APPENDIX A 



The two instruments used for determining the above latitudes were — 



(i) A HurUmann sextant graduated to 10", and 

 (2) A Hurlimann small theodolite graduated to i'. 



The sextant observations are subject to errors of centring, both constant and 

 variable, amounting in the best instruments to i ' ; so that the readings apparently 

 true to lo" are not absolute. The error can only be partially rectified by taking the 

 meridian altitudes of two stars, one to the north and the other to the south, and 

 adopting the mean of the results. This proceeding, however, requires both time 

 and patience, and the care and difficulty attending night observations with the 

 sextant are well known. 



Moreover, in tropical and semi-tropical countries these latter methods are the 

 only ones possible during the greater part of the year. From the 8th of March 

 I was obliged to discontinue the use of the sextant for obtaining latitudes by 

 observation of the sun, the double meridional altitude on that date reaching 126', 

 or almost the extreme limit of the graduations. From latitude 23' the instrument 

 became unavailable until October, .\dded to which the necessity for having a 



