34 
THE VOYAGE OF H.M.S. CHALLENGER. 
Hence: lime in the “ sesquioxide ” = 0T mgrm. for each analysis. 
i fVtr ■ >t< s fr< on the crude oxalates of lime (from 1 G analyses) were mixed and allowed 
• -• in. l i'«>r about six weeks, when a large crystalline crust was found to have separated 
<>ut. consisting chiefly of oxalate of magnesia. This crust was freed from the mother- 
, l . md washed, hy decantation, ignited, dissolved in a little more than the calculated 
w. _ lit of standardised sulphuric acid, and enough alcohol added to the solution to 
produce an incipient precipitate. After a day's standing, a magma of sulphate of mag- 
had formed, which was dissolved by cautious addition of successive instalments of 
water. Xo sulphate of lime remained umlissolved. 
Tht united filtrates from the pure lime precipitates (14 analyses), when allowed to 
: md for about six weeks, deposited a small precipitate, which was analysed and found 
t" in. lude I I mgrrns. of seSquioxides and 4 -4 mgrms. of lime. This amounts to 0'3 
mgrm. of lime for each analysis. 
The ' precipitates were well mixed in an agate mortar, and in four portions 
(w. ighed after re-ignition) the time quantity of lime was determined by the same method 
a- had ■ rv.d to produce the pure lime precipitate from the crude. Results in mgrms. : — 
Substance taken.* 
Real lime obtained. 
Impurity per grm. of substance. 
(1) 
358 4 
357-5 
2-5 
(2) 
309-3 
308-8 
1-6 
(3.) 
11780 
1176-9 
0-93 
(4.) 
1183-9 
1182-7 
1-17 
Mean of (3) and (4), 
105 
I small value (1*05 mgrms.) may almost be accounted for by the solubility of the 
"\alat«- of lime ; I thought the corresponding correction had better be neglected. 
Train the above we see that, in each analysis, the weight of “ pure lime ” as deter- 
Jiiin. 1 by the balance was liable to the following corrections in mgrms. : — 
(1.) On account of the filter-ash, . . - 07 
lit of tlu CiOid lined by the “ aesquioxide,” . . . +0T 
f 1 11 in \ -1 vo<l in tin purification "f the crude lime, . + 0"3 
(4.) For iuipuritie* in the “ pure lime," ...... nil. 
(5.) Tore-error, . . . +04 
Total, +04 
I >nut< correction was applied to the weight of pure lime obtained in each 
■ " uli being thu made somewhat more accurate without appreciable increase 
of laliour. 
b ' • t bowing tabular statement of the results, the numbers all mean “grms. per 
* Corrected for filter-anh. 
