44 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 79 



Barometer No. 7 by Negretti & Zambra (index, etc. correction 

 + .018 in.) was used from January 1866 until February 

 1872 when Barometer A. M. D. No. 26 replaced it. 

 Fortin Barometer A. M. D. No. 26 fixed at a height of 53 ft. 

 above Mean Sea Level continued to be used for the obser- 

 vations until December 17, 191 2, when a Kew Pattern Ba- 

 rometer No. M. O. 1242 was substituted. The correction for 

 index error and capillary action of Bar. A. M. D. No. 26 was 

 stated in 1872 to be —.016 in. and this correction has been 

 applied to all hitherto published values. The station at Gi- 

 braltar was inspected in September 1912 by an official of 

 the Meteorological Office London who found that the index 

 etc. correction of Bar. A. M. D. No. 26 was then + .053 in. 

 A comparison of Cape Spartel and Gibraltar pressure data 

 for the period 1895-1920 confirms the inspection report. 

 Moreover, it appears that by 1895 the correction had changed 

 from —.016 in. to +.013 in. Discontinuities appear at the 

 end of 1897 and in November 1898, after which the correc- 

 tion -I- .053 in. is appropriate. The following adjustments 

 for index error have therefore l:)een used in computing the 

 data: 



1866 —.016 in. 



1867-1894. . . .Graduated from —.016 in. to -[-.013 in. 



1895-1897 -I- .013 in. 



1898 Jan.-Oct. +.033 ; Nov.-Dec. +.0^3 in. 



1899-N0V. 1912 +-053 ii^- 



All values have been reduced to a common height of 53 ft. by the 

 application of corrections as shown below. 



Jan. Feb. Mar. Apr. May June July Aug-. Sejjt. Oct. Nov. Dec. Year 



Feb. 18.-)2-July ]So5 aud March 1856-Dee. 18G0 



+ + + ++ + ++ + + + + 



Inches -1-.008 .OdS .008 .008 .007 .007 .007 .007 .007 .007 .008 .008 .008 



Aug. 1855-Fcb. lsr,C, 

 Inches —.023 —.023 —.023 —.022 —.022 —.022 —.023 —.023 



Li addition an adjustment has been made to the individual read- 

 ings for the period February i852^June 1908 when only 

 two observations were taken each day in order to reduce 

 them to ;U7'' + i3'' + 2i'i). The corrections were deduced 

 from Lisbon pressure data 1800- 1919 given by Lima in 

 " O Clima de Portugal Continental " and are as follows : 



Jan. Feb. Mar. Apr. I\Iay Juno July Aug. Sept. Oct. Nov. Dec. Year 



— + + + + + + + + + + — + 



Inches .005 .003 .001 .007 .003 .004 .oof .000 .003 .001 .001 .006 .003 



