mår Fara NG ry sv 
å 
1906. No.3. AMUNDSEN’S OCEANOGRAPHIC OBSERVATIONS IN 1901. II 
the samples were taken; whilst on May 9, in the Barents Sea, very little 
ice had melted, though much ice had been formed and exposed to low 
temperatures, in this region, during the winter.’ It is also a striking 
coincidence that in Makaroffs samples from the upper water-strata of the 
sea near Franz Josef Land and between these islands and Novaya Zemlya 
(see below), the chlorine values also give higher salinities than the specific 
gravities, and the discrepancy is greater than might be explained by errors of 
observation. It is also striking that the determinations of Makaroff’s samples 
from the deeper waters, from 300 metres Stat. 77, and from 100 and 200 
metres Stat. 82, agree very well, the titrations giving slightly too low 
salinities. The. values obtained for the samples from 250 and 300 metres 
Stat. 83 agree also fairly well, but those for the sample from 350 metres 
show a greater discrepancy. The determinations by titration and by 
Hydrometer of Total Immersion of the samples from Wollebæk's Stat. II 
(see below) agree also fairly well, the discrepancy being between + o'or 
%oo near the surface (10 metres) and — o'or5 °/oo near the bottom (120 
metres). We see that in this region of the sea, the disagreements go on 
the whole in the same direction, the chlorine gives comparatively high 
values of salinity near the surface, but comparatively lower salinities for 
deeper water. 
The Accuracy of Determinations of Temperature and Salinity in the 
Vertical Series at the Stations. 
The temperatures taken with the Richter “Reversing Thermometer No. 
113, are probably correct to within a few hundredths of a degree Centi- 
grade. The instrument was made of very good glass (Jena Glass 59"", 
see above), was divided neatly, by very thin and distinct marks, into fifths 
of a degree centigrade, and the indications could easily be read off by 
means of a specially arranged lens, with an accuracy of + o'01* C. And 
by means of the enclosed small thermometer the temperature of the broken 
off mercury was always determined at the moment of reading. 
1 By some rough experiments made at the Central Laboratory for the International Study 
of the Sea, at Christiania, we have found that where ice is formed in sea-water no 
appreciable alteration is produced in the relation between the quantity of chlorine and 
the specific gravity of the sea-water, although the salinity of the latter be much in- 
creased. The recent investigations of Dr. W. Ringer: „Ueber die Veränderungen in 
der Zusammensetzung des Meereswassersalzes beim Ausfrieren“ (No. III: Verhandelingen 
uit het Rijksinstituut voor het onderzoek der Zee, 1906, Helder) prove that it is only 
after the ice is exposed to. temperatures below — 8°o C. that the relation between 
the chlorides and the other salts (especially sulphates) in the brine of the ice, is 
appreciably altered. 
