352 THE CANADIAN NATURALIST. [Yol. viii. 



by drying to 100° C. suffer a loss corresponding to the quantity 

 of the dye-stuff. Pale sea-green fluorspar from Friedrich Chris- 

 tian (Sp. gr. 3.169) showed a loss of 0.202 per cent., while pale 

 violet from the same place with sp- gr. 3.184 lost only 0.2005 

 per cent. 



While fluorspar is almost never wanting in the metallic veins 

 that traverse the gneiss of the Black Forest, it is almost entirely 

 unknown in those that similarly carry galena, copper, pyrites 

 and blende in the Rhenish slate plateau. As the sandstones and 

 clayslates of that region are doubtless nothing but the finely tri- 

 turated debris of older rocks, viz., of gneiss and granite, we must 

 assume that the mica of the latter was before its re-deposition 

 very much decomposed and already deprived of its fluorine ; as 

 also the feldspar of its baryta {vide supra). It seems to me that 

 in this way a chief difference in the mode of formation of metallic 

 veins in primitive rocks and in sedimentary fragmental rocks 

 would be naturally explained. The occurrence of both minerals 

 in the English Mountain Limestone is quite different ; but I do 

 not know the conditions sufficiently to venture to express an 

 opinion. 



I come now to the most important part of my task, viz., the 

 proof of the presence of the heavy and precious metals in silicates. 

 It was known in this connection that the olivine rock and the 

 serpentine formed from it always contained copper, nickel and 

 cobalt : e. g., according to Stromeyer the olivine from the basalt 

 of Giesen 0.37, that from Kosakow in Bohemia 0.33, and precious 

 chrysolite from Egypt 0.32. I have found that this nickelous 

 oxyd is always accompanied by cobalt, but in far smaller quan- 

 tity. Herr Geh. Rath Woehler had the kindness to have both 

 metals determined in the olivine from Nauroth near Wiesbaden, 

 furnished by me in 1869. In 100 parts there were 0.307 

 nickelous oxyd and 0.006 cobaltous oxyd. Still more recently 

 has it been shown that nickelous oxyd occurs very constantly also 

 in the older eruptive lime-olivine rocks — the palaeopikrites : in 

 the palaeopikrite of Dillenburg, 0.162 — 0.666 per cent., and is 

 accompanied by copper, cobalt and bismuth. The palaeopikrite 

 from Ullitz and other places of the Fichtelgebirg behaves similarly, 

 and the far younger pikrites from Maehren and Austrian Silesia 

 according to my repeated experiments contain the same ele- 

 ments ; but no quantitative determinations have yet been made. 

 In the Nassau palaeopikrites it can be proved that the percentage 



