1889.] Africa from Barometric Observations. c21 



certain corrections whicli may add somewhat to its substantial 

 accuracy. In the course of a few years, with the experience which 

 has been learned as to sources of error, there is a reasonable prospect 

 that a considerable increase of accuracy may be attained in these 

 results. Meanwhile what has been done shews that barometric 

 observations, continued over a sufficient space of time, are capable 

 of furnishing very useful results as to sea level. Probably however 

 the approximate accuracy of these results is due to a great extent 

 to the somewhat exceptional climatic conditions of the southern 

 extremity of the African continent. There is no doubt that owing- 

 to physical circumstances there is quite an unusual degree of unifor- 

 mity of average temperature over the region to which these obser-^ 

 vations apply, and it is this approximate uniformity of average 

 temperature that renders the calculation of sea level from barometric 

 observation as satisfactory as it seems to be. It is very doubtful 

 whether a similar method could be applied with equally satisfactory 

 results to such a continent, for example, as North America, where 

 approximate uniformity of average temperature could not be assumed 

 without leading to entirely erroneous results. 



