PHOSPHORUS, SULPHUR, AND SELENIUM. 35 



applied at the upper end of this arm. Mercury was thus forced up to take 

 the place, in large measure, of the oil or water in the jacket, and in this 

 way the jacket was made to contain a minimum of supplementary liquid. 

 The lubricated stopper was now put in place, and after wiping, the jacket 

 was reweighed. Thus were obtained sufficient data for obtaining the 

 weight of water or oil and the powder used. The compression was then 

 carried on in the usual way. 



This method must sometimes be applied also in cases where the mate- 

 rial contains fine cracks difficult to free from air, as in many rocks and 

 other substances with crystalline structure. It was used in the case of 

 all the substances discussed in this paper, except carbon. 



This exceptional substance, in the form of graphite, needed especially 

 elaborate treatment on account of its porous nature. In order to remove 

 moisture, adsorbed air, and any other volatile substances present, it was 

 inclosed in a hard glass tube and was heated to redness for several hours 

 in a Sprengel vacuum. After weighing, the carbon was subjected again 

 to the same treatment for several hours more. Boiling distilled water 

 was next allowed to run in upon the carbon and remained in contact with 

 it for about thirty-six hours. It was hoped by this means to remove the 

 air from the pores of the solid and fill them with water. As much of this 

 water as possible was then replaced by mercury in the jacket, and the com- 

 pression conducted in the usual way. Even with all these precautions, it 

 was found impossible to obtain wholly consistent results with graphite, 

 and the value given in the table must be looked upon as the maximum 

 value. Apparently the substance contains a multitude of pores of sub- 

 microscopic minuteness which only fill with water under high pressure. 

 It may be that these pores are only developed by the compression of the 

 substance, or at least only opened up by this compression. 



Both in the cases of carbon and silicon, several preliminary determina- 

 tions gave values for the compressibility of each element somewhat lower 

 than those given in the table ; but those recorded are the last and most 

 careful experiments, and are chosen for that reason. 



