Inj <>}>s< rvations in parallel light. 



57 



tioDB, and a varying angle to measure, it would be hard t<> 

 what weight would attach to the probable error as generally 

 computed. I have, however, picket! out a number of slides 

 containing the common hornblende of the amphibolites for 

 which k 2V has been determined. In live slides I have noted 

 the number of sections used in determining '2\ and weighted 

 the results accordingly. Twelve other slides have been com- 

 bined without weighting. Each set gives — 2V=79° with 

 probable error in one case of 0°11, in the other of i o# 08'. The 

 possible error is certainly much larger. 



Deducing p— a by comparison, and from that and '2\ , y — fi 

 and -i—"- (this may also be done graphically by the figure), we 

 find by these methods in the epidiorites and hornblende schists 

 a -cries often intergrown, tints: 





Ik 



!^ 



\\a 



: -« 



a-(3 



7-13 



-2V 



1. Actinohte 



slightly 



colorless 



colorless 



0-026 



■ 0-013 



0.013 



88° 





bluish 







decreases 

 always 



inereases 

 sometimes 



decreases 

 always 





to 



to dull 



to 



to 



to" 





to 



to 



2. Common 



bluish- 



olive 



yellow 



0.0 1 i) 



0-012 



0007 



79° 



hornblende 



green 



green 









and 



and 



and on 



to 



to 









keeps 



lower 



toward 



| 











on? 





glaucophane 

















The relations of increase and decrease may he seen very 

 clearly in cases of intergrowth. The series is not a linear one, 

 but we leave the full discussion* for the report of the Michigan 

 Geological Survey. 



EXPLANATION OF THE FIGURE. 



This figure is to illustrate the graphical determination of 2V from the double 

 refractions — or vice versa — as above referred to. It is founded on the formula' 



/cos 2V \/<i'—P \ 



eos 2V 



— cos 2 



2( ■ =1 — 1— cos 20— /;-2v/ 



\y— a I 



V — 2( — — -)— 1 — 1 -cos 2<p=x— 



2.V 



where 2Y=optical angle about c= . a' and }' arc at right angles to each other 



7 



and to >3. and <?=]_' 



* The change seems mainly due to imbibition of Na 9 Al 9 Si .,().. but this requires 

 chemical proof. If this is so, however, in view of this rapid falling off in }. con- 

 tinued in glaucophane. which contains about 65 per cent Nn. A i.Si ,<),, and of 

 the — extinction of the undetermined glaucophane mineral, described by Roten- 

 busch (Mikr. Phvs. ii. 319), it seems likely that the pure Xa,Al,Si ,< ). amphibole 

 has c=a. | Between writing this and reading proof, the description of it as Rie- 

 beckite, essentially Xa,.Al,Si,<).. has been given.] We have analogous relations 

 in acmite and segirine. 



