WHOLE VOL. NOTES. SOUTTT AMERICA I 15 



The temperature data for Galera are the mean of the max.4- 

 min.( ?) 



The earlier Santiago data, 1861-1891 were means of 7'\ 14'' 

 and 22^'' corrected to mean of 24 hours, l)oth ftjr pressure 

 and temperature. After i8q2 the data are the means of 24 

 hours throughtjut the .series. 



Data for the years i(;03 to 1906 at Iquique were derived from 

 observaticjns taken every two hours pr(jl)ahly measured from 

 a thermograph trace controlled by eye readings. 



Data for the years 1900, 1901 and 1902 were from mean of obser- 

 vations taken at 8''-i4''-2['' brought to the mean of 24 liours 

 by the following corrections derived from com|)arison with 

 the above two hourly observations from January 1904 to 

 July 1906 as follows: 



Data for the years 1907 to 1910 are from the mean of the max. 

 and mill., corrections having been determined from a com- 

 ])arison of the mean max. and min. with the means of 



^ "^ ~- — which figure in the tal)U- for the period 



4 



191 1-1924. The years selected for obtaining the corrections 

 were 1911 to 1914, the following being the corrections ap- 

 plied, to the mean of the max. and min. : 



It is not possible to use any one series of data throughout, as pre- 

 vious to 1911 in common with all other stations in Chile 

 except Santiago and Punta .A^renas, different systems of 

 hours were in vogue, and the minimum thermometer often 

 out of order or broken for long periods. This latter is a very 

 frequent defect with the Chilean data i)revious to 1911 and 

 was due apparently to a lack of systematic inspection at the 

 coastal stations. Fortunately the detection of errors in the 

 minimum thermometer is easy as it is at once apparent from 

 the increased mean daily range. 



SANTIAGO 

 Pressure. 



Earlier data. 1861-1892, are the means of observations at 7'', 14'' 

 and 22^^ corrected to the mean of 24 hours, then the mean 

 of hourly observations till I<;i5. From i()i6 to 1921 they 



