382 Scientific Intelligence — Geology and Mineralogy. 



usual character of the mountain-streams to be, that they rose at the 

 foot of some little elevation, which stood upon a lofty table-land, com- 

 posed of sandstone, then flowed in a sandy bed for a short distance, 

 and afterwards mysteriously sank in the cracks and crevices made 

 in the rocks from atmospheric influences, and did not again reappear 

 until they had reached the foot of the precipice which terminated the 

 table-land whence they sprang ; here they came foaming out in a 

 rapid stream, which had undoubtedly worked strange havoc in the 

 porous sandstone-rocks, among which it held its subterraneous course. 

 What the amount of sand annually carried down from the north- 

 western portion of Australia into the ocean may be, we have no means 

 whatever of ascertaining : that it is sufficient to form beds of sand of 

 very great magnitude, is attested by the existence of numerous and 

 extensive sand-banks all along the coast. One single heavy tropical 

 shower of only a few hours' duration, washed down, over a plot of 

 ground which was planted with barley, a bed of sand nearly five 

 inches deep, which the succeeding showers again swept off*, carrying 

 it farther upon its way towards the sea. — Grey's Journals of his Ex- 

 peditions of Discovery. 



3. On the Chemical Constitution of Wolfram. By M. Marguerite. 

 — Chemists are agreed as to the nature of the elements which are 

 contained in Wolfram, but they appear to be ignorant of the degree 

 of oxidizement in which the tungsten exists. In a late sitting of the 

 Institute, M. Pelouze stated the results of the experiments which M. 

 Marguerite had performed to determine this point. 



Admitting, of which, indeed, no doubt can be entertained, that the 

 analyses of Vauquelin, Berzelius, and Ebelmen, are correct, the fol- 

 lowing formulae may be assigned to Wolfram : — 



1st, 3 Wo' FeO. MnO Wo^ 

 2d, 3 W2o5 FeO'MnO Wo^ 

 3d, 4 (W2 0^) 3 (Fe2 0") Mn' 0^ 



The author decides in favour of the last formula, which represents 

 Wolfram as a compound of blue oxide of tungsten, and of two-thirds 

 oxide, of either iron or manganese. These oxides being isomorphous, 

 if R represent iron or manganese, or a mixture of these two metals, 

 we arrive at the formula 



W2 0^ R2 03, 



which will represent all the varieties of Wolfram. 



The first formula, which represents the tungsten as in the state 

 of tungstic acid, and the iron in that of protoxide, is not admissible ; 



