titreres med Oxalsyre. Tillige bør det erindres, at de som 
Udtrvk for den surt bundne Kulsyre opførte Tal indebolde 
Feilene saavel i Bestemmelsen af den neutralt bundne 
som den samlede Kulsyremængde. 
Som man let vil overbevise sig om. vise de i Tabellen 
opførte Tal især for den neutralt bundne Kulsyres Ved- 
kommende en mærkelig ( )verensstemmelse, naar nemlig 2 
Observationer. begge udførte paa en "V andprove bentet lige 
ved Grønlandsisen, undtnges, ligger i alle de øvrige 133 
Vandprøver den neutralt bundne Kulsyre mellein Grænd- 
serne 51.4 og 55.4 Mgr. per Litre, saaledes at den største 
Difte rents kun beløber sig til 4 Mgr.. hvad der maa siges 
at være meget lidet i Betragtning af. at disse Tal .ere 
fremkomne ved Undersøgelse ai et Hav paa betydeligt over 
200 geografiske Mile i UdStrækniug saavel i syd og nord 
som øst og vest. Hvad derimod den surt bundne Kulsyre 
angaar. da ere de optrædende Differentser betydelig større 
og beløbe sig i Ydertilfældene til omkring 8 Mgr. pr. 
Litre. 
Jeg bar længe bestræbt mig for at opdage nogen 
Lovmæssigbed i disse Variationer. uden at det dog saaledes 
som for Luftens Vedkommende bar lykkes at erbolde klare 
og paalidelige Resultater, i saa Henseende og det ligegyl- 
digt. bvadenten man vælger Dy-bdeforholdene eller den 
geografiske Beliggeidied til IJdgangspunkt for sin Betragtning. 
Da saaledes ingen Del af det undersøgte Felt ud- 
mærker sig fremfor den. anden ved nogen tydelig Forskjel- 
ligbed i Kulsyregebalt, og da de optrædende Differentser 
overalt ere smqa. ville de erboldte Resultater naturligst 
være at benytte til Udiedning af GjennemsnitsværUier, der 
kunne o])fores som Udtrvk for Kulsyregebalten i det under- 
søgte Hav i sin Helbed betragtet. De Gjennemsnitsværdier, 
som saaledes blive at opstille . som Hovedresultater, ere lor 
den neutralt bundne Kulsyres ^ edkommende 
52.78 + 0.083 Mgr. pr. Litre 
med en sandsynlig Afvigelse herfra af en enkelt Observa- 
tion af + 0.(562 Mgr. pr. Litre 
og for den surt bundne Kulsyres Vedkommende 
43.64 + 0.16 Mgr. pr. Litre 
med en sandsynlig Afvigelse berfra af en enkelt Observa- 
tion af + 1.20. Mgr. pr. Litre. 
Da det først var bragt paa det Rene. at de af Dr. 
Jacobsen iagttagno Egenskaber hos Søvandet skrev sig der- 
fra. at den i Søvandet existerende Saltblanding ved Kog- 
hede decomponerede neutrale Carbonater. maatte det ogsaa 
fremstille sig som en meget interessant Opgave at finde 
den nærmere Forklaring dertil. 
Den nærmestliggende Tanke, som i denne Anledning 
først paatvang sig mig, var den at soge G runden i* Chlor- 
magnesiumens bekjendte Egenskaber. At denne under 
Kogningen selv ved Tilstedeværelsen af Overskud afCblor- 
natrium skulde have Tilbøieligbed til lidt efter lidt at 
spalte sig, og at der af den dannede Saltsyre skulde kunne 
u d drives noget Kulsyre, kunde jo ikke synes umuligt. Der 
borne in mind, that the figures representing the carbonic 
acid forming bicarbonates also include the error in both 
titrations. 
A glance at the Table will show that the . figures 
therein set-down, more particularly those representing the 
carbonic acid forming carbonates, exhibit a remarkable uni- 
formity : save in 2 observations, both with a sample of water 
drawn in close proximity to the ice oft the coast of Green- 
land. the carbonic acid forming carbonates determined in 
the remaining 63 samples lies between the limits 5 1 A"' ar and 
55 . 4 »"<jr p er litre and the greatest difference amounts there- 
fore to only 4'"0 r , which must be regarded as very small, 
considering that the said figures refer to the examina- 
tion of water from a tract of ocean which, measured 
both from north to south and from east to west, extends 
for considerably more than 200 geographical miles. As 
regards the carbonic acid forming bicarbonates, the dif- 
ferences in the amounts determined are, however, much 
greater, reaching S"" jr per litre. 
I have long had my attention directed to the possible 
discovery of a law controlling these variations, similar to 
• that which 1 found to regulate those of air, but have not 
yet succeeded in obtaining conclusive results; and it is 
quite immaterial whether depth or geographical position bo 
made the basis of investigation. 
The water in no part of the ocean-tract explored 
being characterised by properties plainly distinguishing it 
from that of any other, and the differences in the results 
obtained having invariably proved small, the latter will 
obviously serve for the computation of average formulae 
representing the amount of carbonic acid present in the 
water of the sea investigated. These average formulae, set 
down accordingly as the final results, were ‘found to be — 
52.78 + O.083 w * ,r per litre 
for the carbonic acid forming carbonates, with a probable 
error in a single observation of + 0.662 per litre; and 
43.64 + 0. 1 'é m * r per litre 
for the carbonic acid forming bicarbonates, with a probable 
error in a single observation of + l.26"‘° r - per litre. 
Having now obtained conclusive proof that the pro- 
perties observed by Dr. Jacobsen in sea-water were the 
result of the property possessed by the saline compounds 
present therein of decomposing at the boiling-point neutral 
carbonates, the next step was to find a satisfactory explana- 
tion of the interesting phenomenon. 
My first thought in this direction was to seek the 
cause in the known properties of chloride of magnesia. 
Assuming this body to have a tendency of gradually de- 
composing during the process of boiling, carbonic acid might 
possibly be driven oft’ by the hydrochloric acid formed. 
Against such an hypothesis, however, various objections may 
be raised. Dr. Jacobsen and others having shown that sea- 
