l 78 EFulett, Mack and Smyth — 



tube. This insert tube contained the solid CO;,, which 

 was added from time to time and kept well packed down 

 so that the ice which formed on this tube was maintained 

 at — 78 C, thus reducing the partial pressure of the water 

 vapor in the apparatus to about -0005 mm. 4 The mois- 

 ture formed as a sheath of ice on the lower end of the 

 insert tube, and after dry gas had been allowed to fill 

 the apparatus, this insert tube was removed with its 

 sheath of ice and rapidly inserted into its weighing 

 tube and weighed, 5 the insert tube was loosened and the 

 system placed in a vacuum desiccator until the water 

 disappeared, and then reweighed. In view of the fact 

 that gases and vapors are not soluble in ice, the method is 

 evidently exact and capable of wide application in 

 moisture determinations. If vapors other than water are 

 condensed, their nature and amount may be determined. 



After the coal had been inserted and the lower end 

 of the insert tube was at the temperature of — 78°, the 

 air was removed from the apparatus without fear of loss 

 of moisture from the coal, but the rapid liberation of 

 moisture and gas from some of the coals, when first 

 evacuated, caused some of the fine powder to "boil" 

 and be carried along and spread throughout the appa- 

 ratus. This difficulty was overcome by inserting a wad 

 of dried glass wool in the top of the coal bucket. In 

 order to prevent a loss of gas from the coals during the 

 evacuation of the apparatus, we inserted the end of the 

 tube containing the bucket in a cold mixture ( — 20° C, 

 or in some cases cooled it to — 78° before evacuat- 

 ing.) Fig. 3 shows the electric heater and vapor bath; 

 it was found most satisfactory to use the condensing 

 vapor of a constant boiling liquid for heating the coal. 

 We were able to find satisfactory materials for the 

 various temperatures needed. 



When the tube was first inserted into the vapor of the 

 bath there was, in some cases, such a rapid evolution of 

 water vapor that it was not condensed on the insert tube 

 as fast as liberated from the coal, and the pressures 

 observed were due not only to the gas liberated but to 



4 Dushman, Phys. Eev., 5, 223. 



6 It was found necessary to carefully clean out the inside of the insert 

 tube, and to wipe in a definite way; if very accurate results were desired 

 we found soft filter paper very good. After the ice had melted and room 

 temperature had been attained we allowed the pressure inside and out to 

 equalize, and made sure that the weight was constant. 



