TRAVELS IN BRAZIL. 171 



chlorite-slate. In order to make a still greater confusion 

 of ideas, we have only to put the questions — Where was the 

 matrix from which they were separated? What power could 

 that be which so broke to pieces the matrix and themselves, 

 that not one stone remained connected with another, but 

 each appears wholly insulated ? If they were rent from 

 another place and here brought together again, how comes 

 it that lithomarge, as it were, prepared a bed for them in 

 which they were deposited, as in their original situation ?" 

 In Gilbert's Annalen der Physik, vol. i. p. 4., M.Von Esch- 

 wege again says, that in the chlorite-slate, topazes are en- 

 veloped in lithomarge; nay, in his latest geognostic de- 

 scription of Brazil, where he himself says, that he has 

 formerly written various things on these subjects with which 

 he is not now quite satisfied, he still observes: — " That 

 talc and chlorite-slate appear inseparable ; where the rock 

 is entirely decomposed into fullers'-earth they find, in nests 

 or groups, enveloped in lithomarge, the beautiful yellow 

 topazes, and also the rare euklase ; and often, in large fine 

 six-sided tables of crystaUised ironglance, with crystalhsed 

 talc, rock-crystals with topazes immersed in them, or topaz- 

 crystals with rock-crystals immersed, also cyanite, &c." 



John Mawe, in his Travels in Brazil, has described the 

 occurrence of topazes very differently, and, in our opinion, 

 more correctly. According to his observations, the topazes 

 are found at Capao in narrow veins in a clay-slate passing 

 into mica-slate. He thought that they were no longer in 

 their original repository ; that the crystal had but one acu- 

 mination, and consisted of fragments ; some were, indeed, 

 grown together with quartz, but even the quartz itself ap- 

 peared but in fragments. The topazes were enveloped in 

 friable earthy talc and large crystals of ironglance, &c. 



If our observations on the occurrence of the topazes 

 are compared with the preceding remarks, it appears that 

 they are found neither in chlorite-slate nor in a fuller's- 



