14 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 79 



1887 to 1888; 1891 to 1895 July. Barometer in use, is un- 

 known, but it is assumed that the necessary correc- 

 tions have been made. 



1895 August to 1911 December. Barometer No. 3 A. M. D. 

 Index error correction -I-.004 in. appHed. 



1912 August to 1920 December. Barometer No. M. O. 1233 

 Kew Pattern. Index error correction —.003 or 



— .004 in. (according to the reading) appHed. 

 CJiaugcs of Hours of Observation: 



1877 March to 1910 June 9'\ 15''. 



1910 July to 1920 December 9^*, 17''. 



All values have been corrected to the mean of 24 hours by 



corrections (Table A) based on observations at 



Duala and Sansane Mangu. 

 Notes: All values have been corrected to normal gravity (Lat. 



45°) by applying a correction of —.073 in. or 



— .074 in. according to the barometer reading. 



* The value for I5hr. was thought to be doubtful and the 

 pressure was corrected to the mean of 24 hours 

 from the mean at 9^ (Table A). 



t The value for 9^^ was thought to be doubtful and the 

 pressure was corrected to the mean of 24 hours 

 from the mean at 15'^ by applying a correction of 

 + .046 in. 



The years 1887 and 1888 are already reduced to Mean Sea 

 Level in the Army Medical Department reports 

 and the values have been reduced to station level 

 by applying the corrections given in Table B. 

 Temperature. 

 Authorities : 



1874 October to 1877 February. Manuscript returns com- 

 municated by the Royal Army Medical Corps and 

 filed in the Meteorological Office, London. 



1877 starch to 1920. As for Pressure. 

 .S"i7r." As for Pressure. 



Observations: The standard adopted is the mean of the mean 

 daily maximum and mean daily minimum. These 

 figures were unreliable or wanting in certain months 

 and the values in italic have been computed by 

 applying corrections (Table C) based on a number 



* t Sec Tables. 



