IS'EW VARIETY OF CARBONATE OF LIMIL. 361 



equal axes, each of the six lower edges of Mhich was re- 

 placed by a bevel with two facets, s, s". Hence it follows, 

 that, if we suppose these facets prolonged till they meet, 

 so as to conceal the faces g, g-, of the equiaxal rhomboid, 

 the crystal would bo a dodecaedron with scalene triangles, 

 analogous to that of the metastatic variety, commonly called 

 dog-tooth spar ; and if we were farther to imagine planes 

 passing through the edges x, .r, &:c., these planes would in- 

 tercept a rhomboid similar to the equiaxal, and which with 

 respect to the dodecaedron would have the same position, as 

 the primitive rhomboid has with respect to the metastatic 

 dodecaedron; so that the equiaxal rhomboid may be consi- 

 dered as a hypothetical nucleus with respect to the dodeca- 

 edron before us. We have already an instance of a hypo- 

 thetical nucleus of the same kind in the paradoxal carbonate 

 of lime discovered by the learned Mr. Tounellier, keeper 

 of the mineralogioal collections of the Council of Mines. 

 But in the latter variety the hypothetical nucleus is the in, 

 verse rhomboid ; and it is remarkable, that the forms hi- 

 therto exhibited by these sorts of hypotheses are engrafted, 

 as it were, on the two secondary rhomboids, the most sim- 

 ple among those that belong to carbonate of lime. 



It was easy to see at once, that the facets s, s", must de- ^^^ structure. 

 pend on a law intermediate to the angles E E of the nucleus, 

 Fig. 2, which is likewise the case with the paradoxal variety. 

 Now in this there are two lines of particles subtracted from 

 the edges D D, and only one from the edges B B, which is 

 the most simple of combinations of this kind: and if we 

 add the condition, that the hypothetical nucleus is the in- 

 verse rhomboid, it necessarily follows, that the intermediate 

 decrement takes place with a single row. In the variety 

 discovered by Mr. Mabru the two terms of combination are 

 greater by unity than in the preceding, that is, there are 

 three lines of particles subtracted from the edges D D, and 

 two from the edges BB; and combining these data with the 

 condition, .that the hypothetical nucleus is the equiaxal 

 rhomboid, we find, that the intermediate decrement is at 

 the same time mixed, and takes place by five rows in breadth 

 and six in height. Any other law would give for the nu- 

 cleus a rhomboid diflferent from the equiaxal. For instance, 



if 



