490 T. L. WATSON— WEATHERING OF ALLANITE 



ally free from grit. The two parts of the allanite, fresh mineral and 

 decomposed crust, are sharply defined from each other, with no indication 

 of gradual passage from one into the other. This is characteristic, with 

 one exception, of all weathered portions of allanite that the writer has 

 studied from many different localities. 



Powder is soluble in hot dilute HCl yielding gelatinous silica; is in- 

 fusible before the blow -pipe and non-magnetic; yields water on heating 

 in closed tube, which reacts faintly alkaline; and reacts slightly for car- 

 bonates. 



Microscopically the powdered crust was composed of a granular waxy 

 material of deep red color, which was partly birefracting and partly iso- 

 tropic, each type exhibiting variable physical properties and probably 

 variable in chemical composition. Separation of the two types in suffi- 

 cient quantity and purity for separate chemical analysis proved unsuc- 

 cessful. After digesting a small portion of the powder for a few minutes 

 in dilute HCl, the residue was washed and examined under the micro- 

 scope. It was found to be of the same character as the undigested portion 

 of powder described above, with no indication whatever of the presence 

 of any undecomposed particles of allanite ; but there appeared a sprinkling 

 of minute, colorless, and double refracting, crystalline grains, which were 

 referred to quartz. A portion of the powdered crust yielded Prof. F. P. 

 Dunnington 11.03 per cent total silica, 7.78 per cent of which was soluble 

 in a 5 per cent solution of sodium carbonate, leaving 3.25 per cent in- 

 soluble and reckoned as crystalline silica. 



Chemical analyses and their discussion. — In columns I and II are 

 given the analyses of the fresh and weathered mineral, and in columns 

 III, IV, and Y the loss and gain of the various constituents calculated 

 on a ferric oxide constant basis.'^^ 



■^8 Method employed in calculations. — In order to trace the changes that have taken 

 place in the weathering of siliceous crystalline rocks, recent investigators (G. P. Mer- 

 rill : A Treatise on Rocks, Rock-weathering, and Soils. The Macmillan Co., N. Y., 1906, 

 400 pages; see especially pp. 187-188) have assumed one of the essential constituents 

 to remain constant as a basis for comparing analyses of the fresh and weathered rock. 

 Investigations have shown that of the essential constituents in siliceous crystalline rocks 

 alumina and iron oxide are the most refractory and have suffered least from leaching 

 and removal in solution. For all cases thus far studied either alumina or iron oxide 

 has been assumed as the constant factor for calculating the percentage loss and gain of 

 each constituent as compared with that in the original rock. 



This method is employed by the writer in the present paper for comparing analyses 

 of fresh and weathered allanite in the attempt to trace the changes in the decomposi- 

 tion of the mineral. While this method has not been applied to the weathering of min- 

 erals, so far as the writer is aware, there appears to be no valid reason why it is not 

 as applicable to the weathering of certain minerals as to rocks. In each of the cases 

 described in this paper ferric oxide appears to have suffered the least amount of loss 

 during the weathering of the mineral and has been taken as the constant factor for 

 comparison. 



