Cross and Hillebrand—Cryolite from Colorado. 281 
sequent careful examination fully revealed, though the crystals 
analyzed were not entirely free from foreign admixture. It 
seems certain also that the compact bluish portions (I) consist 
almost entirely of pachnolite, although it cannot be positively 
asserted that Some thomsenolite may not be intergrown with it. 
That no possible doubt might exist in the mind of any one as 
to the homogeneity of the material used for analysis III, a 
further analysis was made upon crystals from specimen B above 
described, particular care being taken to identify each as pach- 
nolite by the rhombic section. The results of this analysis ap- 
pear under IV. 
Hillebrana. Brandl.t Calculated 
ode es 
i Il. IIL. IV. H20. — 
Al | 12°02 13-02 | 13°01 | 12°23 12°36 13°60 12°32 
Ca | 19°32 15:27 | 15°17 | 18-06 18°04 18°83 17-98 
ug | O13 153 
Na | 10-43 10-28 10°23 10°25 {| . 11°73 10°34 
Ka 0-13 
H,0} 17.87|7°95} 38-64} 8-79! $10! 811| 8-05 810 
F 51°33) 51°28| 51:30* 55°69 51°26 
99°95 100°00 99°85 100-00 
Further determinations of water on material from specimen 
B gave 7-95; 7°99; 8°14, and 8°15 per cent. Still other deter- 
2°968, was 
equally pure at 22° C. gave 2962. The transparent crystals, 
well as all the other portions analyzed, decrepitated vio- 
B 
m that mineral. Should this prove to be the case, a plausible 
xplanation of the difference in crystallization of the two min- 
erals is offered without recourse to the theory of dimorphism. 
_A satisfactory explanation of Brandl’s results so opposed to 
those presented by all earlier analyses and the ones above 
* By difference. Ann. a, Chem. cexiii, p. 6. 
Am. Jour. 
t 
ie Tap Serres, VoL. XXVI, No, 154.—Oocr., 1883, 
9 ; 
