394 LATITUDE AND LONGITUDE OF CAMPOS RODRIGUES. 



" The combination of all above calculated data with the posi- 

 tion ot Ipoy, which was siipi)lied to me direct in Pretoria at the 

 Department of Trigonometrical Survey by Mr. H. E. Schoch, 

 Surveyor-General, is summarised in Table IL" 



The determination of the longitude was confirmed by obser- 

 vations of lunar culminations made at Lourengo Marques by the 

 Sub-Director of the Astronomical Observatory of Lisbon, Major 

 Frederico Oom, of the Engineers, in combination with identical 

 and simultaneous observations made at the Lisbon Observatory 

 by Major Teixeira Bastos, of the Engineers, first-class 

 xA.stronomer of the said Observatory. A small, unimportant 

 difi^erence was found, as the last limit of approximation that may 

 be expected from the process of lunar culminations, even after 

 two or three whole years of observations of the Moon, is not 

 inferior to one second of time. T am unable to supply any 

 more information concerning this determination of longitude, 

 because its solution was obtained at the Lisbon Observatory. 



The geodetic latitude was confirmed twice by astronomical 

 observations, first by Commander Gago Coutinho and then by 

 myself. 



I again quote Commander ( iago Coutinho's remarks : — 



" Astronomical Observations of Latitude. — These obser- 

 vations were made with two Troughton and Simms theodolites, 

 with microscopes and telescope of 33cm., alternately mounted on 

 the pillar of the transit instrument. Four series of observations 

 were made, during four nights, a dift'erent position of the vertical 

 circle being utilised on each night, and a different theodolite from 

 the one utilised during the first two nights was used during the 

 last two. 



" Each unit of weight results from the average of the two 

 meridian observations of the stars, the co-ordinates of which 

 appear calculated in either of the three astronomical almanacs, 

 Nautical Almanac, Berliner Jahrbuch and Connaissance des 

 Temps, their zenith distances being approximately equal — one to 

 the North and the other to the South of the zenith — w^hich had 

 in view the elimination of the errors of the co-efiicient of re- 

 fraction, and all the systematic errors due to the instruments. 



" The results of these observations were as shown in 

 Table IIL 



" The extreme values of the latitude in yy pairs w^ere i".o and 

 9". 6; the probable error in each pair about rh ^".05. Thus the 

 probal)le error of the average of the yy pairs comes to about 

 dz:0".I2. 



'' The average of four nights, allowing the same weight to the 

 average of one night, is 



25°53'o4"-y<'^-lr.o".22 



this value Ijcing accepted, with the reduction of — o"i to the 

 mean ])ole ( Albrecht)." 



