38 HILLEBRAND, MERWIN AND WRIGHT. [April 25, 



above is a single specimen from Paradox Valley, shown us by Mr. 

 Frank L. Hess, of the U. S. Geological Survey. This consists of a 

 single bundle of interwoven fibers implanted on a layer of crystal- 

 lized gypsum. This specimen, on account of its markedly fibrous 

 structure, bears a closer resemblance to the best of the hewettite 

 from Peru than does the mineral described above. Indeed, optical 

 evidence shows that it is hewettite. 



Analyses of Hewettite and Metahewettite, and Their 



Discussion. 



Long after the first analysis of hewettite was made, it was found 

 that both it and metahewettite are extremely sensitive to atmos- 

 pheric changes in humidity, especially within a certain narrow range. 

 It was therefore essential that they be brought always to the same 

 definite state of saturation with respect to water.^ It was found by 

 experiment that this condition could be satisfactorily attained by 

 exposing the mineral powder at a definite temperature over sulphuric 

 acid of vapor tension near that of pure water until equilibrium was 

 established. The strength of acid over which the mineral was placed 



9 Failure to observe this precaution may lead to serious error in estab- 

 lishing a probable formula for minerals with variable water content. For 

 instance, if, as with hewettite and metahewettite, the mineral is analyzed in 

 air-dry condition when the air humidity is high, a very different result will 

 be obtained than when the air is dry. The variations in moisture content of 

 these minerals when left exposed to the air may vary 8 or g per cent, between 

 September and December in Washington. This is not efflorescence as usually 

 understood, for although the reaction reverses itself with return of humidity 

 the loss of water is not accompanied by breaking down of the crystal structure. 



Calcio-carnotite from Colorado (the tyuyamayunite (?) of Nenadkevich) 

 and probably also the original carnotite (essentially the potassium salt) show 

 similar wide differences in water content at different seasons of the year. 

 This may be and probably is true also of other minerals. If so, an explana- 

 tion is afforded of some of the conflicting statements of different analysts 

 in regard to the water content of certain minerals. 



It is, further, important to make a series of exposures over sulphuric 

 acid of increasing concentrations, up to the maximum, at a fixed temperature, 

 and then to carry out tests at rapidly increasing temperatures, in order to 

 detect, by the losses at each step, the number of hydrates that may exist, and 

 also, if possible, what proportion of the water may be differentiated from 

 water of crystallization or of absorption. The curves resulting from sucb 

 tests with these minerals are shown in Fig. 2. 



