48 The American Geologist. Jaimary. i89S 
trace of lOO is in fact in the obtuse angle ooi A loi or ooi A 201 
But, most frequently, it is known a priore where the direc- 
tion of negative extinction falls. Such is the case when it ex- 
ceeds 20° and reaches the values characteristic of labradorite 
or of bytownitc; such also when the feldspar is bordered by 
oligoclase. 
(3). The search for sections 010 is often very long. The 
stage devised by M. de Federov gives a useful means of cor- 
recting the position of sections near 010. He has pointed out 
a process for determining the direction of extinction; it con- 
sists in the search for the direction of rotation necessary to 
bring the nearest optic axis into the field, but it is not applica- 
ble to albite, nor to oligoclase No. 2, nor to anorthite. 
It is in sections 010 that it is easiest to study the growth- 
increments of the feldspars, and their changes of composition 
step by step with their successive consolidations. It is not 
rare to see labradorites, andesines and even oligoclases thus 
succeeding each other. But the dominant type is relatively 
very stable in each stage of the consolidation, and it is very 
easy, generally, to specify without any uncertainty the nature 
of the dominant feldspar. Whenever large crystals are sus- 
ceptible of this examination in the face 010 (and the case is 
very frequent) it is to be advised. 
(4). To what extent do the errors of orientation afifect the 
readings of extinction? The general epures (plates I-VII) 
can reply to this question with precision. Let us suppose an 
error of 10° made in the orientation of a section to be studied; 
in other terms, let us carry forward the pole of the section to 
the parallel 10° from that of 010. The epures give us, for each 
position of that pole, the extinction referred to the trace of 
010, and the angle of the trace of the cleavage 001 referred to 
this same trace. 
For albite this error of 10" in orientation of the chosen 
sections gives rise to extinctions varying from 15° to 25°. 
The mean error in oligoclase is, in the same manner, ±5°; 
it descends in andesine to ±4°, and then rises slightly to la- 
bradorite and anorthite. 
The mean error of one degree in orientation in the face 
010 does not amount to half a degree in the angle of extinc- 
tion on the cleavage 001. The theoretical reason for such a 
