8 BULLETIN 551, U. S. DEPARTMENT OF AGRICULTURE. 
PETROGRAPHIC EXAMINATION. 
In the samples analyzed a search was made by the petrographic 
method for the presence of the more important soil minerals. Those 
sought for and their formulas are: 
Quartz = Si0 2 . 
Magnetite =Fe 3 4 . 
Orthoelase = KAlSi 3 O s . 
Microcline= K AlSi 3 8 . 
Biotite=About (H, K) 2 {Mg,Fe) 2 Al 2 Si 3 12 . 
Muscovite = About H 2 KAl 3 Si 3 12 . 
P]agioclase=Isomorphous mixtures of XaAlSi 3 8 and CaAl 2 Si 2 8 . 
Hornblende=Chieny Ca(Mg, Fe) 3 §i 4 12 with Xa 2 Al 2 Si 4 O l2 and (Mg. Fe) 2 (Al, Fe) 4 
Si 2 12 . 
Epidote=HCa 2 (Al, Fe) 3 Si 3 13 . 
Titanite=CaTiSi0 5 . 
Garnet=R 3 R 2 (Si0 4 ^ 3 in which R is Ca, Mg. Fe, and Mn, and R is Al, Fe, and Cr, and 
rarely Ti. 
Apatite=(CaF) Ca 4 P 3 12 . F may be replaced by CI. 
Zoisite=Ca 2 (A10H) Al 2 (Si0 4 ) 3 . 
Cordierite=H 2 (Mg,Fe) 4 Al 8 Si 10 O 37 . 
Chlorite = Probably a mixture of H 4 (Mg,Fe) 3 Si 2 9 and H 4 (Mg,Fe) 2 Al 2 SiO a . 
Cyanite=Al 2 Si0 5 . 
Sillimanite=A] 2 Si0 5 . 
Tourmaline = A complex silicate of B and AL with Mg, Fe, or the alkali metals promi- 
nent. 
Rutile=Ti0 2 . 
Zircon=ZrSi0 4 . 
This search, it should be understood, involved an examination 
of a reasonable number of subsamples and the expenditure of a. rea- 
sonable time, the results being relative rather than absolute. Expe- 
rience in this work supports the conclusion that by extending the 
search indefinitely, nearly all the minerals sought would be found in 
traces in most soils. 
The results of the petrographic examination are presented in Table 
3. In this table P. indicates that the mineral is plentiful, S. present 
in small amounts, VS. in very small amounts, T. in traces. A plus 
( + ) or minus (— ) sign after a letter has the usual significance. 
The petrographic method as at present developed throws no light 
on the mineral character of soil material finer than silt, and as a 
consequence not only does it fail to give any information regarding 
the chemical character of that portion of the soil that probably is 
most reactive chemically, but may give rise to apparent discrepan- 
cies when the petrographic data are compared with the results of 
chemical analysis. 
