44 HILLEBRAND, MERWIN AND WRIGHT. [April 25, 



darker red than hewettite and has two directions of strong absorp- 

 tion for hght, while hewettite has only one. Metahewettite crystal- 

 lizes in well-defined tabular forms or broad blades, hewettite in 

 slender blades. 



Loss of Water at Room Temperature. — There is a real difference 

 in the behavior of the two minerals during progressive dehydration 

 over sulphuric acid. Reference to Fig. 2 (curve III.) shows that at 

 35° metahewettite loses 10.65 P^^ cent, of water in one drop from 

 17 mm. to 12.7 mm. vapor tension, nearly as much as hewettite loses 

 in the two stages from 17 mm. to 7.4 mm. (curve II.). The points 

 of observation shown in the several curves of the figure in almost 

 all instances represent 24-hour exposures and are approximately 

 equilibrium points. It was unexpectedly found that equilibrium at 

 these points was practically attained in one day, for even after sev- 

 eral days' exposure the losses in weight were seldom appreciably 

 greater than after one day. This fact seems to support the argument 

 that the curves indicate a real difference between the two minerals.^* 

 The curves shown are as a rule quite closely reproducible with dif- 

 ferent samples of mineral. 



The very important effect of varying temperature at a fixed 

 vapor tension of water in the desiccator is strikingly brought out by 

 comparing curves I. and II. for hewettite. It is seen, for instance, 

 that at 35° and 12.7 mm. vapor pressure the mineral loses nearly 

 8 per cent, of water against a little over half of i per cent, at 25° and 

 the same pressure. Again, curve I. shows an equilibrium point at 

 6.05 per cent, water, curve II. at 7.85 per cent. To the question of 

 why the two points do not appear in the same horizontal line if we 

 are dealing with water of crystallization, two answers suggest them- 

 selves: Either (i) an intermediate hydrate which forms at 25° may 

 not have appeared at 35° or, (2) the concentrations of acid used 



1* The lines connecting points of observation in the figure do not signify 

 that the rate of loss was uniform for each unit-lowering of vapor pressure 

 for the interval between two points. The actual loss for each interval would 

 probably be more correctly expressed by a vertical Hne corresponding to some 

 vapor pressure intermediate between those of the actual points of observation. 

 Just where this line should fall might be determined by making observations 

 with acids of smaller variation of vapor pressures and allowing more time 

 for the attainment of equilibrium. 



