54 
METHODS OF PETROGRAPHIC-MICROSCOPIC RESEARCH. 
quadrants in the image move perceptibly the movement in the one quad- 
rant being away from the center while in the adjacent quadrants it is toward 
the center. On further rotation of the analyzer these directions of motion 
are reversed, the effect in each quadrant, during a complete rotation of the 
nicols, being a pendulum-like movement, the points swinging out as the 
principal plane of the analyzer passes from a position normal to the diagonal 
of the quadrant to one parallel with it, and swinging in as it returns to its 
original position. This periodic contraction and expansion of the different 
quadrants in the image field means a difference in objective magnification 
for the two principal positions of the analyzer, as is indicated by the series 
of measurements of Table 2. These were made with the microscope (Plate 
i, Fig. 3) by using a o.i mm. micrometer scale as object and a similar o.i 
mm. scale in the lower focal plane of a Ramsden ocular as micrometer scale 
FIG. 38. 
for the image. Tests were made with three different objectives (Fuess No. 
O and Zeiss apochromats i6mm. and 4 mm. respectively), and with two 
different analyzers (Thompson type, the one 27 mm. long and 10 mm; wide; 
the other 19 mm. long and 7 mm. wide). After careful adjustment of the 
microscope, the micrometer scales were placed parallel with the vertical 
cross-hair and centered; and readings taken, on the upper micrometer scale, 
of the number of divisions from the center of the field to a given division 
in the image of the lower object scale. Similar readings were then taken 
on the same points after the analyzer had been rotated through 90, in which 
position its principal section coincided with the graduated line of the 
micrometer scale. Thus with the large analyzer and the Zeiss apochromat 
1 6 mm., the distance from the center of the field to the 4th division of the 
object micrometer scale measured 47.75 divisions on the eye-piece microm- 
