8 Mr. Levy o;j the Identity 



The ordinary form of the crystals of Heulandite drawn in 

 parallel position with fig. 1, is represented by fig. 2, each 

 plane bearing the sign which indicates the decrement by which 

 it is conceived to be derived of the primitive. The following 

 are the measurements of the angles. 



P,/i'=114.° 5' 7?^,/^'=138= 49' 30" ra,^=131° lO'SO" 



P, a* =130° 19' 50" a\ h^ - 64° 24' 40" a\ b^ =147° 22' 

 P, 6'=147° 8' 7H,6'= 75° 9' i',^*' =135° 52' 



b\g' = 112° 4'. 

 P, d' = 155° 25' 10" m, d^ = 132° 36' d", d' = 146" 31' 20" 



cP, o' = 106° 44'. 



The dimensions assigned to the primitive form of Heulandite, 

 which have been determined so as to make the results of 

 calculation agree as nearly as possible with those of obser- 

 vation, are such, that the line joining the angle o, (fig. 1.) 

 with its opposite, does not make a right angle with the edge //. 

 It is well known, that Haiiy had erroneously supposed that 

 in every primitive oblique rhombic prism this angle was right. 

 In the pi'esent case, however, as in many others, this angle 

 differs but very little from 90°, being equal to 89° 5'. If the 

 property here alluded to, and assumed by Haiiy, did really 

 exist, it would result from it, that every modification pro- 

 duced by simple or intermediary decrement on the edges or 

 angles of an oblique rhombic prism might be supposed to 

 be composed of half the number of planes of a modification 

 derived by some decrement on the edges or angles of a right 

 rhombic prism ; and consequently that to every such modifi- 

 cation there would correspond another produced by a dif- 

 ferent decrement, the planes of which would be situated with 

 respect to one pair of the lateral faces of the primitive precisely 

 as those of the first were with respect to the other pair ; the 

 planes of one modification measuring with each other, or with 

 the lateral faces o^ the primitive, exactly the same angles as 

 the planes of the other. It is easy to see that if 



\d'^ d'y h^J 



represents generally the sign of one of the modifications, the 

 sign of the other will be 



(& ^ + = b y + ~ h ~ ) 



And without entering into any discussion with respect to the 

 results offered by these formulae in particular cases, I shall 



only 



