THE VOYAGE OF H.M.S. CHALLENGER. 
The Hydrometer Error. 
Mr. Buchanan, like most of his predecessors, determined his specific gravities 
I »v means of very delicate hydrometers; hence it is worth while to consider whether 
Mirli hydrometer- results can confidently he assumed to he the same as would have been 
..htain* d by a plunger and a chemical balance, — the method which I used in determining 
the dependence of \ on S at t°. 
At first sight it would appear that a result obtained with a properly standardised 
livdrometer could be wrong only on account of the capillarity, i.e., the error arising from 
tin fact, that the instrument, when used, in addition to its own weight, carries also the 
w.-ight of a small hillock of liquid drawn up around its stem. As it happened that I 
had occasion, some years ago, to inquire into this matter from a more general stand- 
point. it gave me little trouble to apply my experience to the special case of Buchanan’s 
instrument. I speak in the singular, because 1 examined only one of his set of four 
hydrometers, which, with his journal before me, I had no difficulty in identifying as 
I - ing the one w hich he used. Buchanan’s hydrometer has a cylindrical body of about 
L60 c.c. displacement, and a narrow stem (periphery = 1 0 ■ 1 mm.), provided with a 
millimetre scale. There is besides a little table which can be fixed to the top of the 
• mi. and loaded with weights. The weight of the instrument is so adjusted that, when 
in -• i -water, it must be loaded with the table, and a greater or less weight on the table, 
to bring the surface of the liquid within the range of the millimetre scale. Supposing 
the hydrometer has to be charged so as to weigh W 0 grms. to sink it in pure water down 
t.' a c rtain mark, w hile, cateris paribus, a weight is needed in the case of sea-water, 
then the specific gravity S is 
r the capillarity. To compensate the capillarity error, we must imagine, 
ch of th< riments, the weight /x of the hillock to go, say to the top 
• •f the instrument, and the hillock itself to vanish. The reading will remain the same, 
but the correct specific gravity will be =S 0 , and 
Whence 
q _ W L±fh 
8 ,- 8 =^ 
“ S A*0 
very nearly. 
T< d- 1 rrrtiii' the quantities /x, and /x 0 , 1 took a piece of wide, thin glass tubing, 
■ • • -nd.q mad< a diamond-mark round the middle, and provided it with a 
platinum wir«- susp<nd« r fused <>n at caeh end, so that it could be suspended from a 
i ’■ to plunge into the liquid exactly up t" the diamond mark. A 
