eller den sandsynligste Værdi af f 
pe XI +m) — I 4 
Få X (M—m). p 
Til Bestemmelsen af Factoren f har jeg udvalgt føl- 
gende Stationer, der nærmest kunne repræsentere det aabne 
Nordhav: 
103 
or the most probable value of f, 
yee X 3 (M+m) — 5 4 
TS 
For determining the factor f, I made choice of the 
following Stations, which may be taken as most nearly re- 
presenting the conditions of the Norwegian Sea at large. 
P M m Ay M-m å (M-+m) Jf (M-m) VAR A,-A, 
Lindesnes RO BOAT 8°.7 12°.6 9°.4 0°.8 8°.6 oT 
Udsire ER Å 42 9.0 Rit 2 9.8 | 0.7 9.1 -0.I 
Hellisø 13.9 45 8.7 9.4 9.2 0.6 8.6 JEG oj 
Ona 13.0 4.0 8.0 9.0 8:5 | 0.6 dg OD 
Fruholm 8.5 2.4 5.0 6.1 5 8 | 0.4 B olf -0.1 
Fast Yell 1D, 2 7] PD 9.4 5.0 9-7 0.3 9.4 0.0 
Sandwick 12.9 6.7 93 6.2 9.8 0.4 9.4 -0.I 
Thorshavn 10.8 5.4 7.9 54 8.1 0.4 7.7: 0.2 
Papey 6.8 Oo BES 6.1 3.8 0.4 3.4 -0.I 
Grimsey 8.1 I .4 4.6 6.7 4.8 0.4 4.4 Lo .2 
Jan Mayen 2.8 -1..6 0.0 4 +4 0.6 0.3 0.3 -0.3 
Sum 73.9 82.1 79.2 0.0 
. 79:2—73:9 JG 
Vi 82.1 Bat 01005 
Den midlere Afvigelse mellem 
regnede Aarsmedier bliver + 0913. 
Man har saaledes til Beregningen af Aarsmediet 
de observerede og be- 
| 
The mean deviation between the observed and the 
computed annual means, will be + 09,13, 
Hence, for computing the annual mean, we have 
A = 3 (M+m) — 0.065 (M—m) = m + (0.5—0.065) (M—m) = m + 0.435 (M—m). 
Beregningen har jeg udført paa grafisk Vej, idet efter 
Kartet Pl. XXVIII Curverne for Marts, for Parallel- 
grader i Grønlandshavet og det norske Hav og for hver 
5. Længdegrad 1 Østhavet, sattes under de for August 
tidligere optrukne. Med en Proportionspasser, stillet efter 
Forholdet 1 : 0.435, toges paa den større Side Værdierne af 
M—m, Passeren vendtes om og fra Martsordinaterne m 
afsattes med den anden Side af Passeren opad Værdien 
0.435 (M—m). Herved erholdt jeg en Række Ordinater 
for Å, der forbandtes med en Curve. Fra denne 
afsattes dens Skjæringspunkter med Horizontallinierne for 
de hele Grader paa et Kart, og Isothermerne for Aaret 
indtegnedes gjennem disse Punkter. Ved Hjelp af Kyst- 
stationerne og for Nordsøens Vedkommende Resultatet af 
Skibsobservationer fra alle Aarets Maaneder completteredes 
Kartet. Saaledes er Kartet, Pl. XVI, over Havoverfladens 
aarlige Middeltemperatur kommet istand. 
Om Nojagtigheden af dette Kart for de Strækningers 
Vedkommende, hvor der foreligger Observationer baade for 
August og for Marts (det norske Hav og en Del af Grøn- 
landshavet)kunne følgende Beregninger anstilles. — Sættes 
Usikkerheden af saavel August- som Marts-Kartets Tempera- 
turangivelser for et enkelt Punkt i Havet til + 0.38, saa 
bliver den deraf flydende Usikkerhed i Aars-Kartet + 09.27. 
The caleulation I made diagrammatically, placing the 
curves for March drawn from the isotherms, Pl. XX VIII, 
under the curves previously drawn for August, for the par- 
allels of latitude in the Greenland Sea and the Norwegian 
Sea, and for every 5th degree of longitude in the Barents 
Sea. With a pair of proportional compasses, set at 1 : 0.435, 
I first measured, with the long legs, the values of M—m; 
the compasses were then inverted, and from the March ordin- 
ates m I marked off upwards, with the other or shorter 
legs of the compasses, the value 0.435 (M—m). In this way 
I obtained a series of ordinates for A,, which were com- 
bined into a curve. From the latter, its points of inter- 
section with the *horizontal lines for the whole degrees were 
marked off on a map and the isotherms for the year drawn 
through the said points. By means of the Coast-Stations, 
and, as regards the North Sea, the result of merchant ships’ 
observations for all months in the year, the map was 
completed. Thus it proved practicable to construct the 
map, Pl. XVI, showing the annual mean temperature of 
the sea-surface. 
; Concerning the accuracy of this map in regard to the 
tracts where observations had been taken both for August 
and for March (the Norwegian Sea and part of the Greenland 
Sea), the following calculations can be made. Putting the un- 
certainty of the statements, alike in the maps for August and 
March, with respect to the temperature for a single point of the 
sea at + 0°.38, the uncertainty thence resulting in the map for 
