APPENDIX A 



The result of this calculation gives the horizontal projection, that is to say, the 

 distance on the map, and hence the latitude of C. It is enough that the difference 

 of position of the two points should be sufficient to prevent any slight intermediate 

 barometric change producing more than an insignificant effect on AB. In cases 

 7 and lo these differences were considerable, amounting to 3,136 feet and 

 1,394 feet respectively. 



II. DECLINATIONS 



The declinations were obtained with the compass-theodolite with additional 

 piece. Previous observations of the sun had given its azimuth, and consequently 

 the true north. The mean of forty readings of the needle (5 point south ; 5 point 

 north ; needle above, telescope on the right ; id., needle above, telescope on the 

 left ; id., needle below, telescope on the right ; id., needle below, telescope on 

 the left) gave the magnetic north. The declinations of Tali-Fou and Ssumao were 

 obtained by eighty readings, forty with each of my two needles. Their error wa.s 

 from ^o" to i'. 



As was to be expected, the N.-E. declinations increased almost steadily 

 in proportion as we advanced in a north-westerly direction. It was only near 

 Muong-le (Chinese Laos) that minerals were in sufficient evidence to falsify 

 completely the indications of the needle. There I found three declinations so 



