THE VOYAGE OF H.M.S. CHALLENGER 
:>6 
; > iin Aug the water A to very near its own temperature. The stopper was now inserted, 
..wrflowing liquid quickly wiped off, the bottle suspended within the mass of 
Anal " water, and its weight os soon as it had become constant noted down (as W A ). 
1 thus obtained the data for solving the equation — 
W 0 -W A = V(S 0 -S A ), 
whence 
V 
when V st mde for the capacity of the bottle ingrms. of water at 4° (at the temperature t°). 
Small variations of temperature, of course, were unavoidable; these, however, were easily 
allowed for by calculation. The n suits thus corrected were as follows: — 
X- 
4&19T 
20-345 
1026-29 
19-290 
1024-83 
18-322 
1023-50 
17-322 
1022-125 
lg a bottle full of 
pure water 
Nome. 
The Original Water, 
A, 
B, . 
C, . 
“»•: Anal." The plunger experiment (made at 10°‘3), when corrected to 19 0, 7, gave 
102<V295, which is practically the same number. The “original” water, after having 
i medium a- described, was again analysed, and found to contain 20‘346 grms. 
pel kilo., showing that the slight diffusion into it of less concentrated and 
n water had not sensibly affected its strength. From the four values for x 
and I deduced the following relation : — 
4 S IM = 998*32 +l*3745x, 
f r-c-k by the f.:.-t that the constant term came to exactly the specific gravity of 
/ iter at 19 7. Acting upon this hint, I calculated the whole of my experi- 
i <>!i tli' 1 * • j of the assumption that for a given temperature t the excess of the 
ific gravity of a sea-water over that of pure water (W,) is proportional to i.e., that 
Mi=D, 
X 
and that I ) = a + bt + ct*. 
' the b< it of my number-, I adopted the method of the least squares; but saved 
" trouble, without, I am sure, losing in precision, by uniting certain sets of 
' ■ : iii' ti made with the same water at nearly the same temperature into one 
neai t integer temperature, which my provisional interpolation 
• tie t" do. Thus, for instance, four determinations of scries II. at 
i z 
l = 
were reduced to 8', and gave 
*8,-1000 = 25-447 
•3 
25451 
7°-7 
25-452 
7°-8 
25-457 
