the water project, according to requirement, more and more 

 above the sawdust. The ice cylinder shows on its basis a veiy 

 regular hemispherical cavity, which progresses unaltered until 

 its edges touch the mercury by 5, and the last portions of water 

 have been frozen from above downward. After the formation 

 of ice is at an end, the instrument is still exposed some time to 

 a temperature below 0° C. in order to freeze the last traces of 

 water which exist by h between the mercury and the glass walk 

 The ice cylinder thus formed is perfectly free from air bubbles, 

 and equals in clearness and transparency the finest crystal glass. 

 The piece of apparatus designated in the figure with the letter 

 B is now combined, by means of the cork e forcibly pressed in, 

 with the open limb of the apparatus A in such a manner that 

 not a trace of air is included between the cork and mercury, 

 whereby the displaced mercury flows through the capillary tube 

 / into the glass vessel, which is filled with mercury to g. The 

 capillary tube is cemented with the finest sealing-wax into the 

 smooth and perfectly poreless cork. To cement the cork like- 

 wise on the wider tube in which it sits, would be wholly super- 

 fluous, as a displacement is as little to be feared as an elastic 

 eftect afterward, as I have convinced myself by direct expen- 

 ment. The instrument thus arranged is placed in a room hav- 

 ing the greatest possible constancy of temperature, and is siu- 

 rounded on all sides with a thick envelope of snow which ha?. 

 at a temperature above 0° C, become completely coherent with- 

 out at the same time becoming saturated with water. When, 

 after 6 to 12 hours, the entire instrument has attained 0°C.. 

 the mercury vessel is removed from the cork h, weighed with 

 the mercury it contains, and, after any mercury which may still 

 adhere to the capillary tube, has been carefully removed, 

 returned to its place. The apparatus is then removed from its 

 envelope of snow, the ice it contains is melted by radiation 

 from a non-luminous gas flame brought in its neighborhood, 

 and It IS permitted to attain as before in an envelope of snow, 

 the temperature of 0° C. The mercury vessel is now removed 

 and reweighed. The loss in weight, compared with the fif t 

 weighing IS the weight of the mercury, whose volume calcula- 

 ted tor C. expresses the diminution in volume which the ice 

 cylmder at 0° C. has suff-ered in melting to water at the same 

 temperature. ^ 



Let G„ be the weight of the water frozen, , 



^5 the weight of mercury which entered the instrument 

 during the melting of the ice 



b„ the specific gravity of water at 0° C, 



g, the specific gravity of mercury at 0° C, 



be the specific gravity of ice at 0° C , 



