320 PROCEEDINGS OF THE ACADEMY OF [1880. 



active desiccator as chloride of calcium, indicate a remarkable 

 hygroscopic force in the dried substance ; a property not easy to 

 explain. It will be noticed that this force is exercised much more 

 powerfully immediately after desiccation than it is after a lapse of 

 time. Experiment No. (2) shows that two-thirds of the water is 

 absorbed during the first hour. It has been found that the amount 

 of water in the powdered mineral varies with the hygrometric state 

 of the atmosphere at the time of weighing. It is interesting to 

 note that several of the zeolites, a class of hydrous silicates whose 

 exfoliation by heat is very like that of the vermiculites, also have 

 strong hygroscopic powers, losing and regaining part of their 

 water with ease.^ 



Water of Crystallization — The water in Philadelphite probably 

 exists in three theoretical conditions, viz. : — H^-groscopic water, 

 water of crystallization and water of constitution. The first is 

 driven off by drying at 100° C. or by exposure to dry air over 

 HaSO^; the second by gentle ignition, and is accompanied by ex- 

 foliation ; the third by strong and prolonged ignition. The latter, 

 which probably does not much exceed 1 per cent., and whicli 

 the analyses have shown is not needed with the basic radicals to 

 complete the unisilicate formula, will be regarded with the water 

 of crystallization. The most satisfactory determinations of the 

 water of crystallization have been made by subtracting the hygro- 

 scopic water from the total water. 



The following experiments haA'e been made upon the amount 

 and condition of the water. 



(1). The dry mica, which had been out of the quarry for more 

 than a year, was cut into pieces about 5 mm. square, heated in a 

 platinum crucible to a bright red heat for 25 minutes, cooled in a 

 desiccator over HsSOi for half an hour, and then quickly weighed. 

 It lost 7.58 per cent., which will be regarded as the total amount 

 of water. 



(2). The finely powdered mica holds more water. Different 

 experiments gave:— 7.81 (ignited 10 minutes), 7.89, 7.90, 8.11 

 (ignited 25 minutes), 7.50 (powdered just previous to ignition). 

 Strong ignition of the powdered mica probably volatilizes some of 

 the alkalies in addition to the water. 



>• Damour (Ann. d. Mines, IV, x, 208) shows by an experiment similar to 

 those given above, that the water lost by lieulandite exposed over H^SO^ is 

 all regained in 1^ days. 



