40 HILLEBRAND, MERWIN AND WRIGHT. [Ap"i '5. 



lower vapor pressures a vacuum was employed. The loss observed 

 at the end of the last exposure was practically identical with that 

 occasioned by a temperature of ioo° for a few hours in dry air. 

 Then the fractionation was continued at temperatures above ioo°. 

 The results of some of these tests will be given later (pp. z^4-45). 



The following table of analyses does not show all the determina- 

 tions that were made. In fact, the first complete analysis of hewet- 

 tite is omitted, because it was made before the need for bringing the 

 mineral to a definite saturation as to water content was realized. 

 That analysis confirmed, however, in all other essentials than water, 

 the data contained in analysis I. below. 



Analyses of Hewettite and Metahewettite. 



Hewettite,l2 Peru, in Equilibrium with Water- Matahewettite, Utah, in Equilibrium with Water- 

 vapor Tension of 21 8 mm. at 25°. vapor Tension of 38.8 mm. at 35°. 

 I. 11. 



V2O5 68.19 70.01 



V,04 I.2I 



V.Oa -35 



M003 1.56 .13 



CaO 738 7.25 7.2s 



MgO none .03 



K2O none .09 



Na,0 15 -08 



H2O (total) 21.33 21.30 (mean of 21.24, 



21.31 and 21.34) 



Fe^Os, etc 11 .19 



fO. — } .80 



Insol 17 \ 



100.10 100.23 



Trace Li in I., none in II. A very little CI in I. and II. Trace P2O5 

 and Se in II. No Ba or Sr found in either. V2O4 and V0O3 assumed, their 

 amounts measured by consumption of permanganate when the mineral was 

 dissolved in dilute sulphuric acid (see discussion of molecular ratios, p. 41). 



addition to that condensed on its surface. In the third place, in cases when a 

 vacuum is employed this should be relieved by air bubbling through acid of 

 the same strength and temperature as that in the desiccator, if possible while 

 the latter is still in the thermostat. 



^2 An earlier analysis of more compact material from Peru, devoid of 

 crystalline appearance, gave the following results : 



