56 



Foråt undgaa de her omtalte Feil. benyttedes til Be- 

 stemmelse af Saltmængden i Søvahdet følgende Fremgangs- 

 maade. 



I en med tætsluttende Laag forsynet tyk. veiet Por- 

 cellaindigel indveiedes 30 til 40 Gr. Søvand, som afdunste- 

 des paa Vandbad. Etterat Saltene vare nogenlunde vel 

 tørrede, ophededes Digelen med Laaget paa ca.|5 Minutter 

 over en Bunsens Lampe, afkjøledes og veiedes panny. Der- 

 etter bestemtes paa den forud beskrevne Maade den ved 

 Decomposition af Chlormagnesium dannede frie Magnesiaj 

 hvorved de til Beregning af den samlede Saltgehalt for- 

 nødne Data erholdtes. 



I en tidligere Afhandling 1 har jeg paavist, at Car- 

 bouaterne i Søvandet ved Kogning omsætter sig til kulsur 

 Magnesia, som ved tnddampning eller under enhver Om- 

 stændighed ved Glødning efterlader Magnesia, og man 

 skulde altsaa strengt tåget for den saaledes dannede Del 

 af den frie Magnesia beregne en anden Correction end for 

 den ved Decomposition af Chlormagnesium dannede Hoved- 

 mengde. Den Feil. man begaar, ved at undlade dette er 

 •imidlertid baade meget nær constant og desuden saa liden. 

 at den uden videre kan negligeres, idet den kun bidrager 

 til at formindske den samlede Saltgehalt med omkring 

 0.0015 °/ . Det er saaledes fuldstændig tilstrækkeligt til 

 den ved Veiningen fundne Mængde tørret Salt at addere 

 L.375 Gangfi den ved Titreringen bestemte Mængde fri 

 Magnesia, for af det saaledes fremkomne Tal at beregne 

 Saltuehalten i Procenter. 



Mod denne Methode kan der dog reises Indvendinger, 

 idet det kunde befrygtes. at mindre Qvantiteter Ohlorna- 

 trium. Chlormagnesium eller Chlorkalium under Glodningen 

 kunde forflygtiges. eller at en Del af den svovlsure Mag- 

 nesia ved den boie Temperatur kunde decomponeres og 

 gi ve Anledning til Tab af Svovlsyre. Man kan imidlertid 

 let forvisse sig om. at dette ved Anvendelse af en tyk Porcel- 

 laindigel med tætsluttende Laag ikke bevirker nogen Feil 

 af mærkbar Indflydelse. Saaledes fandt jeg. at L.2 (ir. af 

 en passende Blanding af Chlorkalium og Chlornatrium ved 

 P/i Times stærkest mulig Glødning over en Bunsens Lampe 

 i den samme Digel, som jeg benyttede -til mine Saltbestem- 

 melser, kun tabte '2 Mgr. i Vægt, det vil sige, islandingen 

 tabte ved Glødningen ikke fuldt 0.14 Mgr. pr. 5 Minutter. 

 Ligeledes paa vistes ved Bestemmelse af Svovlsyre og Mag- 

 nesia saavel ' det benyttede Søvand som i det glødede. 

 Etesiduum, at man selv ved en meget Lengere fortsat Glød- 

 ning end den. der udfordres foråt skaffe fuldstændig vand- 

 frit Salt. ikke risikerer nogen skadelig Feil foranlediget 

 ved Forflygtigelse af Chlormagnesium eller Decomposition 

 al svovlsur Miuneswi. 



1 lm Rulsyrei i Søi indet" Side I" fa 



In order to* guard against the above-mentioned errors. 

 the following mode of operation was adopted for determin- 

 ing the amount of salt in sea-water. 



From 3(1'"' to 40^ r of sea-water were introduced into 

 a thick porcelain crucible of known weight, furnished with 

 å tiglit-nttting cover, and evaporated on a water-bath. So 

 soon as the salt was sufficiently dry. tbe crucible. with the 

 cover on. was heated for about 5 minutes over one of 

 Bunsens gas-burners, then cooled and weighed with its 

 contents. The free magnesia liberated by the decomposi- 

 tion of the chloride of magnesium was now determined in 

 the manner previously described, and the last factor neces- 

 sarv for computing the total amount of salt åccordingly 

 found. 



In a former paper 1 I drew attention to the fact. tbat 

 the carbonates present in sea-water are transformed during 

 the process of boiling into carbonate of magnesia, which 

 after evaporation. or. at least. on the salt being thoroughly 

 heated. leaves a residue of magnesia ; and hence the pro- 

 portion of free magnesia thus formed would, strictly speak- 

 ing, seem to involve tbe need of a correction different from 

 tbat adopted for the principal amount liberated by the 

 decomposition of the chloride of magnesium. But the error 

 which arises from applying the same correction to botli is. 

 on the one hand. very nearly constant, and, on the other. 

 so small as to admit of being safely ignored, seeing tbat 

 it reduces the total amount of salt only about 0.OU15 per 

 cent. It is. therefore. amply sufficient, if to the amount of 

 dried salt found by weighing be added 1.375 times the 

 amount of the free magnesia determined by titration: the 

 hgure thus obtained will serve to compute the percentage 

 of salt. 



This method cer-tainly is so' far opén to objection. 

 tbat small quantities of chloride of sodium. chloride of 

 magnesium, or chloride of potassium may be volatilized 

 during the process of heating. or some portion of the sul- 

 phate of magnesia be decomposedfat the high temperature, 

 and thus occasion a loss of sulphuric acid. The error. 

 however. arising from this source will not exert an v ap- 

 preciable influence on the result. provided the crucible 

 used for the operation be of thick porcelain. and have a 

 tight-titting cover. Thus. for instance. I found tbat -1.2" 

 of a proportionate imixture of chloride of potassium and 

 chloride of sodium. on being heated for the space of an 

 hour and a quarter over one of Bunsens gas-burners in 

 tlie crucible I bad used for my salt-deterniinations. lost 

 only '2"" jr in weight, or 0.14"'" every 5 minutes. More- 

 . over. it was manifest on determining the sulphuric acid 

 and magnesia both in the water itself and in the heated 

 residue. tbat. even in the event of the heating-process being 

 much more protracted than is necessary to obtain salt free 

 from the smallest trace of water, no serious error can 

 result from the volatilization of chloride of magnesium or 

 the decomposition of snlpliate of magnesia. 



"On lin' Carbonic Acid in Sea- Water," p. 40. 



