14 
at være 43.6 Milliontedele. Herefter faar man | 
No = (43.6 + 2.7825 + 0.0962) 106 = 46.4787 x 106, | 
J. Y. Buchanan, Challenger-Expeditionens Chemiker, 
har fundet, at Sovandets Coefficient er 92.3 Procent af det 
rene Vands. Da dette kan settes! til 50.153 x 10 ved 
0° og almindeligt Lufttryk, faaes heraf 1,— 50.153 x 0.923 x 
NOS = BOP SK NOs, 
Tait? har fundet den samme Procent at vere 92.5, 
hvoraf », = 46.392 x 106 V 
Middel af disse Bestemmelser er 46.387 x 10%, 
Den følgende lille Tabel giver en Oversigt over Stør- 
relsen af Factoren 7 1 Formelen for p paa næste Side i 
Milliontedele, ved forskjellige forekommende Temperaturer 
I 
og Tryk. | 
Dybde i Fayne. t 
(Depth in Fathoms.) 
(0) 5°.0 
) 0.0 
416 2.0 
1333 0 off 
1985 —1.7 
» vil saaledes falde mellem 45 og 46 Milliontedele. | 
- | 
I mine Beregninger har jeg antaget 7, constant lig 45 « 10%, | 
Den heraf flydende Fejl har, som senere skal vises, 
ingen stor Indflydelse paa det beregnede Tryk. end mindre 
paa Tryk-Forskjeller i samme Dybde-Niveau. | 
Er 1 Dybden h Favne Trykket af det overliggende 
: | 
Vand p Atmosfærer, og er Søvandets Tæthed, ved alminde- | 
oO 
saa er dets Tæthed, i Dybden h, —“—. 
G 1—7p | 
Det Tryk, dp, som en vertical Vandsøjle af Højden dh | 
udøver ved sin Gravitation, er proportionalt med Tyngdens 
Størrelse. Man faar saaledes 
ligt Lufttryk, /,, 
S, SE 
20 ple = ee 
1—np 045 1—np 
Of) = 
For at kunne integrere denne Ligning maatte man | 
kjende den Lov, hvorefter Tætheden S, varierer med Dyb- 
den. Som Snittene Pl XXXIX til XLI vise, er denne 
forskjellig i de forskjellige Verticallinier. Forskjellerne ere 
imidlertid ikke store, og saavel numeriske Beregninger som 
theoretiske Betragtninger, hvis Resultat senerehen skal 
meddeles, vise, at man kommer til den ønskede Nøjagtig- 
hed, om man regner med en constant Værdi af Vandets 
Tæthed og sætter denne lig Middeltallet af Tæthederne i de 
I 
i 
Travaux et mémoires du bureau international des poids et 
mesures, Tome IT, D. 30. 
2 
> Proceedings of the Royal Society Edinburgh, 1883, S. 224. 
} 
| 
| 
7 
gravity of 1.0264, to be 43.6 millionths. 
result we get 
Np = (43.6 + 2.7825 +. 0.0962) 10% = 46.4787 x 106, 
Mr. J. Y. Buchanan, Chemist to the Challenger Ex- 
pedition, found the coefficient of sea-water to be 92.3 per 
cent compared to that for pure. Now as this may be 
put? = 50.153 x 109 at 0° and ordinary atmospheric pres- 
sure, we get n, = 50.153 x 0.923 x 10 *= 46.291 x 106. 
Tait? has found the same percentage to be 92.5, 
whence 1, = 46.392 x 106. 
The mean of these determinations is 46.387 x 10—*. 
The following short Table will give a general view 
of the value of the factor 7 in the formula for p, next page, 
in millionths, at different actual temperatures and pressures. 
According to this 
p ? 
Ot 45.58 
O 46.39 
75-968 45-74 
244.482 45-43 
365.c08 45.03 
Thus » will vary between 45 and 46 milliontbs. In 
my computations, I have regarded 1 as constant, and equal 
to) 24) SX UO, 
The errors arising therefrom have, as will subsequently 
be shown, no considerable influence on the computed pres- 
sure, and far less on differences of pressure throughout the 
same level. 
If, at the depth h fathoms, the pressure of the su- 
perincumbent water is p atmospheres, and if the density 
| of the sea-water at ordinary atmospheric pressure is S,, 
0 
then its density at the depth Å will be — The pres- 
1—7 P 
sure, dp, which a vertical column of water of the height 
dh exerts by its gravitation, is proportional to the force 
of gravity. Hence we get 
(1 — 8 cos 2) (1 +0. h) dh. 
In order to integrate this equation, it is necessary to 
know the law according to which the density S, varies 
with the depth. As shown by the sections Pl. XX XIX 
to Pl. XLI, this differs along the different vertical lines. The 
differences however are not considerable, and alike numerical 
computations and theoretical considerations, the result of 
which will be subsequently given, clearly prove that the 
desired accuracy is reached, if we calculate with a constant 
value for the’ density of the water and put the latter 
1 Travaux et mémoires du bureau international des poids et 
mesures, Tome II. D. 30. 
2 Proceedings of the Royal Society Edinburgh, 1883, p. 224. 
19* 
