LATITUDE AND LONGITUDE OF CAMPOS RODRIGUES. 393 



on the final comparisons of the invar wires, with which the base 

 was measured. These comparisons are being made at the 



International Observatory of Weights and Measures of Sevres. 

 In the medium length of the wires, which were 24 metres long, 

 a maximum error of one-tenth of a millimetre may l)e found, 

 which would possibly bring the error in the 76,829 metres up to 

 ;^2 centimetres. Besides this error, a maximum error of 

 1/100,000 may also be feared in the transport by triangulation 

 of the distances to Lourenqo Marques. In the end these errors 

 make up a total of less than 



or half a tenth of second of arc, which has been taken as the 

 probable error in the longitude. 



" These observations of azimuth distances having been con- 

 cluded at the Observatory, amongst which the Inhaca lighthouse 

 was thoroughly intersected, repeated observations of the 

 azimuths of circumpolar stars in proximity to the greatest elonga- 

 tion were made (and within such limits that one second of time 

 would not introduce in the azimuth an error of more than 

 1" of arc) in connection with the same lighthouse, taking advan- 

 tage of the periods when its light was less intense, and therefore 

 more clear in the focus of the telescope. The mean value of 

 this azimuth, after correcting the alierration and approximate 

 reduction to the mean pole, was 



27i°o6'45".2 -L-o". 32 



" From the compensation of directions observed, made at the 

 Head Department of Geodetic Works in Lisbon by the process 

 which was used in the compensation of the fundamental triangu- 

 lation of Portugal, the conclusion was arrived at that the angle 

 between the Inhaca and Ipoy was 



T25"24'5o".6 -bo".64 



" By coml)ining these two angles, we get the following as the 

 most probable true direction f roni the Observatory to Ipoy : — 



36°3i'35".8-ho".72 



the azimuth being counted from South to West, as is customary 

 in geodesy. The probable error o",72, at the distance of yy 

 kilometres, corresponds to a deviation in direction of 2/ centi- 

 metres. 



" Coming now to the calculation of the above calculated 

 azimuth and distance, and adopting the form of the Earth used 

 by the United States in the " Coast and Geodetic Survey " 

 (Clarke, 1366). the following differences are obtained between 

 the Observatory and Ipoy: — 



Latitude - 33'23".57o 



Longitude + 27'3i".657 



