The Barometer in South Africa. 



97 



Therefore if we use Stone's figures for 1841-70, we must apply 

 - 0.038 to them. 



P'rom 1871-80, capillarity • alone, and from 1881 to date 

 capillarity and -0.013 index error were applied. In a tabular form 

 these reductions are shown as follows : — 



Reduction previously used. 



Reduction to Prov. Standard. 

 in. 



- 0.038 



- 0.041 



- 0.004 

 + 0009 



The barometer at the Cape, at least during the winter months, is 

 lower during heavy rains. Table 2, based on the observations 

 1841-1870, shows the three heaviest and three lightest monthly 

 rainfalls for each month with the corresponding barometer means. 



The tables which follow require no further explanation. The cost 

 of compiling the Cape monthly means has been met by a grant from 

 the Association. 



P.S. — Mr. J. Power, of the Royal Observatory, has been good 

 enough to read the foregoing. He considers that from 1847 to date 

 my provisional standard should be corrected by -f-o.oo3in. Since the 

 above was put in type I have come across a paper by Maclear giving 

 the monthly means for 1842 -1855 (Board of Trade Meteor. Papers, 

 No. I, 1856). An inspection of Maclear's figures shows that Mr. 

 Power is right, and that from 1847 (inclusive) onwards o.oo3in. 

 should be added to all the printed figures in Tables 2-3-4. 



