208 
med følgende Resultat: 
No. p beregnet 
(comp.) 
I 16.62 
2 76.66 
3 223.15 
4 , 234.45 
5 
6 
at. 
244.69 
245.88 
7 QI 2022 
8 292.69 
9 309-19. 
(0) 305.27 - 
Og af de saaledes beregnede Tryk beregnes Dybden 
efter Formelen: 
/ 
with the following result: — 
p efter Lodskud 0— B 
(by sounding) 
16.95 at. +-0.33 at 
75:97 —0.60 
222.89 —0.26 
234.66 —+0.21 
244.48 — 0.21 
246.33 +0:45 
272.90 0.68 
291.93 —0.76 
309.66 0.47 
365.01 —0.26 
WO = SP OA, 
| 
And from the pressures thus computed the depth is 
caleulated according to the formula: 
1—— (1— = p 
| ee NE \ Sis s VOR 
a, S(1 —8c0852 9) (2 4. = h) 
med folgende Resultat: with the following result: — 
No. h ber. h obs. Q = IB 
(comp.) 
I 91.2 Fv. (Fms.) 93 Fy. (Fms.) —+1.8 Fv.(Fms.) 
2 419.8 416 — 3.8 
3 12701 1216 = Toit 
4 1278.9 1280 —+1.1 
5 1334-1 1333 ott 
6 1340.6 1343 + 2.4 
7 1483.3 1487 iO 
8 1594.1 1590 —4.1 
9 1683.5 1686 +2.5 
IO 1986.5 1985 —1.5 
Det er hidtil forudsat, at Lodskuddene angive den 
rigtige Dybde uden at vere beheftede med nogen constant 
eller systematisk Fejl. Den i Forhold til dens egen Værdi 
store sandsynlige Fejl, hvormed & udkommer af Lignin- 
gerne, opfordrer imidlertid til at undersøge, om muligens 
den hele fundne Værdi af & skulde være et Resultat af 
systematiske Fejl ved Lodskuddene og saaledes savne Rea- 
litet. Prøven kan gjøres ved at sætte & = 0, beregne den 
under denne Forudsætning udkommende sandsynligste Værdi 
for % og dermed de tilsvarende Værdier af % til Sammen- 
ligning med de observerede Lodskud. 
Man faar da 7 = Summen af Coefficienterne for x 
i Betingelsesligningerne divideret med Summen af Leddene 
paa højre Side af Lighedstegnet eller 
po SLB 40 
~~ 20.9195 
og dermed 
= 2.8128 X 10 
I 
| 
MF = fl 2.3 = Panne (Hms.); od = + 1.94 Favne (Fms.). 
It has hitherto been assumed that the soundings give 
the true depth, not being affected by any constant or 
Meanwhile, the relatively large probable 
error of ¢, as compared with its actual deduced 
from the equations, calls for investigating whether, perhaps, 
the whole value of & that I have found may not be a 
result of systematic errors attaching to the soundings, and 
The proof can be made by putting 
systematic error. 
value 
thus want reality. 
é = 0, working out the most probable value of % under 
that supposition, and with this also the corresponding values 
of h, to compare with the observed soundings. 
We then get x’ = the sum of the coefficients for x 
in the conditional equations, divided by the sum of the 
terms to the right of the sign of equality, or 
log % = 4.3999884 —10, 
and with this value of zx 
