Palache and Wood — Crystallographic Study of Millerite. 349 



Forr, (1011) 



10 readings, very good signals, average value of p — 20° 43 '-4 



11 " good " " " /o=:20 42'-8 

 4 " fair " " " p = 20° 43'-5 



12 " poor " " " p = 20° 4:\'-b 



37 " all sorts " " " p-20° 42'-6 



For 79, (0221) 



3 readings, very good signals, average value of jo = 37°.07' J- 

 6 " good " " " " p = 37°03'i 



fair 



P : 



6 " poor " " " p = 37°17'£ 



12 " very poor " " " p = 37°12'| 



Meanof24 " selected " " " /0 = 37° 09'-0 



Mean of 15 " best " " " ^=37° 05'-8 



(The word selected as used above means that, using the 

 averages found in the columns above, the whole number of 

 readings whose signals were very good, good and fair was 

 summed with half the number of readings whose signals were 

 poor and very poor, and the general average then taken, 

 weighted in this manner to avoid giving undue influence to the 

 relatively large number of inferior readings.) 



Using the formula pp V3=tsm p and the above values, we 

 obtain, 



jD=i, p = 20° 42'-6, j!? = 0-21828 

 2 7=2, p = 3l° ()5'-8, ^ = 0'21830. 



Taking the value of p = 0*2183 we have from the relation 

 e=f p Q , a:c=l : 0-3274. 



This value of p , which is based on a large number of measure- 

 ments, is probably to be given preference over Miller's value ; 

 it is exceedingly near the value of p for beyrichite found by 

 Laspeyres. 



Found, millerite jp =0'2183 



Miller " p o =0-2197 



Laspeyres, beyrichite ^ = 0*2185 



Accordingly a table of angles based on the new value is 

 presented to replace that given in Golclschmidt, Winkeltabel- 

 len, p. 242. It contains the new forms, and in a supplement 

 those doubtful ones not yet fully established. This constitutes 

 Table I. 



