470 PROCEEDINGS OF BALTIMORE MEETING. 



remelted and in a viscous condition. Carefully examined, they show a reticulated 

 surface. A partial analysis by Professor R. L. Packard gave — 



SiOa 34.84 per cent. 



FeA 16.78 " " 



AI2O3 (by difference from 30.04 per cent sesquioxides) . 13.26 " " 



Bases CaO, MgO, etcetera (by difference) 35.12 '' " 



100.00 " " 



Not much stress can be laid on this analysis, as the crystals prove to be merely a 

 shell with a rectangular network of enclosed matter. These ramifying tubes are 

 in places full of greenish birefractive matter. In patches iron oxides have separated 

 out. There are also occasional globules of copper. A basal section of one of these 

 large crystals shows in convergent light a uniaxial cross, very vague even in a thick 

 microsection, thus indicating a low birefraction. The clearer parts of the melilite 

 substance, which run like needles through the inclosed mass, are not parallel to the 

 side faces, but to a prism of the second order. 



2. These crystals pass gradually into others of the type commonly hitherto ob- 

 served in slags, in which the melilite has been more successful in disentangling 

 itself. We see that the previously mentioned larger crystals are in the nature of 

 poikilitic patches, bounded by cavities, the rounding of whose faces is not due to 

 remelting, but to inability of the melilite to mould into its form the whole inclosed 

 mass and completely master the drop-forming tendency of surface tension. 



The bright reflecting crystals are piled up in thin tablets, something like tridy- 

 mite, but in form bounded by faces at right angles to the basal plane and each other. 

 Their angle is rarely truncated by a small prism of the second order. Goniometric 

 observations confirm the tetragonal character of the mineral. 



A random section of the devitrifled slag shows globules of copper, magnetite 

 or iron growth forms and melilite. The latter is negative, uniaxial, with a refractive 

 index about 1.6, and birefraction varying from 0.008 to 0.020 decreasing toward 

 the center. 



These specimens are quite suggestive of the state a rock is in when the poikilitic 

 texture and lava stalactites were formed. The variable and rather high birefrac- 

 tion of this melilite match very well the facts encountered by C. H. Smyth, Jr.* 



3. Another interesting occurrence is of small scales of hematite lining a hollow. 

 They stand generally with their breadth perpendicular to the surface of cooling, 

 and hence lend to the surface a bloom like that which gives its name to grape ore, 

 and is produced by the same cause. Frequently a scale is found out of place, lying 

 flat to the surface. The}'' are but a couple of millimeters or so in diameter and verj^ 

 thin. 



The next paper was entitled — 



NOMENCLATURE OF THE FINE GRAINED SILICIOUS ROCKS 

 BY LEON S. GKISWOLD 



Mr Griswold's paper provoked an interesting discussion, in which J. E. 

 Wolff, B. K. Emerson and A. C. Lane took part. 



*Am. Jour. Sci., August, 1893, p. 104. 



