trykket for det absolute Lufttryk, eller N (1 — 0.00259 cos 2), hvor 
g er Observationsstedets Bredde. Med andre Ord, den normale Ba- 
rometerhgjde (den til 0° og Normalbarometret reducerede Kviksølv- 
højde) skulde, for at bringes i Overensstemmelse med Aneroidbaro- 
metrets Angivelser, corrigeres med den af Observationsstedets Bredde 
afhængige Tyngdecorrection, — N >< 0.00259 cos 2g, en Størrelse, 
der paa vore Bredder altid er positiv og voxer med Bredden. Und- 
ladelsen af at tage det nævnte Hensyn har imidlertid ikke medført 
nogen Fejl af praktisk Betydning i Reductionen af Expeditionens 
Barometerhøjder, da de udførte Sammenligninger mellem begge In- 
strumenter ere blevne behandlede i Grupper, indenfor hvilke Obser- 
vationsstedernes Bredde ikke varierer inden vide Grænser. I 1877 
var den midlere Bredde af de Steder, hvor Sammenligningerne mel- 
lem Kviksølv- og Aneroidbarometret foretoges, 68.°1. Den gjennem- 
snitlige Forskjel mellem Tyngdecorrectionen ved en enkelt Sammen- 
ligning og ved denne Bredde er, uden Hensyn til Fortegn, + 0.7066 
og den største forekommende Forskjel er —0.””21 (64% Bredde) og 
+- 0.mm13 (paa 71° Bredde, ved Jan Mayen). I 1878 ere de tilsva- 
rende Tal for Aneroidbarometer A: 75.9, + 0.7056, — 0.12 (70.96) 
og + 0.mm14 (77.99), og for Aneroidbarometer B: 78.94, = 0.7027, 
— 0.711 (74.%) og + 0.04 (80.90). Det vil sees, at disse Værdier 
for Middelafvigelsen gruppere sig i samme Orden som de ovenfor 
fundne Værdier af den sandsynlige Fejl af en enkelt Sammenligning 
mellem Aneroid- og Kviksølvbarometer (0.17, 0.14 og 0.11). Det er 
derfor sandsynligt, at de sidste vilde kunne formindskes noget ved 
Indførelsen af Tyngdecorrectionen. Ved Rejser, hvor Barometrene 
sammenlignes under forskjellige Bredder, bør, naar der er Spørgs- 
maal om en større Nøjagtighed, altid Tyngdecorrectionen tages med 
i Beregningen, en Fremgangsmaade, der overhovedet anbefaler sig 
ved sin rationelle Begrundelse, og som det er at haabe vil vinde 
gjennemgaaende Indgang i den meteorologiske Praxis. 
3. Luftens Temperatur og Fugtighed. 
Ombord i et mindre Dampskib er det ingen let Sag 
at finde en passende Plads for et Thermometer, der skal 
angive den rigtige Lufttemperatur. Den første Dag, vi 
vare under Damp, den Iste Juni 1876, paa Vejen fra 
Bergen til Sognefjorden, en klar Solskinsdag, gik jeg med 
et Slyngethermometer rundt Skibets Dæk for at prøve alle 
mulige Steder som Plads for Thermometrene. Resultatet 
blev den følgende Ordning, der viste sig meget tilfredsstil- 
lende. 
Begge Thermometre, det tørre og det vaade, opstil- 
ledes i et Hus af bronzeret Kobberblik, paa lignende Maade 
som ved det norske meteorologiske Instituts Stationer.  Kob- 
berhuset (Fig. 10) var 57 Cm. højt, 22 Cm. bredt og 15 
Cm. dybt. Det havde pyramideformet Tag med Skorsten, 
Jalousier over hele Bagvæggen, over hele Døren (der vises 
lukket op i Figuren) og nedentil, i Højde med Thermo- 
metrenes Kugler, paa begge Siderne og paa Forsiden samt i 
Bunden. Thermometrene stode meget frit inde i dette Hus, 
idet de oventil gik gjennem Huller i en Tverbjelke af Tre, 
og nedentil hvilede i Bøjler af tyk Messingtraad, der vare 
indskruede 1 den midt efter Huset staaende verticale Træ- 
bjelke. I den nederste Del af Huset hang ved den Væg, 
der var nærmest det vaade Thermometer, en Kop med Vand, 
hvorfra en Væge sugede Vand til det vaade Thermometer. 
Den norske Nordhavsexpedition. H. Mohn: Meteorologi. 
barometer, should be substituted the expression for absolute pressure, 
or N (1 —0.00259 cos 2g), in which ¢ signifies the latitude of the 
place of observation. In other words, the normal height of the baro- 
meter (the height of the mercury, reduced to 0° and to the standard 
barometer) should, to make it correspond with the indications of the 
Aneroid, be corrected with the gravity correction (— N < 0.00259 cos 2 gp), 
which is dependent on the latitude of the place of observation, and 
which, in our latitudes, is always positive and increases with the 
latitude. Meanwhile, this omission has not given rise to any error 
of practical importance in the reduction of the readings of the baro- 
meter noted on the Expedition, since the comparisons between both 
instruments have been treated in groups, within which the latitude 
of the places of observation does not vary to any considerable extent. 
In 1877, the mean latitude of the localities in which I made the 
comparisons between the mercury and the Aneroid barometers, was 
68.°1. The average difference between the gravity corrections for a 
single comparison and for the said latitude, is, without regard to sign, 
-— 0.""066, and the greatest differences that occurred are — (0.™™21 
(64° N. lat.) and + 0.713 (719N. lat., off the Island of Jan Mayen). 
In 1878, the corresponding figures for the Aneroid barometer A are 
as follows: 73.96, + 0.056, — 0.12 (70.96), and + 0.14 (77.99), 
and for the Aneroid barometer B, 78.°4, + 0.027, — 0.™™11 (74.°6) 
and + 0.™m04 (80.°0). It will be’ seen that these values for the mean 
deviation are ranged in the same order as the values, given above, 
of the probable error of a single comparison between the Aneroid 
and the mercury barometer (0.17, 0.14, and 0.11). Hence, it is prob- 
able that the latter might be somewhat diminished by introducing the 
correction for gravity. On voyages during which the barometers are 
compared in different latitudes, the correction for gravity should, when 
greater accuracy is desirable, be taken into account, a mode of pro- 
cedure manifestly the most rational, and which, it is to be hoped, 
will in future be generally adopted in meteorological practice. 
3. Temperature and Humidity of the Air. 
On board a comparatively small steamer, it is no easy 
matter to find a good place for mounting a thermometer 
that has to indicate the true temperature of the air. The 
first day we were under steam, June Ist 1876, on our 
passage from Bergen to the Sognefjord, in bright sunshiny 
weather, I made the tour of the ship's deck with a sling 
thermometer, to try every available place in which the ther- 
mometers might be mounted. The result was the following 
arrangement, which proved in every respect satistactory. 
Both thermometers, the dry and the wet, were set 
up in a bronzed copper case, in the manner adopted at 
the Stations of the Norwegian Meteorological Institute. 
The copper case (Fig. 10) was 57” high, 22” broad, and 
15” deep. It had a pyramidical-shaped roof, provided 
with a funnel, louvres across the whole of the hind wall, 
across the door (open in the figure), and below, on a level 
with the bulbs of the thermometers, on both sides, and 
in front, as also at the bottom. Within this case the thermo- 
meters had an exceedingly free position, passing as they 
did, above, through holes in a cross-piece of wood, and 
being supported below in rings of thick brass wire, screwed 
into the vertical cross-prece stretching through the middle 
of the case. In the lower part of the case, suspended near 
the wall nearest to the wet thermometer, hung a cup filled 
with water, from which a wick sucked up the fluid to the 
wet thermometer. 
