57 



being stirred about from time to time during a couple of days, the 

 smaller fragments nearly all dissolved, while the larger lumps were 

 only superficially acted upon. The solution thus formed, having been 

 found strong enough for the purposes for which the silicate was pre- 

 pared, was poured otf, and fresh water poured upon the lumps, which 

 were frequently stirred during two or three days, by which a second 

 solution, but very much weaker than the first, was obtained. At this 

 period my experiments were interrupted, and the jar containing the 

 solution and the undissolved lumps was put away in a cupboard, where 

 it remained undisturbed for nearly a year. I then found that some of 

 the lumps still remained, to a great extent, undissolved ; but a great 

 number had softened into a pasty mass, in which were disseminated 

 here and there the un softened lumps. The whole of this pasty gelati- 

 nous mass was not immediately derived from the softening of the lumps, 

 as a part appeared to have been precipitated from the supernatant liquor, 

 so that the uneven sui'face formed by the original pasty mass was filled 

 up and partially covered over by a thin layer of gelatinous silica, like 

 that formed by precipitating a solution of basic silicate by soluble car- 

 bonates, or by a solution of sal-ammoniac. Upon the top of this pasty 

 mass, beautiful white warty concretions had formed, the whole being 

 covered by about six inches of water. The borders of the warts were 

 serrated, the serrations being produced by the projecting ends of fine 

 prismatic needles. In every instance the warts formed over a lump of 

 undissolved silicate, being largest where the lump came closest to the 

 surface of the pasty mass. 



The jar, tightly covered with writing-paper, was again laid aside, 

 but in a place where it could be frequently examined. The warts gra- 

 dually increased in number, each new one appearing to commence over 

 a lump, or where the pasty mass was thickest and most granular, until 

 at length they extended into a continuous snow-white crust. The po- 

 sitions of the warts in this crust were marked by raised prominences. 

 The crust thus formed continued to increase in thickness, the fresh 

 depositions appearing to begin, as at first, over the lumps, so that the 

 raised prominences became more and more marked, until a distinct 

 reniform structure was developed. "While this growth was taking place, 

 the water had gradually evaporated, until not more than an inch covered 

 the crust, and the pasty mass had become quite gelatinous. 



The supernatant liquor, which was a solution of carbonate of potash, 

 containing only a mere trace of silica, was poured off, and the crust re- 

 moved as carefully as possible. The latter was very fragile, the slightest 

 pressure reducing it to a pulpy mass. The gelatinous mass upon which 

 the crust rested had a yellowish colour ; left in the jar, it gradually 

 dried and cracked. Part of it, when dried, consisted of an opaque 

 whitish-grey substance, mottled with pure white, which was very friable 

 when dried for some minutes in a water-bath. Another part, however, 

 was semi-translucent, hard, and very like some varieties of opal, and 

 contained water even after having been exposed to dry air for several 



B. I. A. PEOC. VOL. VIII. I 



