OP ARTS AND SCIENCES. 257 



Not a trace of hornblende was observed in any of these rocks ; it 

 occurs with ortlioclase in some related coarser-gi'aiued rocks of the 

 Keweenaw series near the Wisconsin State line. 



The almost general acceptance of Tschermak's theory, which breaks 

 down the boundaries between the triclinic feldspars, has rendered the 

 determination of these a matter of wholly secondary importance in 

 diagnosis. 



But the feldspar plays such an important part in the series of 

 changes that have taken jjlace in these rocks, that it seemed desirable 

 to determine its position in the plagioclase series. This is no easy 

 task. Except in a few instances, chemical analysis must necessarily 

 fail, because the rocks have generally changed in character, — either the 

 felds[)ar or augite being more or less altered ; the composition of the 

 augite is an unknown quantity ; and isolated crystals of both minerals 

 are too small to be got out clean for analysis. The analyses I had 

 made were of little use in determining the feldspar. The entering 

 angle on the basal cleavage seemed to offer a ready means of deter- 

 mination by direct measurement. I therefore made a long series of 

 careful measurements with a reflecting goniometer, with very unsatis- 

 factory results; for the cleavage surface in one or both sets of hemi- 

 tropic individuals is apt to be either curved or crossed by innumerable 

 smaller striae, each giving an image, so that the result lies between 

 extremes often from thirty minutes to one degree apart, or more 

 than the difference between albite, oligoclase, and labradorite, and in 

 addition to this, different authorities give different measurements for 

 this angle. Still I give the results in the general table. 



Des Clolseaux's beautiful method is much more promising. This 

 consists in cutting very thin sections parallel to the basal cleavage, and 

 measuring, between crossed nicols, the angle between the edge | u. 

 and the principal section — maximum extinction of light, — or between 

 the maximums extinction in the alternating hemitropic bands to guard 

 against errors that might arise from obliquity in the section. Des 

 Cloiseaux determined this double angle, from many measurements, 

 to be: — 



Oligoclase 0° — 2° 



Andesine 3° — 4° 



Albite 7° 40' — 9° 40' 



Labradorite 10° 84' — 13° 56' 



Anorthite 40° — 80° 



Microcline 30° 54/ 



VOL. XIII. (n. s. V.) 17 



