VOL. LXXX.] PHILOSOPHICAL TRANSACTIONS. 679 



SO much of the fluid was not required, proved a material convenience. The par- 

 ticular disadvantage in the method of weighing in a vessel, is the difficulty of 

 filling it with extreme accuracy; but when the vessel is judiciously and neatly 

 marked, the error of filling will, with due care, be exceedingly minute. By 

 several repetitions of the same experiments, Mr. Gilpin seemed to bring it within 

 the I ., ^ o tb part of the whole weight. The above-mentioned considerations in- 

 duced the gentlemen employed in the experiments to give the preference to 

 weighing the fluid itself; and that was accordingly the method practised both by 

 Dr. DoUfuss and Mr. Gilpin in their operations. 



The vessel chosen, as most convenient for the purpose, was a hollow glass 

 ball, terminating in a neck of a small bore. That which Dr. Dollfuss used, 

 held 5800 grains of distilled water; but, as the balance was so extremely accu- 

 rate, it was thought expedient, on Mr. Gilpin's repetition of the experiments, 

 to use one of only 2965 grains capacity, as admitting the heat of any fluid con- 

 tained in it to be more nicely determined. The ball of this vessel, which may 

 be called the weighing-bottle, measured about 2.8 inches in diameter, and was 

 spherical, except a slight flattening on the part opposite to the neck, which 

 served as a bottom for it to stand on. Its neck was formed of a portion of a 

 barometer tube, .25 of an inch bore, and about 14^ inch long; it was perfectly 

 cylindrical, and on its outside, very near the middle of its length, a fine circle 

 or ring was cut round it with a diamond, as the mark to which it was to be filled 

 with the liquor. This mark was made by fixing the bottle in a lathe, and turn- 

 ing it round with great care, in contact with the diamond. The glass of this 

 bottle was not very thick; it weighed gi6 grains, and with its silver cap 936. 



When the specific gravity of any liquor was to be taken by means of this 

 bottle, the liquor was first brought nearly to the required temperature, and then 

 the bottle was filled with it up to the beginning of the neck only, that there 

 might be room for shaking it. A very fine and sensible thermometer was then 

 passed through the neck of the bottle into the contained liquor, which showed 

 whether it was above or below the intended temperature. In* the former case the 

 bottle was brought into colder air, or even plunged for a moment in cold water; 

 the thermometer in the mean time being frequently put into the contained liquor, 

 till it was found to sink to the right point. In like manner, when the liquor was 

 too cold, the bottle was brought into warmer air, immersed in warm water, or 

 more commonly held between the hands, till on repeated trials with the ther- 

 mometer the just temperature was found. It will be understood, that during the 

 course of this heating or cooling, the bottle was very frequently shaken between 

 each immersion of the thermometer; and the top of the neck was kept covered, 

 either with the finger, or a silver cap made on purpose, as constantly as possible. 

 Hot water was used to raise the temperature only in heats of 80° and upwards, in- 

 ferior heats being obtained by applying the hands to the bottle; when the hot 



