160 PROCEEDINGS OP THE AMERICAN ACADEMY 



CaClg, aud CaSO^. It has the appearance of a boiler scale from one 

 to two inches thick, semitransparent, with a vitreous lustre and of 

 greenish gray color, although sometimes black on the surface. The 

 scale is usually covered with crystal planes, which have at first sight 

 the appearance of octahedral forms projecting from the surface, but on 

 closer examination the planes were found to be the terminations of 

 prisms extending down into the body of the scale. The faces were 

 somewhat curved, and so irregular as to render accurate measure- 

 ments with the reflecting goniometer impossible. Attempts were 

 made to obtain approximate values by cementing to the faces bits of 

 thin microscope glass, but the results were discordant. The crystals 

 seemed to have a monocliuic habit, and to consist of an oblique 

 rhombic prism terminated by two pairs of planes of the positive and 

 negative hemioctahedrons. For the prismatic angle we obtained 

 57° 46', and for the positive and negative octahedral angles an average 

 of 69° and 75° respectively, but with a variation of more than three 

 degrees between measurements on different crystals. We observed a 

 well-marked cleavage parallel to the assumed plane of symmetry ; also 

 a second cleavage — inclined to the prismatic edges, and mai-ked by 

 striations on the prismatic planes — parallel to the assumed basal sec- 

 tion. There w^ere also indications of both ortho- and clino-domes ; 

 and from these features, as well as from the mode of twinning, it is 

 highly probable that the crystallization is monoclinic, but the evidence 

 is not conclusive. 



A qualitative analysis showed that the material was composed 

 chiefly of sodium and magnesium in combination with carbonic acid 

 and chlorine, with a small amount of calcium, and a trace only of 

 iron. The scale when pulverized was decomposed by water, all of 

 the sodium salts and part of the magnesium passing into solution upon 

 digestion with a sufficiently large volume of boiling water, but from 

 this solution all the magnesium was thrown down on concentration as 

 carbonate. 



In the quantitative analysis, the alkali was determined in two ways ; 

 first, by extraction of the sodium salts with water, and the separation 

 of the magnesium by concentration ; secondly, by the regular Law- 

 rence Smith method ; and concordant results were thus obtained. The 

 magnesium, calcium, and iron were separated in the usual way. Chlo- 

 rine was determined by precipitation with argentic nitrate from a solu- 

 tion of the scale in nitric acid, and the COj was determined by loss on 

 treating with acid in a small apparatus adapted for the purpose, and 

 also by absorption in potash bulbs. The results are given below. 



