II. L. Wells — Composition of Pollucite, etc. 213 



Abt. XXI Y. — On the Composition of Pollucite and its Occur- 

 rence at Hebron, Maine • by H. L. Wells. 



It is a matter of great satisfaction to announce the discovery 

 of pollucite in a new locality. This very interesting mineral 

 has heretofore been found only on the Island of Elba and even 

 there in very small quantities, so that it may be called a min- 

 eralogical rarity. Its composition, in being the only known 

 mineral in which caesium is an essential constituent, adds 

 greatly to its interest. 



Before describing the American material, some account of 

 the history of the mineral may be given. In 1846, Breithaupt 

 described"* two minerals from Elba, which he called Castor and 

 Pollux from their great similarity in appearance. He distin- 

 guished them easily however by their difference in specific 

 gravity. Castor is now considered to be identical with petalite, 

 and it is a fact worthy of mention that the latter mineral is 

 found at Peru, Maine, only a few miles from the new pollucite 

 locality, a fact which points, perhaps, to a new association of 

 " Castor and Pollux." Breithaupt's material was analyzed by 

 Plattner, f but at that time caesium had not been discovered, 

 so that he naturally mistook it for potassium. His results 

 were as follows : 



l. 



Plattner. 



SiO, 46-200 



A1 2 G 3 16-394 



Fe 2 O s 0-862 



K 2 __. 16-506 



Na 2 OJ 10-470 



H 2 2-321 



92-753 



Plattner sought in vain for an explanation of his low results, 

 and, not having enough material to repeat his analysis, he pub 

 lished it as it was. The discrepancy remained unexplained 

 until in 1864, eighteen years later and after Plattner's death, 

 Pisani§ discovered caesium in the mineral. Pisani states that, 

 if Plattner's analysis be re-calculated on the supposition that 

 the caesium was weighed as platinichloride while the soda was 

 calculated in the usual way from the weight of the mixed 

 chlorides, that the results would correspond closely to his own 



* Pogg. Ann., lxix, 439. % With trace Li 2 0. 



f Ibid., p. 446. § C. R., lviii, 714 



