vviiuLi: \(ji.. NOTi';s. south ami:rica 113 



moved to the l\ilacio dos Estados where the heij^ht of the 

 barometer was 18.3 m. above sea level. A correction of 

 — '^f.y was applied to the observed mean values to reduce 

 them to the old level of 61.4 m. 



Hours of Ohscn'ofioJi : The hours of observation are not given 

 but are stated to be 24 hour means in the Reseau Mondial. 



Xofc: A gravity correction of —1.4 mm. was applied to all the 

 mean values to reduce them to Lat. 45°. 

 Ti:mi>1':raturf<:. 



Antlwrity: Robert C. Mossman. 



Notes: The series of cUmatological observations taken at the 

 Astronomical observatory of Rio de Janeiro from 185 1 to 

 1890 have been summarised by Senhor E. Cruls in his 

 well-known work, O Clevira do Rio de Janeiro (Rio de 

 Janeiro 1892). The temperature data were not reduced 

 to a homogeneous system, and there were different .systems 

 of e.x]:)Osure and hours of observation. The foregoing data 

 are from observations after 1870 when the exposure of 

 the thermometers was changed and has remained constant 

 to date, vi::. in a large pavilion with wooden louvres. The 

 hours of observation were 7, 13, and 17'' from January 187 1 

 to December 1873 ; 7, 10, 13 and 16'^ from January 1874 

 to June 1879; 4, 7, 10, 13, r6, 19, and 22'' from July 1879 

 to June 1885 and afterwards at three hourly intervals begin- 

 ning at j''. The reductions to true mean of 24 hours were 

 obtained from a comparison of the values at each of the 

 above combinations of hours with the mean of the 8 observa- 

 tions per day given //; cxtcuso in Bolct'un Mcnsal for the 

 bve years 1900-1904 the corrections being as follows: 



Minus correcti(Mis in Tenths of a Degree C. to bring observa- 

 tions to true mean temperature. 



.Jan. Fill. _\!;ir. Spv. May -liinr .Inly Aiit;. Sop!. Oct. N<i\. Dec. 



I'.Tiii.l 



.laii. 1871 -Dec. 1878 fi ;"> 1 \ i f, r> fi .'') r, ."S r. 



•Jan. ]87t-.Juiie ^879 9 9 G r> 4 6 6 6 6 fi 8 9 



July 1879-June 1885 2 2 11]llliixj 



'J'he original data have all ha<l the above corrections aijplied 

 before entering in the table. J am of the opinion that the 

 minus corrections should be increased liy 0.4° or 0.5° up 

 to June 1879 ^-^ ^'^6 data, /. e. mean monthly temperatures, 

 seem too high but 1 have let the values stand as corrected 

 above. 



