16 THE CANADIAN NATURALIST. [Yol. X. 



amount of liquid thus obtained was considerable. The tubes 

 were carefully weighed, then opened, and after an esamioation 

 of the liquid, thoroughly dried, and weighed again. The weight 

 of the water in No. 1 was thus found to be 13-4 milligrams, in 

 No. 2, 12 milligrams, corresponding respectively to 13 4 and 12 

 cubic millimeters at 4*^. The volume of the quartz in the first in- 

 stance being 8-25, and in the second, 7-41 cubic centimeters, we 

 have for the amounts contained in one cubic centimeter of the 

 mineral, 1-63 and 1-62 cubic millimeters respectively, no correc- 

 tion being made for temperature, as the results are only approxi- 

 mate. This would indicate a comparative uniformity in the dis- 

 tribution of the water, while the amount of the gas varies. But 

 such a conclusion is at best doubtful, inasmuch as the darker 

 quartz is not as thoroughly broken up by the heat as the lighter 

 variety, and the refrigeration of the tube No. 1. was not made 

 complete at first, so that some water doubtless escaped with the 

 gas uncondensed. 



A small portion of the water removed with a minute pipette 

 was dropped upon red litmus paper, where it produced a strong 

 but fugitive alkaline reaction, implying the presence of free am- 

 monia. This was confirmed by adding Nessler's test solution to 

 the remainder of the liquid in the end of the tube, in which it 

 caused the characteristic yellow coloration, and, in one instance 

 a slight precipitate. Before the tubes were opened it had been 

 noticed that the water, though to all appearance prefectly trans- 

 parent and colorless, left a white deposit upon the glass where a 

 drop of it had evaporated. When this was heated by the appli- 

 cation of a small gas flame, it did not fuse, but appeared to shrink 

 or to diminish in amount very slightly, while the glass around it 

 and over it lost its transparency as if corroded. A similar but 

 very slight action upon the glass where the moist gas had come 

 in contact with it had previsouly been observed. This sugges- 

 ted the presence of fluorine. The glass of the tube in which the 

 efi"ect was most marked contained some lead, but the other showed 

 it also to some extent. A special experiment with a tube free from 

 lead, which had been most carefully cleaned, gave the same re- 

 sult, though in somewhat less marked degree. Its appearance 

 would be accounted for by the supposition that the water of the 

 cavities contained some hydro-fluo-silicic acid in solution, result- 

 ino- from the decomposition of silicon fluoride, or, as ammonia 

 was also present, from an ammonium compound of the acid. 



