WHOLE VOL. NOTKS. NORTH ATLANTIC 85 



muda was —.033 in., and in reply the Senior Medi- 

 cal Officer stated (27th April. 1914) that the figure 

 + .010 in. was the correction for index error and 

 that the gravity c(jrrection had in fact not been ap- 

 plied. From this statement, which is supported by 

 a comparison of the whole series of annual means 

 with those at the nearest stations on the coast of 

 America, it was inferred that the observations had 

 been corrected for index error, but not for gravity, 

 and the latter correction (—.033 in.) has been a])- 

 plied throughout. 



Table A. — Corrections Applied for Differences of Height. Indies 



Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. 



CI ft. 



to 1.t1 ft. —.090 —.096 —.090 — .090 —.095 — .094 —.094 — .094 —.094 —.095 —.095 —.096 



120 ft. 



to 151 ft. — .C?.? —.03.3 —.033 —.033 — .033 — .032 — .03-2 —.032 — .032 —.033 —.033 —.033 



Table F,.— Corrections Applied to Reduce A ((/' + ji'') to V (y'' + i .i'' ) . Indies 



Jan. Feb. Afar. Apr. Ma3' June July Aug. Sept. Oct. Nov. Dec. 

 —.017 —.017 —.010 —.013 — .009 — .000 —.004 — .000 —.011 -.015 —.010 — .016 



Based on observations at .'*'' 41"', 15" and 20'' 41"\ 1913-1918 inchi.sive. 



Temperature. 

 Authorities: 



1887 to i8f)o. London. .Army Medical Department. Annual 

 al:)stract of meteorological observations taken at 

 Netley and stations abroad. London. Army Medi- 

 cal Dept. Re])orts. 



T891 to 1803. Manuscri])t data su]i])lied by the Meteorologi- 

 cal Office. Toronto and filed in the Meteorological 

 Office. London. 



1894 to 1900. .As for Pressure. 



T90T to 1910. Toronto. Meteorological Service of Canada, 

 Annual Report. 



1911 to 1920. As for Pressure. 

 Site: See note under Pressure. 



Observations: The combination .', (9'' + 2i'^ + Max. + Min.) ap- 

 pears to give a good representation of the 24-hour 

 mean. From 1866 to 1890 and from 1894 to 

 1900 many of the observations of the daily mini- 

 mum temperature a])pear to be unreliable and the 

 means for 1866-1890 and 1894 to 1900 are the 

 means of the readings at 9'' corrected to the com- 



