6 FRIDTJOF NANSEN. M.-N. Kl. 
Jena Glass No. 5g!ll, and protected by an outer strong glass tube. The 
stem was filled with an inert gas, nitrogen or carbonic acid, in order to 
prevent the division of the mercury thread which was found very trouble- 
some during the voyage with the Michael Sars, in 1900. This gas in the 
stem proved a great improvement. The scale ranged from — 3 to + 8° C. 
in No. 110, and from — 3 to about 10" C. in No. 109. - Both scales were 
divided into tenths of degrees centigrade. Å degree had on No. 110 
a length of ı cm. but was in No. 109 somewhat shorter. 
Both instruments were tested at the Charlottenburg Reichsanstalt in 
March, 1901, and were marked PTR 15655 (for 109) and PTR 15656 
(for No. 110). 
The following determinations of the corrections were made. 
Corrections. 
| | Deep-Sea 
| Deep-Sea Thermometer No. 109. 'Thermom. 
I Nowe: 
Scale. || | 
| March 14, | July ı, Septb. 5. | Septb. 12, Septb. 13, | March 14, 
| 1901 | I 90 I I 901 | 1901 | 1901 | IQOI 
| Charlottenb. | Gjoa. | Gjøa. Christiania. | Christiania. | Charlottenb. 
| | | 
— 3°C.| + 0'06° C | | | + o'or 
©, #005 | fox OE Ca) 0.105) SG | 0'00 
FÅ oy |+ 007 ” | | + 0'02 
+8, | | + 0'02 
+ 10, 4- 007°, | | 
Thermometer No. 110 was lost with the Pettersson-Nansen Water- 
Bottle in May 1901. 
It seems somewhat puzzling that No. 109 should have indicated so 
much lower during the voyage than it did before and afterwards. 
Amundsen says in his journal, that on Juli 1, a piece of freshwater ice 
drifting in the sea, was used for the zero-determination. It may seem 
doubtful whether this ice has been sufficiently pure; but according to 
zero-determinations made with other thermometers in the same ice, simul- 
taneously, it does not seem probable that the ice can have been impure 
enough to have caused any significant error. On the other hand it seems 
more probable that Amundsen has got fairly pure ice in Tromsø for 
his determination on September 5, 1901. 
Thermometer No. 638. 
This was a thermometer which Amundsen had from the Meteoro- 
logical Institut at Christiania for taking surface-temperatures. Its zero- 
| AVR 
