128 II'. G. Mixter — Thermoohemistry of Silicon. 



tight with beeswax. The hydrofluoric acid is weighed in the 

 open bomb and then the top is put on. After the preliminary 

 observations of temperature in the calorimeter the disc, l> is 

 pushed down, thus allowing the silicic acid and cup c to fall 

 into the acid. The mixture is stirred by moving d and h up 

 and down. 



Hydrofluoric acid having a density of 1'079 and containing 

 approximately 22 per cent of II F was used in the work. The 

 specific heat of it was found by cooling about 200 g. in a 

 platinum bottle in snow for hours, in one case over night, and 

 then placing it in the calorimeter used for the work. Two 

 determinations gave - 80 and 0'81. The method is not a good 

 one, but the result is sufficiently accurate for the work as shovvn 

 by - 994 found for the specific heat of water. The heat capac- 

 ity of the hydrofluoric acid solution was so small compared 

 to the total heat capacity, that an extremely accurate value is 

 unnecessary. The calorimetric experiments were made under 

 fairly uniform conditions and hence the results are comparable. 

 One source of a small error was the undetermined specific heat 

 of the solution of hydrofluosilicic acid. 



The line e of Table II shows the heat effect of 1 g. of Si0 2 

 if the water is combined with the silica without heat effect. 

 These e values indicate that little energy is required to separate 

 the water. The heat effect of 1 g. of Si0 2 derived algebraically 

 from the d values of B and C is 594 cal.; from E and F, 588 ; 

 E and G, 587 ; F and G, 588 cal. Now the heat of combina- 

 tion of water in B is calculated thus : 



592 - (594 X 0-99) = 4 cal. That is 4 cal. are required to 

 separate - 01 g. of water from 0*99 g. of silica and for 1 gram 

 molecule, 7200 cal. Likewise we find that 7000 cal. are 

 required to separate 18 g. of water in preparation C. These 

 results are within the experimental errors, but they indicate 

 that energy is required to separate the small quantity of water 

 retained by silica at a red heat. The values found algebraically 

 for the heat of solution of the silica in E, F and G in hydro- 

 fluoric acid are the same as the e values. This indicates that 

 water is combined without heat effect in silicic acid containing 

 7'5 to 21*4 per cent. It should be understood that the silicic 

 acid giving this result was made at 100° and may have con- 

 tained capillary water. Such water would not affect the result 

 essentially. 



Silicon dioxide in its union with water with small or no heat 

 effect resembles the anhydrides of weak metallic acids. The 

 oxides given below are more or less polymerized, hence the 

 difference between the heat of formation of an acid or hydrox- 

 ide and that of a corresponding oxide is not in all cases the 

 heat effect of the combination of water. 



