Linen ind, hvorved Thermometret vendtes, Kviksølvtraaden 
splittedes og registrerede Temperaturen, og Haglene rullede 
ned i den anden Ende, der nu blev den tungere, og tjente 
til at sikre Instrumentet mod en ny Omvending. Til sin 
Accomodation behøver dette Thermometer kun 3 Minuter, 
men det er en nødvendig Betingelse, at det i denne Tid 
staar lodret eller næsten lodret med Kuglen nedad. Dette 
finder Sted, naar Trækassen har Opdrift i Vandet. Denne 
Egenskab have Trækasserne, naar de ikke have været ud- 
satte for sterkere Vandtryk. Men naar Trækasserne havde 
været paa større Dyb end et Par Hundrede Favne, viste 
Erfaring mig, at de mistede sin Flydeevne. Herom over- 
beviste jeg mig ved directe Forsøg. Naar en Trækasse 
kommer under et større Vandtryk, driver dette 
Træcellerne indesluttede Luft, der giver Træet dets Fly- 
deevne, ud, og Kassen mister sin Opdrift. En saadan 
Kasse vil, naar Lodlinen standser, strax vende sig om og 
registrere Thermometrets Temperatur. Hvorvidt denne er 
den samme som det omgivende Vands, beror paa Omstæn- 
dighederne. Dersom Temperaturen i Havets Dyb ikke 
forandrer sig eller kun forandrer sig ganske langsomt paa 
det Stykke, Thermometret gjennemløber i de sidste Minu- 
ter, vil dette kunne registrere den søgte Temperatur. Det 
er her at bemerke, at Accomodationen foregaar forholdsvis 
Da 
Dyb- 
bero 
den i 
hurtig under Thermometrets nedadgaaende Bevægelse. 
det var umuligt at faa gjort nye Trækasser til hvert 
lodskud, ere vore Bundtemperaturer, der væsentlig 
Negretti-Zambra'ske Thermometre, bestemte efter 
fremsatte Princip og kunde derfor muligens være 
Der er imidlertid, som nedenfor 
paa de 
det her 
lidt for højt angivne. 
skal vises, saa god Overensstemmelse mellem den ved Ne- 
eretti & Zambra’s Vendethermometre gjorte Bestemmelse af 
Trykcoefficienten tor Buchanans Kviksølvpiezometer og den 
af Buchanan selv opgivne, at Virkningen af Forskjellen 
mellem begge paa 2000 Farnes Dyb ikke gaar op til mere 
end 091, og Kviksølvpiezometret antyder endog en Correc- 
tion i modsat Retning af den som ovenfor er antydet. 
En anden Control haves 1 de Bestemmelser af Indexther- 
mometrenes Trykcorrectioner, der ere udførte 1 Dybet un- 
der den atlantiske Strøm, jevnførte med dem, der ere 
udførte i Polarstrømmen. I den første aftager Tempera- 
turen raskere mod Dybet end i den sidste. De nævnte Be- 
stemmelser ere gjorte med Negretti & Zambra's Vende- 
thermometre som Normal, og give, som Resultat for 2000 
Favnes Dyb, i Gjennemsnit —0°.31 for den varme Strøm 
og —0.°21 for Polarstrømmen. Dette peger i Retning af en 
mindre fuldstændig Accomodation i Polarstrømmen end 1 
Dybet under den varme Strøm, et Forhold, der strider 
imod det virkelige, og saaledes kun tjener til at bestyrke 
Antagelsen af en fuldstændig Accomodation. Herom se 
yderligere nedenfor. 
15 
we hauled in the line, whereby the thermometer turned 
over, the thread of mercury broke and registered the tem- 
perature, and the shot rolled down to the other end, which 
now became the heavier, and thus kept the instrument 
from turning over again. For accomodation, this ther- 
mometer requires a space of 3 minutes; but it is a neces- 
sary condition that during this time the instrument. shall 
keep perpendicular, with the bulb downwards. And such 
will not fail to be the case provided the wooden frame 
has the necessary buoyancy in the water. This property 
the should it 
exposed to any considerable pressure of water. 
frame is found to have, not have been 
But when 
the wooden eases had been lowered to some depth, 
for instance more than 200 fathoms, they lost their buoy- 
ancy. Of this I obtained proof by direct experiment. 
When one of such wooden cases has to sustain a great 
pressure of water, this expels the air in the pores of the 
wood, which gives the latter its buoyancy, and the case loses 
the property of floating. A frame in this condition will, on 
the stoppage of the sounding-line, immediately turn over and 
register the temperature of the thermometer. Meanwhile, 
whether it be the same as that of the surrounding water, 
depends upon circumstances. Provided the temperature in 
the depths of the sea do not vary, or vary but very 
slowly in the strata passed through by the thermometer 
during the last few minutes previous to stopping, the instru- 
ment will be able to register the temperature sought to 
be measured. I must remark here, that the accomodation 
proceeds quickly on the downward passage of the ther- 
mometer. As new wooden frames could not possibly be 
got ready for every deep-sea sounding, our bottom-temper- 
atures, which ultimately depend on Negretti & Zambra’s 
inverting-thermometers and are determined on the prin- 
ciple here laid down, may therefore, possibly, have been 
given a trifle too high. Yet there is, as will be shown 
farther on, such excellent agreement between the deter- 
mination of the 
mercury-piezometer computed from Negretti & Zambra's 
inverting-thermometers and that given by Buchanan him- 
self, that the effect of the difference at a depth of 2000 
and the 
mercury-piezometer indicates even a correction of an opposite 
character to that suggested above. Another mode of control 
coefficient of pressure for Buchanan’s 
fathoms does not amount to more than 0.°1; 
hes in the determinations of the index-thermometers’ pres- 
sure-corrections — which are made in the deep under the 
Atlantic Current — as compared with those made in the 
Polar Current. In the former current, the temperature 
diminishes more rapidly with the depth than in the latter. 
The said determinations are performed with Negretti & 
Zambra’s inverting-thermometer as standard, and give an 
average result, for a depth of 2000 fathoms, of —0.°31 
for the Warm Current and of —0°.21 for the Polar Current. 
This points towards a less perfect accomodation in the 
Polar Current than in the depths beneath the Warm 
Current — in direct opposition to fact, which thus 
tends to confirm our assuming a perfect accomodation. — 
Respecting this subject, see further on. 
