G6S DR HUGH ROBERT MILL ON THE 



employed. For convenient use at sea, the hydrometer seems to be the most satisfactory 

 instrument. The optical densimeter, in which the refractive index of the sea water is 

 measured, and the density directly deduced, would appear to be at once more convenient 

 and easily used; but although it has been experimented with by the U.S. Coast Survey, 

 I am not aware that it has been found satisfactory. Standard total immersion areometers, 

 to which the water is rendered uniform in density by dilution with distilled water, have 

 been successfully used in some cases.* 



Determination of Density. — One of the " Challenger " hydrometers has been in use 

 during the whole of the work of the Marine Station. At first the observations were 

 made only on land, but nearly all the Clyde work was done on board the little steam- 

 yacht " Medusa." As in Mr Buchanan's sea-work the glass jar was hung by three cords 

 to a hook in the roof of the cabin, but it was only in perfectly calm weather that it was 

 possible to use the hydrometer, the quick motion of the " Medusa " making the stem 

 bob up and down, so that it was impossible to read the scale. In good weather there 

 was usually time to determine three densities while the "Medusa" went from one 

 observing station to the next, and a surface, bottom, and intermediate sample were taken 

 at each stoppage. In bad weather only two samples were usually taken at each station, 

 as with three the accumulation of bottles was too great to be worked off before the next 

 day's trip. 



The following description of the hydrometer used and discussion as to its accuracy 

 is mainly copied from the first part of my thesis for the degree of Doctor of Science 

 presented to the University of Edinburgh in 1886. 



The hydrometer is a glass instrument with a cylindrical body about 5 centimetres in 

 diameter and 12 centimetres long, terminating above in a stem of the same length, 3 

 millimetres in diameter, and loaded below with a bulb containing mercury. The 

 constants of the instrument were ascertained in June 1884 by Buchanan's method,t and 

 all the observations made were reduced in accordance with these values, which were 

 verified in 1886. The mass of the instrument, corrected to a vacuum, is 150*1478 

 grammes, and can be increased by the addition of small hat-shaped brass, weights placed 

 on the top of the stem, and by rings which rest on these, to 155*8384 grammes, through 

 thirty-six gradations. The volume of the areometer up to the first mark on the millimetre 

 scale enclosed in the stem, is at 0°, 150*207 cubic centimetres, and at 25°, 150'321, the 

 volume of the graduated 100 millimetres of stem is 0*850 cubic centimetre, and assuming 

 it to be cylindrical, each millimetre corresponds to an immersed volume of 0*0085 c.c. 



Three tables were constructed from these data. The first gives the volume of the body 

 of the instrument at each tenth of a degree from 0° to 30°. The second contains the 

 volume of stem immersed, corresponding to each half-millimetre of scale, and this quantity 

 is assumed not to vary appreciably with temperature. The third table gives the mass 

 of the whole instrument when loaded with each of the possible combinations of weights. 

 In redetermining the constants of the hydrometer much difficulty was found in 



* Thoulet's Oceanorjraphie. + Chatt. Rcpts., " Pbys. Chem.," vol. i. pt. 2. 



