NO. n.] 



The above numbers lead to the equation 



Corr. = + 0-337 - 0-0127 t + 00012 (760 - h) mm. 

 The last term becomes, when 6 = 730, -f 004 mm. 

 — - 6 = 780, — 0-02 - 

 The correction may be taken as independent of the barometric height. 

 In 1893 we had Corr. = + 0295 — 00151 . t 16 obs. 

 - 1897 - - - = + 0-337 — 0-0127 . < 86 - 

 The weighted mean of the temperature-coefficient is — 0-0131 per degree 

 G. The correction at 0° is greater by 0-042 mm. in 1897 than in 1893. 

 The mean, which I have adopted, is + 0-32 nun. It differs from each 

 of the two values by only ± 0-021 mm., which is practically insignificant.^ 

 The station barometer Adie 763 was used for the current observations 

 from noon, October 12, 1893, to 7 p. m. August 14, 1896. The height of 

 its cistern above the level of the sea was 0-93 metres. The reduction of the 

 height of the mercury to sea-level can be taken as + 0-09 mm., with no error 

 exceeding + 0-05 mm. for any pressure or temperature of the air. 



The observations taken with Adie 763 have been reduced to the true 

 height of the mercury at 0° and to sea-level by the Correction 



+ 0-32 4- tabular (International) reduction to 0° — 00131 t + 0-09 mm. 

 Reduction at 0° = + 0-32 + 0-09 = + 0-41 mm. 



III. The Marine Barometer Adie No. C 764, was used when the 

 Fram was in open water. 



The 16 comparisons in 1893 at the Institute gave, for 2 groups, 

 Att. Therm. Corr. M. E. 



5-2° + 0-11 mm. + 0-099 mm. 8 obs. 



13-3 — 0-14 - ± 0-079 - 8 - 



The comparisons on board gave, at 11°, + 0094 mm. The adopted correction at 11° 

 is + 0-17 mm., dift'erenee 0'08 mm. As the comparisons in Christiania were made 

 under much more favorable circumstances than on board the Fram, I have adopted 

 the above-mentioned vahie, + 0-32 mm. The check obtained by the observations on 

 board may be considered very satisfactory. Tlie Kew correction was -f- 0'12 mm, 

 whicli corresponds to a temperature of 18 ° C. 



