496 T. L. WATSON WEATlIEr.lXG OF ALLAN ITE 



I. Analysis of fresh allanite. J. A. Cabell, analyst. 

 II. Analysis of weathered allanite — inner reel layer (crust). J. R. Santos, 

 analyst. 

 III. Calculated percentage of each constituent saved. 

 IV. Calculated percentage of each constituent lost. 

 Y. Calculated percentage loss for the entire mineral. 



VI. Analysis of weathered allanite — outer white layer (crust). J. R. Santos, 

 analyst. 

 VIT. Calculated percentage of each constituent saved. 

 VIII. Calculated percentage of each constituent lost. 

 IX. Calculated percentage loss for the entire mineral. 



The figures disclose some interesting differences, which are difficult of 

 explanation, for the two parts of the weathered crust. Thus, when the 

 analyses of the two weathered portions of the mineral are compared, it will 

 be observed that the inner red layer contains less SiOg, ALOg, Ce^Og, and 

 BeO, and more FcoOg and HgO, than the outer white layer, wdiile botli 

 agree in the absence of DisOg, La203, YgOg, FeO, CaO, and MgO. These 

 differences are brougbt out in a striking manner in the figures given in 

 columns IV and VIII, which represent the percentage loss of each con- 

 stituent calculated on a ferric oxide basis. AYhen compared in the usual 

 way with the analysis of the fresh mineral, we find that the inner layer 

 has suffered a loss from leaching of 76.52 per cent of the original mineral 

 as against a loss of only 61.29 per cent for the outer layer, just the reverse 

 of what would normally be expected. Each constituent of the inner red 

 layer has been removed in larger amount than that of the outer white 

 layer, as indicated in the comparison of figures in column IV (inner red 

 layer) with those in column VIII (outer white layer) of table on page 495. 

 More puzzling still is the fact that the outer layer contains more than 12 

 per cent of ferric oxide and is white in color. It is suggested that, of the 

 several possible explanations considered, probably the most likely one to 

 account for this anomalous condition is the existence of the iron in com- 

 biiialioji with the other c-onstituents in some new colloidal or metacolloidal 

 form, whose properties are yet unknown. Furthei* investigatioji is needed 

 to clear up this point. 



Discussion of Results 



The reddish brown weathered product of allanite from tbe many locali- 

 ties studied bas been shown, both chemically and niicrosco])ically, to be 

 variable in c(;mposition and not susceptible of deiijiite cJiemical expression. 

 As indicated, by the analyses, essentially the same constituents enter into 

 its composition, but in greatly varying ratios. In, the most advanced 



