I9I4.] HYDROUS CALCIUM VANADATES. 45 



were not sufficiently close together to indicate all the hydrates 

 formed. 



Loss of Water IVIien Heated Above ioo°. — Other differences be- 

 come apparent when the minerals are heated up to the point of los- 

 ing the last molecule of water. The experiments showing this were 

 made at the geophysical laboratory, the heating being done in an oil 

 or a nitrate bath, both in an open tube and in tubes which were 

 evacuated from time to time, and the loss determined by weighing 

 the tubes. 



Metahewettite, after losing 13.8 per cent, (six molecules) of 

 water over strong sulphuric acid, was first heated gradually from 

 100° up to 350° during one hour with weighings at frequent inter- 

 vals. Water was expelled abundantly at three stages near the tem- 

 peratures 185°, 275°, and 340°. The loss at each stage as determined 

 by these weighings and by repeating the heatings and weighings on 

 another sample at favorable temperatures, was found to correspond 

 to about 2.3 per cent., or one molecule of water at each stage^^ (see 

 Fig. 2, lower portion). The sizes of the circles represent the loss of 

 water at each heating divided by the time of the heating. This is 

 the rate of loss of the water. 



Hewettite, on the other hand, exhibits only one well-marked loss 

 in weight when heated rapidly. Near 250° a loss corresponding to 

 two molecules of water was observed. Near 300° and 350° there 

 are evidences of increased rate of loss if the temperature is increased 

 rapidly in 30 minutes. These losses correspond to one-half molecule 

 each. These data are not shown in the figures. 



15 This method of rapid heating and frequent weighings affords surer 

 indications of the existence of water in different stoichiometrical proportions 

 than that of holding the mineral in dry air for a long time at successively 

 increasing temperatures. By either method a rate of dehydration, not a con- 

 dition of equihbrium, is measured. Fig. 2 shows two distinct breaks in the 

 rate of water loss for metahewettite when it is heated rapidly as described 

 in the text. The curves plotted for S-hour periods of heating in a dry-air 

 current (not shown in the figure) are without distinct breaks. 



The temperatures at which water escapes most rapidly bear little relation 

 to those employed in slow heating. Long heating results in a higher loss at 

 a given temperature, and in complete dehydration at a much lower tempera- 

 ture, than very rapid heating. 



