32 



Som man ser. er den lier til Bestemmelse af den 

 neutralbundne Kulsyre anvendte Methode i alt. væsentligt 

 den samme, som allerede for flere Aar tilbage er beskre- 

 ven af Dr. Mohf. 1 Forskjellen stikker kun deri. at jeg 

 har anvendt Classens Kjoler og desuden ombyttet Salpeter- 

 syren med Svovlsyre. Det syntes .mig nemlig ikke tilraa- 

 deligt at anvende Salpetersyre ligeoverfor et saa stærkt 

 cblorholdigt Mineralvand som Søvand, hvori der desuden 

 fandtes en vel mærkbar om end temmelig liden Qvantitet 

 oxyderbare Stoffe. 



Efter denne Fremgangsmaade har jeg senere ved 78 

 Observationer bestemt Kulsyren i 64 forskjellige Vandprø- 

 ver temmelig jevnt fordelte over tlet af den norske Nord- 

 havexpedition undersogte Felt, Resultaterne tindes sam- 

 nicnstillede i Tabel IT. 



Forst skal i Korthed berøres de Feil. hvormed de i 

 Tabellen opførte Tal kan tenkes beheftede. 



Om man vikle antage. at den i Søvaudet indehohltc 

 Kiselsyre ikke forekom opløst som fri Syre, men forefandtes 

 bundet til Baser i Form af Silicater. en Antagelse af me- 

 get tvivlsom Berettigelse, vikle man aabenbart efter den 

 ovenfor beskrevne Methode faa noget feilagtige Resultater. 

 idet den til Kiselsyre bundne Mængde Baser vilde findes 

 som bunden til Kulsyren.. Foråt kunne danne mig en 

 Forestilling om. hvorvidt den Feil. som man under denne 

 Forudsætning skulde kunne begaa. nogensinde vil kunne 

 tænkes at faa særlig Betydning, har jeg bestemt Kiselsyre- 

 mængden i forskjellige af de bjembragte Vandprøver, idet 

 følgende Fremgangsmaade er kommen til Anvendelse. 0.5 

 Litre Søvand inddampedes med lidt Saltsyre i en Platina- 

 skaal først over fri Ild senere paa Vandbad til Tørhed 

 og tørredes ved 110° — 120° ,C. Saltene udtoges derpaa 

 og pulveriseredes bedst muligt i en vel poleret Agatrive- 

 skaal, hvorpaa de atter tørredes ved samme Temperatur, 

 og sluttelig bragtes over i et passende Kar og tilsattes 

 ca. 200 CC. saltsyreholdigt Vand, hvorved al Gips fuld- 

 stændigt opløstes. De paa denne Maade udskilte Kisel- 

 syremængder vare altid meget sinaa og maatte nærmest 

 blive at betegne- som Spor. idet de. hvor jeg forsogte at- 

 veie dem. kun beløb sig til Brøkdele af Mgr. i den an- 

 vendte Portion Vand. Dette stemmer paa det Nøieste 

 med, hvad Thorpe og Morton nylig har fondet i det irske 

 Hav.-' bvorimod de ældre Opgaver giver noget større Tal. 

 ( lg man vil saaledes se. at der ingensomllelst Rimelighed 

 er for, at den i Søvandet indeholdte Kiselsyremængde skulde 

 bidrage til i nuerkelig Grad at gjøre de fundne Resultater 

 upaalidelige. 



Naar man skal danne sig en Mening om Nøiagtighe- 

 den af disse Observationer, er det imidlertid nødvendigt 

 at tåge Hensyn til en anden < (mstændighed, som kunde 

 tænkes at have bavt sin Indflydelse paa Resultaterne. 



As will be seen. the method adopted for the determ- 

 inatiori of the carbonic acid forming carbonates. was es- 

 sentially the same as that described by Dr. Mobr 1 several 

 years previously. The only difference lay in my kaving 

 employed Classens cooler and made choice of sulphuric 

 acid in preference to nitric. It did not seem advisable to 

 use nitric acid when examining a mineral water so rich 

 in chlorine as is sea-water, and which. besides contains a 

 quantity. small indeed but appreciable, of organic matter. 



By this process I subsequently determined the car- 

 bonic acid in 64 different samples of sea-water, drawn at 

 coniparatively regular intervals from the tract of ocean in- 

 vestigated on the Norwegian North- Atlantic Expedition; 

 the number of observations amounted to 78. The results ' 

 are given in Table II. 



I vill tirst say a fev vords respecting the errors 

 that may possibly affect the accuracy of the figures set 

 down in the Table. 



Assuming the silicic acid in sea-water not to occur 

 as a free acid, but combined vith bases in the form of 

 silicates. an hypothesis of very doubtful valne, the results 

 obtained by the method described above would be obviously 

 to some extent inaccurate. inasmuch as the bases com- 

 bined with silicic acid must in that case have behaved as if 

 originally combined with carbonic acid. In order to ascertain 

 what importance could possibly be attached to an error 

 arising on such an assumption. I determined the aniount of 

 silicic acid in divers of the samples of sea-water brought 

 home with me. adopting for that purpose the folloving 

 method. Five-tenths of a litre of sea-water mixed with a 

 little hydrochloric acid were evaporated to dryness in a 

 platinum dish. at first over a common fire and then in a 

 water-bath,. and dried at a temperature of 110° — 120° 0. 

 The salts were then tåken out and transferred to a well 

 polishéd agate dish, in which they were finely pulverised. 

 and again dried at the same temperature ; finally they were 

 placed in a suitable vessel. and mixed with about 200 '■''' 

 of water containing hydrochloric acid. which thoroughly dis- 

 solved all the gypsum. The amounts of silicic acid thus 

 precipitated were invariably very small, indeed but little 

 more than traces; for, on attempting to weigh them. they 

 proved to be but fractions of a niilligramme. This result 

 agrees exactly with the observations of Thorpe and Morton 

 on vater from the Irish Sea.-' ivhereas the figures in ear- 

 lier statements are somewhat bigher. Hence. there is no 

 reason whatever to assume. that the silicic acid present. in 

 sea-water should to any considerable extent influence the 

 results obtained. 



When judging of the accuracy of these observations. 

 regard must. liowever. be had to another circumstance that 

 might possibly in some measure affect the results. The 

 water examinéd did not cousist of freshly drawn samples, 



1 Mohr Titrii-mothnde ?Ac Aufl. — ".--'I. 

 - Ann. Chom. Pliarin. 1 .".s — 1-2-2. 



Mulir's Titrirmethodeii :;te Aufl. 

 Ann. Chem. Pharm. 158, p. 122. 



524. 



