184 G. II Williams — Celestite from West Virginia. 



represent former concretions. They are now partially empty 

 and partially filled with clay. The celestite crystals adhere to 

 the walls of these cavities or lie unattached in the clay. The 

 latter are the most perfectly developed, being in most cases 

 doubly terminated, and occurring either singly or in groups. 

 With the celestite, minutely crystalline calcite is associated, ex- 

 hibiting the combination: oo P (1010) and - \R (0112). No 

 other mineral was observed within these ovoid cavities except 

 gypsum, minute crystals of which sometimes incrust the celes- 

 tite crystals, especially near their acute ends. 



The celestite crystals vary in length from less than a milli- 

 meter to three inches or more. The largest are two inches in 

 thickness. In color they range from a deep blue to the palest 

 possible tint. Some are even colorless. One specimen shows 

 layers parallel to the macropinacoid which are alternately of a 

 lighter and darker color. Many of the smallest or medium- 

 sized crystals are clear and transparent, but the large majority 

 are more or less opaque from internal impurities. The clay, 

 in which the crystals appear to have grown, is frequently in- 

 cluded in them in large amount. 



The most striking feature presented by this occurrence of 

 celestite is, however, its crystal-habit. This is pyramidal, which 

 is not common for the species, and is always due to the prepon- 

 derance of the acute brachy-pyramid #, P% (144). On a large 

 proportion of the crystals, especially the smaller ones, this form 

 occurs alone, when it becomes almost lenticular from a round- 

 ing of the faces (fig. 6). On other crystals other forms were 

 observed fn combination with the above-named pyramid. These 

 are, in order of their frequency and importance, as follows : 

 a,ooPoo (100); d, \P oo (102); c, OP (001); o, Poo (011); 

 and m, oo P (1 10) (figs. 1 and 2). Of these forms, cl is the only 

 one whose planes are sufficiently even and bright to yield satis- 

 factory measurements, a is always striated parallel to the ver- 

 tical axis, probably by oscillatory combination with m. c, the 

 principal cleavage plane, is uneven and drusy, when it occurs 

 at all. o is usually present only as a rounding of the edge : 

 (144) : ( 1 44), but in a few cases yielded a tolerably sharp reflex. 

 m is of extremely rare occurrence and even then is so poorly 

 developed as to leave its identification doubtful. 



The following angles, obtained with a reflecting goniometer 

 are unsatisfactory, owing to the uneven nature of all the planes 

 except d. They are compared with the calculated angles given 

 by Auerbach in his monograph on celestite,* because none of 

 the data requisite to the calculation of a new axial-ratio could 

 be obtained. 



* Sitzimgsberichte der Wiener Akad., vol. lix, p. 549. 1869. 



