III. — Aleteoroloiiical, etc., Observations. 



valuable studios on liuwenzori. But if the ilatu have nevertheless been referred 

 to Bujongolo, it was because they offered greater guarantees of accuracy, and 

 this for the following reasons : — 



First of all, the difference of altitude between Bujongolo and Fort Portal 

 may be regarded as near enough, having been obtained from a considerable 

 number of observations. On this account we may consider as partially com- 

 pensated those sources of error which are due to atmospheric disturbances, 

 such as may have a great influence on isolated measurements, especially when 

 dealing with very remote stations. Such would precisely have been the case if 

 the data observed on Ruwenzori had Ijcen directly compared with those 

 corresponding hour for hour with Fort Portal. But by making the comparisons 

 with the data obtained at Bujongolo, a much nearer place, one may fairly 

 assume a greater uniformity of atmospheric conditions. 



Then there is another fact which shows the greater convenience of the 

 course adopted. It is seen in the following example to which many others 

 might be added. 



On 7th July, at 12 o'clock, on Edward Peak, the pressure (reduced to 0') 

 was 428-5 mm., and the temperature I'Q Celsius (34'"4 F.), the corresponding 

 readings being at Bujongolo 489-lG mm. and 3°-9 Celsius (39° F.), and at Fort 

 Portal 638-69 mm., and 23°-3 Celsius (74° F.). 



Calculating from these data the difference of level between Edward I'eak 

 and Fort Portal, and then separately lietween Edward Peak and Bujongolo, and 

 between Bujongolo and Fort Portal, we get : — * 



Difference of level between Edward Peak and Fort Portal ... 3,355 metres 

 Difference of level between f]dward Peak and 



Bujongolo 1,074 metres 



Difference of level between Bujongolo and Fort 



Portal 2,249 „ 



Total difference of level between Edward I'eak 



and Fort Portal •"5,323 metres (1 0,900 feet) 



Therefore, with the direct calculation, and omitting Bujongolo, we have a 

 difference of over 32 metres (105 feet). 



Such a difference arises from the fact that whereas the law of Laplace 

 is based on the hypothesis of a static equilibrium of the atmosphere, and of 

 a temperature and lumiidity which decrease regularly with the altitude, this 

 does not take place in the present instance. 



* In this estimate no account is taken of the tension of the aqueous vapour. 



369 2 B 



