4 FRIDTJOF NANSEN. M.-N. Ki. 
+ indicates that the error has to be added to the observed temperature 
in order to find the correct temperature. By several determinations it was 
also found that if the mercury be broken off at temperatures about zero, 
the correction due to the difference of temperature at which it is read off, 
is about 0'0098° C. for each degree the broken-off mercury is cooled or 
heated. 
Amundsen made one observation of the zero-correction of the instru- 
ment during the voyage, on July ı, 1901, but as he himself remarks it 
is of no value as it was too difficult to make the mercury break off. As 
the thermometer was made of Jena normal glass 1611 it may be assumed 
that it has altered its corrections fairly gradually during the time of the 
voyage, and as the instruments were, as a rule, exposed to lower tempera- 
tures than in the laboratory it seems probable that the corrections were 
if anything somewhat smaller than as indicated above. It may therefore 
be assumed that the zero-correction during the voyage was very nearly 
+ 0'06° C. 
Negretti and Zambra Reversing Thermometers Nos. 72012 and 72620. 
These thermometers were several years old, and were of the ordinary 
type delivered by Negretti and Zambra. They were made of ordinary English 
glass, and the scales were devided into whole degrees. The graduation 
was rather rough, so that it was somewhat difficult to read off the 
temperature accurately, with a reading lens, or even with a reading 
microscope. 
The following determinations of the zero-correction were made. 
Zero-corrections. 
Date. N. Z. No. 72012. NZ. Now 72020: 
March, 1901, Christiania. . . . = ore) C. — 016 ° C. 
JON own Giga. FN — 0135 » —AO'T4 0, 
September 5, 1901, Gjøa, Tromsø OO — 0:20 17 
September 12, 1901, Christiania . — 0'136 , 
September 13, 1901, Christiania . — 016 , — 012 , 
The mercury broken off at zero expanded by an amount equal to 
about o‘or * C. of the scale, for each degree it was heated. 
