TKANSACTIONS OF THE SECTIONS. 75 



gangue is a compact micaceous haematite ; tlie locality, traced to an old sale 

 catalogue of Heuland's, is the Rosario Mine, Chili. 



It is singular that other specimens of the same mineral should have been 

 found among the specimens preserved in the British Museum. One of these is 

 associated -with chrysocolla and ochre on a quartzose veinstone, another occurs as 

 a thin crust on a schorlaceous rock, both being from a Chilian locality. A specimen 

 recently obtained is associated with quartz and eisenkiesel, and is from the Mina 

 del Limbo, Del Salado, Copiapo, Chili. 



On Anclretusife. Bj Professor I^. Story Maskelyne, F.R.S. 

 A somewhat well-marked group of minerals would seem to justify the designa- 

 tion of the Dufrenite gToup, by reason of their having, as a common constituent 

 (or being capable of being so represented), a compoimd of which the formula ia 

 II3 Pj Og -}- R2 Hu Og ; R being Fe in the case of Dufrenite, 



Dufrenite being Fe, Pj O^ + Fe, H^ 0^, or, in Berzelian symbols, Fe P, Fe K3. 

 Peganite is II P,' il H3 + 3SE. 

 Fischerite is A\ P.'id lEj + 5H. 

 Cacoxene is Fe P, Fe Hj + 9H. 



Wavellite is 2lip, Al K3 + oil. 



A mineral recently found in Cornwall, and sent to the British Museum by Mr. 

 Tailing, may perhaps be referred to this group. It has been analyzed in the 

 Museum Laboratory, and Professor Maskelyne named it Andrewsite, in honour of 

 the distinguished President of the Chemical Section of the British Association, 

 Dr. Andrews, of Belfast. 



Andrewsite occurs in occasional association vtdth a bright green mineral in 

 brilliant minute crystals, presenting a strongly marked resemblance to those of 

 Dufrenite. This green mineral not ha'^dng been as yet, from the small amount 

 obtained of it, submitted to an analysis, is only provisionally termed Dufrenite. 



The Andrewsite which it sometimes thus accompanies presents itself in globular 

 forms or in disks with a radiate structure, and in habit cmiously resembles Wavel- 

 lite. Its colour is a slightly bluish green ; its svirface is generally formed of a very 

 thin layer of the mineral provisionally termed Dufrenite, crystals of which occa- 

 sionally stand out of the globules. 



The interior of the globules is sometimes homogeneous, and consists of radiating 

 crystalline fibres ; oftener one perceives an almost sudden transition from an outer 

 shell of some thickness, which consists of Andrewsite, into an inner core, formed 

 of a brown mineral. 



Seen imder the microscope, the two minerals appear to a certain degree to inter- 

 peneti-ate each other, so that the selection of material for analysis is a work of 

 much caution. 



The spherules usually stand on the projections of a quartzose veinstone, protruding 

 into a hoUow, and covered with a mass of limonite, sometimes carrying a drusy 

 crust of Gotliite, and studded occasionally with a few brilliant little crystals of 

 cuprite. The spherules are met with in one or two cases on cuprite formed round 

 a nucleus of native copper. Andrewsite, in fact, contains copper, four analyses of 

 separate specimens giving the percentages of 10-651, 10-702, 10-917, and 11-002. 



The analyses of Andrewsite have proved sufficiently concordant to justify the 

 formida 



3{Fe, P, Os+Fe, H, OJ + CuP, 0,. 



or .^iFeF", Fe HjI + Cuj'p; 



in which, however, a portion of the ferric phosphate is replaced by ferrous phosphate, 

 as in Yi^danite is frequently the case -with the two phosphates. 



