RKsn/rs r.ASKn on TXsor.uFu.r: ai,.<>:, and Ai..n, , [.-...o, loT 



In the Georgia granitic rocks the iron and aluminum oxideshave i)r()ved 

 tlie most resistant of the constituents present toward the weathering 

 agents, and while the ratio of increase for the two is nearly the same, 

 that of the iron is, with several exceptions, sliglitl}^ greater than alumina, 

 and has therefore been assumed to remain constant for convenience of 

 comparison of the analyses of the fresh and decayed rock in tables F. II, V, 

 VI, VIII, and IX. The figures in the tables of analyses disclose the 

 fact that the })ercentage of iron oxide in the fresh and decayed rock is 

 strikingly smaller than that of the alumina. While the ratio of increase 

 in the iron oxide, as pointed out above, is slightl}^ greater than that of 

 the alumina for the decayed and fresh rock, the writer has recalculated 

 in a similar manner all of the results : first, on the assumption of an 

 Al^Og constant basis; and, second, on the assumption of an Al^Og + Fe-^Oj 

 constant basis in order to make comparisons and point out the possible 

 ditierences resulting. The percentage amount of alumina is greater in 

 every case of the Georgia granites for the decayed product than for the 

 corresponding fresh rock, and is present in larger amounts than any 

 other single constituent where increase in the decayed over the fresh 

 rock is indicated. Briefly stated, a comparative stud}?" of the tables based 

 on the assumption of, first, an ALO3, and next, an Al,03 + Fc^Og con- 

 stant, shows : 



1. That on the assumption of constant AI2O3, the Fefl.^ like the Hfi in- 

 dicates a slight gain, owing to the ratio of increase being greater for the 

 iron oxide than for the alumina. 



2. The calculated percentage amounts of the constituents saved and 

 lost are in similar ratios for the three cases, but differ in actual amounts 

 in each one. The constituent showing the greatest loss under one as- 

 sumed constant likewise indicates similar loss for the other two assump- 

 tions. The percentage amount lost or saved for any one constituent is 

 not the same in the three cases. 



3. The total percentage loss for the entire rock is invariably greater 

 on the assumption of an Fe.^Os constant than for the other two assumpt- 

 ively insoluble constituents. In each case the lime salts have been re- 

 moved in larger amounts than the magnesia, and similarly the soda 

 has suffered a greater loss than the potash. The fact particularl}^ em- 

 [)hasized is the greater loss for the entire rock on the Fe.fl.. assumed 

 constant over that of the Al.^O. and Al.^0.^ + Fe.^Og assumptions. 



The tables show that the conclusions stated under 2 and 3 are gov- 

 erned directly by the numerical value of the ratio of increase — factor — 

 of the constituent assumed to remain constant in the change from fresh to 

 residual rock. 



