H Pp >) 
Favne. (Zms.) Atm. 
Max. 91.16 1.02796 
592 Min. 91.09 1.027 16 
Max. 182.78 1.0283 1 
1900 Min. 182.61 1.02732 
100 Max. 247.72 1.02809 
Min. 247.53 1.02741 
Heraf ser man, at de specifiske Vægter give Trykkene 
i Dybet med en Nojagtighed af 2 til 3 Millimeter Kvik- 
sølvtryk. Usikkerheden i Tyngdefactoren b medfører ingen 
Usikkerhed i Trykkene, der gaar op til en halv Millimeter. 
En sandsynlig Tilstedeværelse af en Fejl i Coefficienten for 
Søvandets Sammentrykkelighed medfører en Fejli Trykkene, 
der er ret merkelig. 
stant over hele Havet i et og samme Niveau, saa at Try k- 
Men denne Fejl er saagodtsom con- 
forskjellerne i Niveaufladerne blive ganske uafhængige 
deraf. Dette sidste gjælder i end højere Grad Coefficienten b. 
Fejlene i de beregnede Tryk-Forskjeller i Dybets 
Niveauflader blive saaledes praktisk talt kun afhængige af 
Fejlene i de beregnede Værdier af X. Forskjellen mellem 
Trykkene i to Punkter vil faa en sandsynlig Fejl af 
7==(d på å 
V2 Ta d 3, da Fejlene 1 hvert Punkt er + 
Disse sandsynlige Fejl blive altsaa: 
i 500 Favnes Dyb 1.51 V2 = 2.14 mm. 
1 WOOO — == YU = 30 
1 1500 — = YR VD = B78 
For at kunne bringe Nojagtigheden op til 1 Millimeter 
Kviksølvtryk, maatte den specifiske Vægt bestemmes 4 Gange 
nøjagtigere, altsaa være nøjagtig i 5. Decimal, eller An- 
tallet af Observationer af den specifiske Vægt være saa 
stort, at den sandsynlige Fejl af samme i et enkelt Punkt 
i Havets Dyb var 4 Gange mindre end den vi have op- 
naaet. Her tales kun om Dybder, der ikke synderlig over- 
skride 1500 Favne. 
Sammenlignes de sandsynlige Fejl af de beregnede 
Tryk med Størrelsen af Trykkene selv, og beregner man 
den tilsvarende sandsynlige Fejl af en Barometerhøjde paa 
760 mm, faar man 
for 500 Favne (fathoms): 1.51: 
(91.13 Xx 760) 
dp ) an) d 2) KD 
as)¢- TB db Bin dn Diff. 
mm. mm. mm. mm. 
pe , 2.37 
51 0.05 2.38 0.01 
I 0.20 6.39 
4 2 5.00 0.01 
.64 0.45 39 0.01 
8.70 
From this Table, we perceive that the specific gravi- 
ties give the pressure in the deep with a precision of 
from 2 to 3 millimetres mercury-pressure. The error in, 
the gravity-factor b does not involve any error in the 
pressure, which amounts to half a millimetre. The prob- 
able presence of an error in the coefficient of compres- 
sion for sea-water, involves an.error in the pressure quite 
appreciable. But this error is well-nigh constant through- 
out the whole extent of ocean at one and the same level, 
so that the differences of pressure at the surfaces of level 
is wholly unaffected thereby. This refers still more forcibly 
to the coefficient b. 
The errors in the computed differences of pressure at 
the surfaces of level of the deep, depend thus, prac- 
tically speaking, only on the errors in the computed 
values of 3. The difference in pressure between two 
points will have a probable error of V2 re dX, the error 
— 
dp 
—=|d 3S. These probable errors 
as 
at each point being + | 
are therefore: — 
at a depth of 500 fathoms 1.51 V2 = 2.14 mm. 
m —= 1000 = 2147 = 36% 
at, —— 1500 — 264 v2 = 3.73 
In order to reach a precision of 1 millimetre 
mercury-pressure, the specific gravity would have to be 
determined with quadruple precision, accordingly to the 5th 
decimal, or the number of observations for specific gravity 
be so great, that the probable error of such at any one 
point throughout the depth of the sea be 4 times less than 
that we attamed. Here the question merely relates to 
depths not exceeding 1500 fathoms. 
If the probable error of the computed pressure be 
compared with the amount of pressure itself, and if the 
corresponding probable error of a barometrical height of 
760mm. be likewise calculated, we shall get: — 
= 0.00002181 = 75257; Bar. 0.0166 mm. 
» 1000 3 2.14: (182.62 x 760) = 0.00001544 = 44,5 » 0.0117 
0 Å 2.64: (274.56 x 760) = 0.00001263 = 7517, » 0.0096 
Normalbarometer (Standard Barometer): 0.01: 760 =(0:0000181818T OL OL00 
Sammenligningen stiller sig saaledes gunstig for Nøj- 
agtigheden af de beregnede Tryk i Havets Dyb ligeoverfor 
Maalingen af Atmosfærens Tryk. 
Vil man tage et fuldere Hensyn til de forskjellige 
The comparison is@accordingly favourable to the pre- 
cision of computed pressure in the depths of the sea as 
contrasted with the measure of atmospheric pressure. 
Should we wish to pay greater regard to the various 
