32 W. M. Mulchings — Willemite in Slag. 



were at first supposed to be aj)atite, like the crystals previously 

 observed by me in a slag of somewhat similar nature.' Under 

 the microscope, however, they proved to have very high double- 

 refraction, of positive character. Chemical examination proved them 

 to consist of silicate of zinc, with a little iron. 



These crystals are as much as half an inch in length and of a 

 thickness up to - 7 h inch. The majority have no definite terminal 

 faces, but many have perfect rhombohedral end-forms. They are 

 pale yellow in colour, and distinctly dichroic, the oi'dinary ray 

 being the more strongly absorbed. Nearly all the larger ones 

 inclose a central core of dark amorphous matter. 



Thin sections of the slag itself show that it consists mainly of 

 olivine (Fayalite), principally in a confusedly crystalline form, but 

 with a fair proportion of definite crystals. There is comparatively 

 little residual amorphous base. Crystals of Willemite, of the same 

 dimensions as those seen free in the cavities, lie in the olivine in 

 all directions, and have been the first to form. The separation 

 parallel to the basal plane is strongly developed, and cross-sections 

 show, in addition to irregular cracks, a well-marked cleavage with 

 intersections of 120°. 



Both Fayalite and Willemite contain a large amount of crystals 

 of magnetite. This latter contains a good deal of intercrystallized 

 zinc-spinel, and is not very readily attacked by acid. It may be 

 separated out by decomposing the slag with acid, evaporation, 

 and solution of residual silica in caustic alkali. If the separated 

 crystals are then digested for some time in strong hydrochloric 

 acid, the magnetite proper is dissolved, leaving a residue of green 

 zinc-spinels, mostly as hollow shells, and fragments of octahedra, 

 but also showing some few perfect crystals. 



The slag in which these Willemite crystals occurred was produced 

 uninterruptedly during about 10 days' work of the furnace, and 

 during the whole of this time the crystals could be seen in any ball 

 of the slag which was examined. It may be interesting to note 

 that during another run of the same furnace, producing a similar 

 slag, with rather more than less zinc oxide, no Willemite was 

 formed, either in cavities or in the massive slag. All conditions 

 were the same, and the only difference in composition of the slag 

 itself was that, whereas on the former occasion it contained ahout 

 lijr per cent, of lead, the percentage of that metal had risen on the 

 latter occasion to 5. Thin sections showed that there was here very 

 much more amorphous, and nearly amorphous base, in which were 

 large numbers of beautiful idiomorphic crystals of Fayalite. 



Thus it would appear that a small amount of lead silicate, itself 

 glassy, had the power to cause much of the slag to solidify in the 

 glassy condition, and to prevent entirely the crystallization of the 

 zinc silicate. Observations of this sort, made on slags, may throw a 

 little light on the manner in which the structure and mineralogical 

 composition of rocks may be greatly modified by apparently un- 

 important amounts of some one constituent. 



1 " Mature," Sept. 15th, 1887. 



