106 Professor Orenville A. J. Cole — On Flame- Reaction. 



Albite, Amelia Court House, Vii'ginia. No result. Kj := '^S. 

 Albite, Zoptau, Moravia. No result. 



Grade 1. — Apophyllite, Squire's Hill, near Belfast. Kj probably 

 := 4 or 5 per cent. Analyses of other apophyllites give 3'10-6'30. 



Biotite, Miask. This is a low result, but one analysis gives KjO 

 as low as 5"61, while the potash in biotite from other localities may 

 sink to less than 1 per cent. 



Grade between 1 and 2 (1'5). — Elaeolite, Brevig. KjO = 5'17. 



Elseolite, Magnet Cove, Arkansas. Kj = 5-91. 



Anorthoclase, Pantelleria. KjO varies from 2-53 to 5'4o. 



Obsidian, Lipari. Kj = 5'1. 



Pitchstone, Corriegills. KjO^^-T. 



Groundmass of Phonolite of the Schlossberg, Teplitz. The bulk- 

 analysis of the rock has Kg = 6-57. 



Groundmass of Phonolite of the Schlossberg, Briix. This is full 

 of small nepheline crystals. 



Grade 2. — Muscovite (probably Eussian). K^ probably := 9 or 

 10 per cent. 



Biotite, Burgess, Canada. Intensified to 25 by H CI. Kj 

 probably about 8 per cent. 



Grade between 2 and 3 (2-5). — Porphyritic Orthoclase (Sanidine) 

 in trachyte of the Drachenfels, near Bonn. Average of five analyses 

 gives Kg = 9-7. 



Groundmass of Phonolite, Schloss Olbriick, Eifel. Rich in 

 minute leucite crystals. Compare with the figures above given for 

 phonolites rich in nepheline or nosean. 



Grade 3. — Microcline, Pike's Peak. 



Orthoclase from drusy cavity in granite, Slieve Donard, Mourne 

 Mountains. 



Orthoclase (Adularia), Schwarzenstein, Zillerthal. 

 Leucite. K^ = about 20 per cent. 



Evidently all true orthoclases, with their Kg := about 13 per cent, 

 (theoretical 17 per cent.), come in grade 3. Soda-orthoclase will 

 give 2'5, and anorthoclase 1'5 or even lower. 



Spodumene, with its good lithium flame visible to the naked eye, 

 gives no result throiigh the blue glass used in these experiments. 



Lastly, the advantages claimed for the employment of sodium 

 carbonate in place of gypsum are: — (1) The certainty in each case 

 that the sodium flame is clearly differentiated from that of potassium ; 

 we have a large quantity of sodium present, and we have eliminated 

 its efi'ects. (2) Complete decomposition of the assay. (3) Security 

 against loss of the assay when picked up on the moistened bead 

 and inserted in the flame. It is quickly fused in and absorbed. 

 (4) Since the operation is always performed in the presence of 

 sodium, there is no need for elaborate cleaning of the wire after 

 each experiment, or for the use of distilled water. 



