of the longitude for January 14 ,. 

 rejected because on that night the time was computed from observations made on January 11 and 16- 



night 



The work of the best seven nights makes 



i 



The work of four inferior nights makes 



I 



4 h 24™ 42'.922 



± .05 1 



4 k 24 ro 42'.905 



±.093 



results inversely proportional to the squares of 



If weights are assigned to these two 

 their probable errors, the final result is 



The correction for ,he difference of the Personal Equation ia taking transit, between 

 Mr. Dean and Mr. Goodfellow is 



The correction for the difference of tl 

 between these observers is 



Applying these corrections, we have 



This result is computed from 1324 



1 



4* 24 ro 42*.918 



± .045 



0M)38 ±.012 



ignals 



I 



0*.018 ±.006 

 42'.867 ±.047 



ignals of Class 



positive, and about half 



Brest to Duxbury and half from Duxbury to Brest 



