216 
Proceedings of the Royal Society of Edinburgh. [Sess. 
Table I. — Density of a Water by the Method of Reversal. 
Series. 
Set of 
Obsers. 
Micromete: 
Tube S. 
Distilled Water. 
Height = H. 
r Readings. 
Tube K. 
Sea- Water. 
Height —h. 
Difference. 
Relative Density. 
a=h /h. 
I. 
No. 1. 
1320T6 mm. 
1283-18 mm. 
- 36"98mm. 
1-0288 
M 2. 
17*66 
80-87 .. 
— 36*79 ii 
7 
1! 3. 
13*84 „ 
77-11 
-36-73 „ 
8 
II. 
u 4. 
1315-81 „ 
1278-87 .. 
-36*94 „ 
9 
• i 5. 
14-79 
77-92 „ 
-36-87 „ 
9 
U 6. 
13-83 .. 
77-06 
-36-77 
8 
Mean = 
1-0288 
After inter- 
Sea- Water. 
Distilled Water. 
changing 
waters. 
Heights A 
Height = H. 
III. 
n 7. 
1283-79 mm. 
1318-90 mm. 
— 35 Timm. 
1-0274 
8. 
80-52 i, 
15-68 
-35-16 i, 
4 
n 9. 
78-82 
13-94 i, 
-35-12 „ 
4 
IV. 
.» 10. 
1284-01 „ 
1319-20 i. 
-35-19 
4 
i, 11. 
81-46 „ 
16-64 
-35-18 „ 
5 
12. 
76*51 „ 
11-55 „ 
-35-04 
5 
Mean = 
1-0274 
Mean of means = 
1-0281 
By pyknometer A 18 = 
1-0279 
Diff. = 
•0002 
The results tabulated above were among the first obtained with this 
particular form of the densimeter. With a fuller knowledge of the in- 
strument, it was found possible to increase the standard of accuracy 
considerably, the means of the density values as determined by the densi- 
meter and pyknometer differing by not more than ± '00005, or 1 in 20,000 
(vide infra). 
Now, although the density of a sea- water may thus be measured with a 
high degree of accuracy, it is quite obvious that the method is both slow 
and irksome — not only because some 60 or 70 revolutions must be imparted 
to the micrometer screw during each set of observations, but also because 
of the time required for reversing the waters in the cisterns and tubes. 
Hence, for routine work, the ordinary process of using the instrument should 
be discarded for the method of differences, as set forth in the next section. 
