'9I4-] HYDROUS CALCIUM VANADATES. 39 



preparatory to analysis was chosen after a preliminary trial had 

 shown about the range of highest vapor pressure over which little 

 change in water content took place. Sulphuric acid of sp. g. about 

 1. 105 (15°) was found to give the desired vapor pressure, about 

 nine tenths that of saturation. ^*^ Two different temperatures were 

 employed for establishing equilibrium over this acid in a thermostat, 

 namely, 25° in winter and 35° in summer (temperatures most easily 

 maintained). 



At 25° the water-vapor tension of this acid is about 21.8 mm. or 

 2 mm. less than that of pure water at the same temperature, and 

 38.8 mm. at 35°, or about 3.5 mm. less than that of pure water. 

 Upon material thus brought to a definite water content the analyses 

 were made. One portion was used for studying the course of dehy- 

 dration, first by exposing the mineral at 25° (or 35°) until equi- 

 librium was reached, successively over sulphuric acid of concentra- 

 tions that corresponded to lowering of the vapor tension by tenths 

 approximately and finally over phosphorus pentoxide.^^ With the 



10 The tables of Domke and Bein were used in this connection (Z. anorg. 

 Cheni., 43, 176, 1905). 



11 The value of conchisions deducible from such a series of fractiona- 

 tions on minerals of the kind in question depends on careful observance of 

 certain precautions. In the first place, the mistake must not be made, as in 

 the present case, of using different temperatures for the initial saturation. 

 It was not expected that our tests would extend from winter into summer, 

 as they did at intervals, thus necessitating the use of two temperatures. Of 

 course, the vapor tension of an acid of given strength is markedly greater 

 at 35° than at 25°. The initial water content of the mineral may therefore 

 differ and the results of tests started at the two temperatures not be strictly 

 comparable. For, although the mineral is under a greater water vapor ten- 

 sion of the acid at 35° than at 25° and might therefore perhaps take up more 

 water than at the lower temperature, it is probable that the higher tempera- 

 ture will have its effect on the mineral also but in a quite indeterminate 

 degree (see Fig. 2 and p. 44 of text, for an instructive illustration of one 

 effect of temperature differences, as shown by curves I. and II.). In the 

 second place, at the conclusion of any one test of a series the mineral must 

 not be allowed to cool in the desiccator after removal of the latter from 

 the thermostat, but must be taken out of it at once and inserted quickly into 

 a capped weighing vessel before it has time to cool, for if allowed to cool 

 before removal the final condition will approach that of room temperature 

 and not be that of the experiment. Moisture may be condensed on the min- 

 eral and its containing vessel and the former may perhaps reabsorb water in 



