76° N. Br., 14° E. Led., en Temperatur i Havfladen af 
over 4°. Her var aabenbart en Tungespids for 4°-Iso- 
thermen. — Samtidig var Temperaturen ved Fruholm 4°, 
hvilket er omtrent den normale Temperatur for Aarstiden. 
Ved Fruholm er den normale Temperatur for Marts 20.4, 
eller 29.0 lavere end Normaltemperaturen for den 24. No- 
vember. En ligefrem Reduction giver Temperaturen for 
Marts paa “Alberts” ovennævnte Plads lig lidt over 2°. Imid- 
lertid fandt Capt. Otto den 7. December under 71° N. Br. 
1605 E. Led. en Temperatur af 60.5, hvilket er 09.5 højere 
end Aarets Medium paa dette Sted. Aarsmediet for Fru- 
holm er 59.0, og Mediet for den 7. December 40.0, altsaa 
skulde den observerede Temperatur paa “Alberts” sidst- 
nævnte Sted være 195 for højt eller den normale Tempe- 
ratur her for den 7. December 59.0 og for Marts (4°—2°.4) 
19,6 lavere eller 3°.4, et Tal, som stemmer godt med mit 
endelige Kart. Jeg setter derfor Temperaturen for Marts 
under Sydkap lavere end 2°, og lægger Tungespidsen for 
Isothermen for 2° paa lavere Bredde, paa 7595.  Forøvrigt 
har jeg ladet Isothermen for 2° danne en Tunge, hvis Axe 
falder sammen med den af *Alberts” Observationer frem- 
gaaende Tunge for 4°. 
Til Constructionen af Marts-Isothermerne for Østhavet 
og Danmarkstrædet har jeg, da Observationer ganske mangle, 
maattet benytte en anden Fremgangsmaade. Idet jeg har 
støttet mig til Observationerne fra Kyststationerne og Be- 
tragtningen af Isgrændserne tidlig om Vaaren, samt efter 
Analogi med Temperaturens Fordeling i Havets Overflade 
om Sommeren og 1 Luften om Vinteren, har jeg først gjort 
et Udkast til Havisothermernes Lob i Marts. Efter dette 
Udkast og August-Isothermerne beregnedes den aarlige Va- 
riation, og bleve Linier for ligestor aarlig Variation forsøgt op- 
construerede i Kartet. Det noget uregelmæssige og mindre 
sandsynlige Lob af disse Linier corrigeredes saaledes, at 
Variationslinierne fik et regelmæssigt Lob (se Pl. XXIX). 
Ved Hjelp af disse nye Linier for ligestor aarlig Variation 
kunde jeg derpaa gaa directe over fra August-Isothermerne 
til de paa Kartet Pl. XXVIII fremstillede Isothermer for 
Havoverfladen 1 Marts. 
Til Isothermerne for Nordsøen har jeg benyttet Obser- 
vationer fra norske Handelsfartojer. 
Aarets Middeltemperatur har jeg beregnet paa føl- 
gende Maade. Kaldes Middeltemperaturen for August M, 
for Marts m og for Aaret Å, saa kan man sætte 
A= % (M+m) — f (M—m) 
hvor f er en constant Factor og M—m den aarlige Varia- 
tion fra Marts til August, fra den koldeste til den varmeste 
Maaned, Factoren f udledes af Observationerne fra Kyst- 
stationerne. Hver af disse giver en Ligning af ovenstaaende 
Form, og man faar, naar X betyder Summationstegn, 
SA= 3 4(M+m) — f 3 (M—m) 
102 
Cape, Spitzbergen, lat. 76° N, long. 14° E, a temperature at 
the sea-surface reaching upwards of 4°. This was clearly the 
extremity of the tongue formed by the 4° isotherm. The 
temperature at the same time at Fruholm was 4°, which is 
about the normal temperature for that time of year. At 
Fruholm, the normal temperature for March is 2°.4, or 
200 lower than the normal temperature for the 24th of 
November. A direct reduction of the temperature at the 
forestated position of the * Albert,” gives for March a trifle 
over 2°. Meanwhile, Capt.- Otto found, on the 7th of De- 
cember, in lat. 719 N, long. 16°.5 E, a temperature of 6°.5, 
or 0°.5 higher than the annnal mean determined for that 
locality. The annual mean for Fruholm reaches 5°.0, 
and the mean for the 7th of December 4°.0: hence the tem- 
perature observed by the “Albert” at the last-mentioned place 
should be 195 too high, or the normal temperature here 
for the 7th of December be 5°.0 and for March (4°—2°.4) 
19,6 lower, or 39,4, a result exhibiting very satisfactory 
agreement with my finally constructed map. Hence, I put 
the temperature for March, in close proximity to South 
Cape, lower than 2°, and draw the point of the isothermal 
tongue for 2° on a lower parallel of latitude, viz., 759.5. 
For the rest, I have let the isotherm for 2° form a tongue 
the axis of which is congruent with the tongue for 4° 
resulting from the observations of the “Albert.” 
In order to construct the March-isotherms for the Ba- 
rents Sea and Denmark Strait, I had necessarily, obser- 
vations being wholly wanting, to adopt another mode of 
operation. Taking regard to the observations from the Coast- 
Stations, as also to the ice-limit early in the spring, and 
in accordance with analogy as shown by the distribution of 
temperature at the surface of the sea in summer and in 
the atmosphere during winter, I first drew a rough sketch 
of the course taken by the sea-isotherms in March. From 
this sketch and the August-isotherms, I computed the an- 
nual range, and drew lines for an equal annual range on 
the map. The somewhat irregular, and comparatively im- 
probable, run of these lines was corrected by giving the 
lines of range a regular course (See Pl. XXIX). With 
these new lines for an equal yearly range, I could then pass 
directly over from the August-isotherms to the isotherms 
drawn on Pl. XXVIII for the sea-surface in March. 
' When constructing the isotherms for the North Sea, 
I made use of the observations taken by Norwegian 
merchant vessels. 
Now, calling the mean temperature for August M, 
that for March m, and that for the whole year A, we 
can put 
A= 4 (M+m) — f (M—m), 
in which formula f is a constant factor and M—m the an- 
nual range from March to August, from the coldest to 
the warmest month. The factor f is determined from the 
observations taken at the Coast-Stations. Each of these 
gives an equation of the above form; and we get, with 3 
as the sign of summation, 
A= 3 3(M+m) — f I (M --m), 
