386 Alf Trolle. 



will also precipitate phosphoric acid, iron and aluminium if present, 

 but qualitative tests showed, that these were not present in any con- 

 siderable quantities. The fdtrate from the calcium oxalate was eva- 

 porated with sulphuric and glowed for some time. The residue was 

 weighed and from an examination of the acidity of the solution of the 

 residue in water (indicator: methyl-orange) it was determined, that 

 neutral sulphates of potassium, sodium and magnesium had been 

 formed. Thus, on determining the magnesium as magnesium-pyro- 

 phosphate in one portion of the solution and in another portion the 

 potassium as potassium-platinum-chloride (after removal of sulphuric 

 acid, lead acetate and hydrogen sulphide), the sodium could be found 

 as difference. 



The density determination was carried out by means of a pycno- 

 meter of ca. 10 cc. (Sprengel-Ostwald model) and the determinations 

 reduced to vacuum, the density of the air being taken as 0.0012. In- 

 stead of the density s the dimension p = (s — 1) X 1000 is noted. 



The average composition of sea-water is known from Forchham- 

 mer's work in the "Tndbydelsesskrift" of Copenhagen of 1859: "Om 

 Søvandets Bestanddele og deres Fordeling i Havet". Using his average, 

 we may estimate from the chlorine determinations, how much there 

 must be of other substances. In this way we obtain the quantities 

 noted in the Table as "calculated". The calculated p value is found 

 by means of Knudsen's hydrographical tables. 





Series LV, 



115ä|4m. 



Series LH, 68 m. 



Samples 



found 





calculated 



found calculated 



CI. . . . 



. 15.67 o/oo 







15.23 7oo 



SO . . . 



1.96 - 





2.230/00 



2.03 - 2. 17 «/00 



Ca ... 



. 0.36 - 





0.33 - 



0.34 - 0.32 - 



Mg . . . 



1.10 - 





1.05 - 



1.06 - 1.02 - 



К . . . . 



. 0.25 - 





0.26 - 



0.24 - 0.25 - 



Na ... 



. 8.60 - 





8.75 - 



8.46 - 8.49 - 





21.16 at 



13.0 



= 21.24 - 



20.52 at 13.5° 20.54 ^ 



The agreement between the observed and calculated values may 

 be said with full certainty to show, that the water-samples possess 

 the character of ordinary sea-water. On going through Forchham- 

 mer's results, we find many analyses, which differ more from the aver- 

 age composition of sea-water than these lake samples analysed 

 here. And similarly, the agreement between observed and calculated 

 p may be said to be quite satisfactory, for the chlorine titration, den- 

 sity determination and the factor p : CI calculated from the hydro- 

 graphical tables may give rise to differences in tlie 2nd decimal 

 of p. 



The sulphate determination is the one which differs most. This 

 is probably connected with the fact, that the water-samples had a 



