396 LATITUDE AND LONGITUDE OF CAMPOS RODRIGUES. 



The table of corrections for differential refraction was ob- 

 tained by means of the formula 



h (r — r') r= 29" sin- Z sin i' (Z' — Z") 



The results obtained are given in Table I\'', which only men- 

 tions seconds; the respective probable errors of the averages of 

 each pair and each day. the probable error of each pair of stars 

 and, finally, of the general average, are also mentioned therein. 



The three results shown in brackets were rejected. They 

 were not considered at the time of observation as worthy of con- 

 fidence, either on account of clouds, or owing to the passing of 

 the stars across the thread of the reticle, which made their bisec- 

 tion with the movable thread most difficult. 



Besides the above-mentioned errors, the following were 

 further determined : 



Probable error of declination for one pair d-o".i30 



Probable error of declination for one star -bo". 184 



Probable error of observation for one pair rko".i27 



Probable error of observation for one star -to". 179 



Through Major Frederic Oom, I received the following 

 reduction to the mean pole, kindly transmitted by the Geodetic 

 Institute of Potsdam, and obtained from the results, not yet jmb- 

 lished, of the international observations of the variation of 

 latitude. 



" For Lourenco IMarques the reductions will be : 

 1912 — January 23rd-26th : 



(j>^~4>=-0".22 



1912 — February 8th-9th : 



</)q-<^=-o".i7" 



By the application of these reductions, we get the following 

 value of the latitude reduced to the mean pole: 



1912— Tanuarv 23rd — 25° 58' 5". 52 -lzo".049 



24th 5".52Hzo".046 



25th 5"-62 -b o".o78 



26th 5".43 -I- o"-049 



1912— February 8th 5".49 4;o".046 



9th 5". 40 -b o".o62 



The mean value adopted for the latitude of the pillar of the 

 transit instrument reduced to the mean pole is : 

 25° 58' o5".5o -bo".022 S 



Thus we have for co-ordinates of the transit instrument's 

 pillar : 



