53 



ombord med Kronometerstanden 15™ 18.*8. Etter Verifi- 

 kationen den 10de i Hammerfest skulde Standen være 15 

 16.*4. Forskjellen, 2.-4 svarer, da Bredden var 72°, til 

 en Aistand af 0.18 Kvartmil. 



Den 24de Juli, da Expeditionen kum tilbage til 

 Hammerfest fra den anden Tur. regnedes ombord med 

 Kronometerstanden 15™ -'il . Den efter Signalerne den 

 28de Juli korrigerede Stand er 15'" 32.*6. Forskjellen, 

 l.ii. svarer, da Bredden er 73". til en Afstand af 0.12 

 Kvartmil. 



Den 24de August, da sidste < fbservation toges paa 

 Tilbagerejsen fra Spidsbergen til Tromso. regnedes Standen 

 af Kronometret ombord til lii'" 2i*0. Etter Verifikationen 

 den 28de i Tromsø skulde den være 16™ 3.*0. Forskjellen. 

 l. s 0, svarer, da Bredden var 74 ' 2 °, til en Afstand at' 0.06 

 Kvartmil. 



Tager man Hensyn til Usikkerheden at Størrelsen at 

 Kimmingdalingen — vi regnede med en konstant Værdi 

 for alle fra Hyttedækket tagne Solbøjder - der beror paa 

 Horizontalrefraktionen, og til Virkningen at [rradiation til 

 at gjøre den iagttagne Solradius niere eller mindre for stor. 

 ettersom Solbilledet ses mere eller mindre lyst, kan vel den 

 sandsynlige Fejl af en paa Soen maalt Solhojde neppe Net- 

 tes under ± 1'. Da i Regelen nedre Solrand maaltes, gjør 

 Irradiationen de maalte Hojder for smaa. Breddérne for 

 store ved Middag, for smaa ved Midnat, Timevinklerne 

 for store paa Vestsiden, tor smaa paa l Istsiden. ostlige 

 Længder for store paa Vestsiden, tor smaa paa Østsiden - , 

 vestlige Længder for smaa paa Vestsiden, for store af Ob- 

 servationer paa Østsiden. 



Saa ofte Anledning gaves, og det paa ganske faa 

 Uhdtagelser nær i Regelen liver Dag. og sandig veil Ankom- 

 sten til en Arbejdssta.tion. korrigeredes Bestikket ved Ob- 

 servationerne. Disse beregnedes altid (undtagen Meridian- 

 højder) med 2 vilkaarlige Bredder eller, naar de var nær 

 Meridianen, med 2 vilkaarlige Længder. Nautical Ahna- 

 nacs Tabeller benyttedes. Stedlinien blev afsat i Kartet 

 som en ret Linie mellem de 2 saaledes bestemte Punkter. 

 Ved Hjælp af den imellem 2 Observationer udsejlede Kurs 

 og Distance bestemtes den paaværende Plads i bver af 

 Stedlinierne. Statiouernes Bredde og Længde bestemtes 

 saaledes af mig og indførtes i Loddejournalen. Bestik og 

 Observationer svarede i Regelen meget godt. Under Skrab- 

 ning og Trawling blev vi imidlertid altid sat ud af Bestik. 

 da det var vanskeligt at bidde Rede paa Kurser og udlobne 

 Distancer under disse Manøvrer. 



Efter det ovenfor anførte om vore Kronometres Gang 



og om den sandsynlige Fejl af en maalt Solhøjde fremgaar 

 det. at vore Værdier tor Statiouernes Bredde og Længde 

 bar en sandsynlig Fejl af mellem 1 og 2 Minuter. For 

 den største Dels Vedkommende overstiger Fejlen ikke syn- 

 derlig 1 Minut. 



Hammerfest, we assumed the error of the chronometer to 



be l."i" is. s. From the verification at Hammerfest on the 

 lOth, it was found to have been 1.")'" L6.*4. The difference, 

 2.*4, corresponds ' on the 72nd parallel of latitude to 0.18 

 of a nautical mile. 



Wben returning to Hammerfest on the second cruise, 

 we assumed the error of the chronometer for an observa- 

 tion tåken July 24th to be 15 m 31*. Corrected' at Hammer- 

 fest mi the 28th by comparison witb the time-signals, it proved 

 to have been 15™ 32.*6. The difference. l. s 6, corresponds, 

 in latitude 73°, to 0.12 of a nautical mile. 



On the 24th of August, wben returning from Spits- 

 bergen to Tromso. we assumed the error of the chrono- 

 meter to be 16'" 2.O. Verified at Tromsø on the 28th, it 

 was found to have been lli'" ■'>/(). The difference. l.*0, 

 corresponds. in latitude 74" '60'. to 0.06 of a nautical mile. 



It' we consider the difficulty of determining the dip 

 •of the horizon, - a constant valne was assumed for all 

 solar altitudes tåken from the roof of the roundhouse, - 

 depending as it does on horizontal refraction, and also 

 take into account the effect of irradiatjon in magnifying to 

 a greater or less extent the diameter of the sun, according 

 as the solar disk appears more or less luminous to the eye 

 of the observer, the probable error of a solar altitude tåken 

 at sea can hardly be put at less than + Y . The altitude 

 of the lower limb having as a rule been observed. the effecl 

 of irradiation would make the measured altitude too low, 

 the latitude too high at noon. too low at midnight. the 

 hour-angles too great west. too small east. the results for 

 longitude east too high on the west side. too low on the 

 east side. fur longitude west too low on the west side. too 

 high on the east. 



YYhenever opportunity offered. which, as a rule. 

 was almost ever)' day, and in particular on arming at a 

 sounding-station, we corrected our dead-reckoning by 

 means of observations. These were. always computed (ex- 

 cepting onlv meridional altitudes) with two optional latitudes, 

 or. when tåken near the meridian, with two optional- longi- 

 tude*. The Tables used were those of the Nautical Alma- 

 nac. The line of equal altitudes was marked out on the 

 chart as a straight line between the two points thus deter- 

 mined. The ship's position on each of the lines of equal 

 altitudes was computed from the course and the distance run 

 between the observations. I thus determined the latitude 

 and longitude of the observing-stations. and entered the 

 results in the sounding-journal. The dead-reckoning and 

 the observations exhibited in general satisfactory agreement. 

 Meanwhile, we were invariably out of our reckoning after 

 working the dredge or trawl: to calculate the days work 

 correctly is next to impossible with such operations in hand. 



From what has been stated above concerning the rate 

 of our chronometers and the probable error of an observed 

 solar altitude. it is evident that the valnes for the latitude 

 and longitude of the observing-stations will have a probable 

 error of from 1 to 2 minutes. For the greater part of 

 them. the error will not much exceed a minute. 



