62 FRIDTJOF NANSEN. M.-N. KI. 
temperatures below zero centigrade (at 200, 220, and 250 metres) but 
his salinities are obviously about 0'04 9/,,, too high; they have here been 
reduced accordingly, and the values thus obtained agree very well with 
those of Amundsen’s Stations. At Station IV, Amdrup also used a 
Sigsbee Water-Bottle and a reversing thermometer; but he himself points 
out that these instruments have possibly not worked satisfactorily; both 
temperatures and salinities seem improbable, which is also indicated by 
the densities; it was therefore thought advisable to leave these observa- 
tions out of consideration. 
During the Kolthoff Expedition, on board the Frithjof, to the 
East Greenland Coast, Mr. Östergren! took, in July 1900, two 
interesting Stations (Æ I and F II, Pl. V) with deep-sea observations to 
the north of Jan Mayen and between Spitsbergen and Greenland. The 
temperatures were taken with a Pettersson-Nansen Insulated Water-Bottle, 
of the first model made by L. M. Ericsson & Co. in Stockholm, in 1900; 
but without the Nansen Deep-Sea Thermometer. The temperature readings 
were taken by a thermometer inserted after the water-bottle came on 
deck, which prejudices the accuracy of the observations. Nor is it stated 
whether any correction has been introduced for the reduction of temperature 
caused by alteration of pressure. It is stated that the bottle was hauled 
up with a velocity of 1000 metres in 10 or 15 minuter. But if it has 
taken as much as 30 minuter or more to haul the bottle up from 2000 
or 3000 metres, the readings obtained cannot be trustworthy, as experi- 
ments have proved that the insulation of the bottle is not sufficient to 
keep the temperature-readings unaltered for such a long time. It is 
thus seen that the temperatures cannot be very accurate, at all events 
from the deep strata. The readings may farthermore have been too low 
owing to cooling by expansion of the water and the solid parts of the 
water-bottle (especially the india-rubber) on hauling up from great 
depths. Is seems, however, more probable that they have been too 
high, owing to deficient insulation during the long period the bottle was 
being hauled up, and owing to the insertion of the thermometer after 
the bottle had come up on deck. It may therefore be expected that 
the temperature of —1°23° C. at 3100 metres, at Station I (77° 11’ N. 
Lat., 0° 55’ W. Long.) is somewhat too high, in spite of the considerable 
cooling caused by hauling the instrument up from this depth. 
Much worse than the inaccuracies of the deep-sea temperatures thus 
caused, is, however, the fact that the water-bottle has not worked properly 
1 O. Pettersson and Hj. Östergren, Vattenprof tagna under "1900 Ars Svenska 
Zoologiska Expedition”, Ymer, vol. XX, Stockholm, 1901, pp. 325—329. 
