CHEMICAL FORMULAS OF MINERALS. 341 



The next step in the usual method, is to determine how 

 these constituents are combined ; how much of the silica 

 with the potash, and how much with the alumina. Refer- 

 ence is made to the possibility or probability of certain com- 

 pounds, which Chemistry alone can teach ; but aid is found 

 in the principle, that the number of atoms of oxygen in each 

 acid and base is usually some simple multiple, the one of the 

 other. If in the above compound, 1 of silica be united with 

 1 of potash, the ratio alluded to is 1 to 3 ; and if the alu- 

 mina be combined with the remaining 3 atoms of silica, the 

 same ratio holds. This is the mode of combination com- 

 monly adopted ; it is expressed in the following formula, the 

 dots as explained, indicating the oxygen : 

 &Si+£lSi 3 . 



The index 3 expresses the number of atoms of silica : had 

 the 3 been written as a prefix, thus, 3&1 Si, it would have 

 meant 3 atoms of a compound of silica and alumina. 



The formula might also be written with equal precision, 

 and without dividing the silica between the bases, as follows : 



(&+£l) Si*. 



In a similar manner, an analysis of Garnet affords the ra- 

 tio of ingredients as follows : 



3 of lime (3Ca), 1 of alumina (l£l), 2 of silica (2 Si), 



corresponding to the oxygen ratio 3 : 3 : 6 or 1 : 1 : 2. Ap* 

 portioning the silica to the bases, we have the formula 



Ca 3 Si-f-£lSi, 



in which the oxygen ratio for each member is 1 : 1. Idocrase 

 and Meionite afford other simple examples. 



3. In Feldspar, above cited, the protoxyd portion is often 

 not potash alone, but part soda or lime. Again, in Garnet, 

 the protoxyds, instead of being all lime, may be part magne- 

 sia, protoxyd of iron, &c. In each case, however, all the 

 protoxyds added together, make up the same specific number 

 of atoms as if there were but one alone. So the peroxyd 

 portion may not be all of it alumina, but part peroxyd of 

 iron, the amount of this peroxyd of iron being just equiva- 

 lent to the deficiency in the alumina; that is, an equivalent 

 not in actual weight, but in atomic weight. In Garnet, as 

 stated above, the oxygen ratio is 1 : 1 : 2 ; and whatever the 



