406 
MESSES. T. E. THOEPE AND A. W. EUCKEE ON 
rately calibrated. The dilatometers were constructed of flint glass containing 33 per cent, 
of oxide of lead, and the coefficients of expansion were determined for the limits of 
temperature between which our observations were made by observing the increase in 
the apparent volume of mercury. The calibration was made by a method substantially 
the same as that employed by Kopp — that is, by determining the weights of mercury 
contained in the various parts of the graduated stems. In several particulars, however, 
the method actually adopted possessed some points of advantage over that upon which 
it was based, both in improved modes of keeping the temperature of the bath constant 
and in withdrawing the mercury, and also in the plan adopted of weighing the mercury 
withdrawn so as to avoid removing the instrument from the bath. The thermometer 
used was one of a set made by Casella ; 1° on each instrument had a length of about 
9 millims. They were previously compared with Kew Standards belonging to Sir W. 
Thomson. Frequent determinations were made of the zero-point of the thermometer 
during the progress of the observations ; indeed the alteration in the fixed point had 
been noted sufficiently often over a period of nearly six years to make it possible to 
represent it by means of an equation like that obtained under similar conditions by 
Prof. W. H. Miller (“Determination of Standard Pound,” Phil. Trans, for 1855). 
The bath in which the instruments were immersed contained from thirty to forty litres 
of water, which was agitated by a stirrer worked by a small hydraulic engine. Direct 
experiments have shown that the temperature of the whole mass was thus made sensibly 
uniform. 
The bath was heated by steam driven in from a small metallic boiler. When the 
temperature at which an observation was to be made was nearly attained, the delivery 
of the steam was regulated so as to maintain the temperature as nearly as possible 
constant ; and at least six observations were made with a very slowly rising and an equal 
number with a very slowly falling thermometer. The readings were obtained by the 
aid of a telescope furnished with a micrometer-eyepiece, and the result given in each 
case is the mean of at least six observations. 
As we desired to make our observations on the sea-water in its natural condition, it 
was considered unnecessary to attempt to extend them over a wide range of temperature. 
It was found that the water might be heated to about 50° C. before the air in solution 
made its appearance, provided that the dilatometer was kept perfectly still ; but from the 
unavoidable vibration produced by the action of the stirrers in the bath, minute bubbles 
were formed at four or five degrees below this point. Accordingly the observations 
have not been extended beyond 40°. For the purpose of reducing specific-gravity 
observations this limit is amply sufficient, since sea-water never attains so high a tempe- 
rature even in shallow places within the tropics. 
The specific gravity of the sea-water at 0° C., compared with distilled water at the 
same temperature, was found, by the bottle, to be F02867. 
The results of the first series of observations are contained in Table I. 
