Appendix B. 



In fact, ill spite of every care taken to make continual comparisons with the 

 Foitin liarometer, the term of correction did not keep constant, not only from 

 clay to day, Imt even during the same day whenever the instrument got shaken 

 or was exposed to sudden changes of altitude. 



Hence, to avoid errors that might even be serious, a cautious and limited 

 use has been made of the data obtained with the aneroid. Thus, the 

 simultaneous indications of the Fortin and the aneroid being noted, and the 

 altitude obtained from the former being taken as correct, the indications of 

 the latter have served to establish differences of level, which were inconsiderable 

 relati vel}' to the point of comparison. When this process was completed the 

 comparison was renewed, so that the data first obtained were brought under 

 control. 



In general the reported data result from the mean of two or more 

 determinations, and may consecpiently be regarded as sufficiently accurate. 



As to the altitudes of the places passed by the expedition on the return 

 journey, that is, from and after the 14th July, these were all ol)tained by 

 means of comparisons with the data ol)served simultaneously at I])anda.* Then, 

 for the sake of uniformity, they were reduced, like the previous ones, to the 

 common level of Bujongolo. 



^Vmongst the determinations made relatively to Ibanda was that of 

 Iolanda Peak, the altitude of which will consequently not be so near the truth 

 as that of the other peaks. And, besides the inconvenience of Ibanda lying 

 still lower than Fort Portal, there was also the trouble caused by the breaking 

 of one of the mercurial barometers, instead of which we had to use the 

 hypsometer, which yields a less degree of approximation in the re.adings. 



NOTE. 



In the western districts of the Uganda Protectorate the geodetic survey 

 has IK It yet been cari'ied out, l)ut the far-seeing British Government is taking it 

 in hand, and no doubt it will soon be an accomplished fact. Hence it mio-ht 

 seem reasonable to refer the various altitudes of the Kuwenzori group, not 

 to Bujongolo, but to Fort North Portal, the exact height of which above 

 sea-level will soon be known. In fact, this very critical point is being seen to 

 by Messrs. H. Y. Tegart and II. E. Mtuldox, who have recently published some 



• For the altitudf of Ibmuia ivlativclv to F"ort Portal- 148 wasi assumed, this being tlic 

 mean of the observations made on goicg and returning. 



3G.S 



