168 G. P. MERRILL — WEATHERING OF MICACEOUS GNEISS. 



fresh and decomposed rocks, so far as available material was at hand, the 

 loss on ignition, minus the loss in the fresh rock, being tabulated as 

 water and calculated as one of the soluble constituents. 



The series given is not large enough to permit safe generalizations, but 

 it is sufficiently large and varied to be more than merely suggestive, 

 particularly when we compare these figures with those representing the 

 " insoluble residues " of ordinary soil analyses, and which vary between 

 55 and 80 per cent in those rich in free lime carbonates and organic 

 matter and between 65 and 90 per cent in others. 



Possible Form of Combination of Soluble Constituents. 



The possible form of combination of these soluble constituents is cer- 

 tainly an interesting question. Undoubtedly a large part of the iron 

 exists as free hydrous oxides. The small amounts of lime and magnesia 

 may exist partly as carbonates, and a part of the silica has been dissolved 

 from the free quartz by the alkaline carbonate used for the extraction of 

 the gelatinous silica after digesting in acids. Undoubtedly, too, a portion 

 of all the constituents is derived from the partial decomposition by the 

 acid of a variety of existing silicates — shreds of mica, hornblende, augite, 

 flakes of feldspar, etcetera — shown by the microscope to be still existing. 

 Eliminating from consideration in the analyses of the decomposed Bo- 

 hemian phonolite and Virginia gneiss the Fe 2 3 , and noting only those 

 other constituents found in sufficient quantities to be considered essen- 

 tial, we shall get, by calculating back to a basis of 100, formulae approx- 

 imating Al 2 Si 3 3 -j-5 H 2 0, and Al 2 Si0 5 4-3 H 2 0, respectively, suggestive 

 of compounds allied to montmorillonite and allophane. The possible 

 errors in these calculations are so obvious that it is unnecessary to enu- 

 merate them, but they are of interest, and sufficient to suggest a possible 

 combination of these soluble constituents other than those of zeolites. 



It would appear to the writer that the soluble potash in soils exists, 

 not in zeolitic compounds, but in some of the numerous decomposition (?) 

 products of feldspar, nepheline, scapolite, etcetera, to which the name 

 pinite is commonly applied. Such at least is the case in the potash-rich 

 soil of Maryland examined by R. L. Packard.* 



These remarks are made, not with a view of individual criticism, but 

 to call attention to points which seem in danger of being overlooked. 



The writer, assuming that his mental processes are similar to those of 

 other workers, and that language used and statements made in such form 

 as to convey to him erroneous or misleading impressions are likely to 

 produce the same results on others, believes that the using of certain 

 terms with a meaning more precise is a subject worthy of consideration. 



*Bull. 2i, Maryland Agricultural Experiment Station, 1895. 



