308 REVIEWS 



charcoal added to reduce the hematite present to the ferrous condition. 

 The smaller portion, after crystallization for twenty days, gave a 

 well crystallized yellowish-brown mass. Pyroxene crystals could be 

 seen with the naked eye. In the vesicles of the slag were remnants of 

 unmelted hematite, as well as newly crystallized hematite flakes and 

 brilliant spicular pyroxene crystals, sometimes i mm long. These 

 crystals showed distinct prismatic, pinacoidal and pyramidal faces, 

 pleochroism, and parallel extinction. In thin section, as shown in Plate 

 IV, Fig. 2, distinct porphyritic structure was observed, with idiomorphic 

 enstatite and olivine in a groundmass consisting of monoclinic pyroxene, 

 plagioclase, magnetite, and a small quantity of glass. The olivine 

 was in short crystals, completely transparent and colorless, of very 

 strong double refraction and parallel extinction. The greater part of 

 |he olivine crystallized in spherical concretions. The plagioclase of the 

 groundmass showed twinning with extinctions varying from io° to 27 , 

 hence, a labradorite. Its crystallization was earlier than the other 

 components of the groundmass. The augite formed aggregates of 

 prisms partly as small phenocrysts, but principally in the groundmass. 

 The order of crystallization was thus olivine, enstatite, monoclinic 

 pyroxene, labradorite, magnetite and augite, glass. The larger mass 

 (over 100 pounds), gave also an enstatite basalt (Plate IV, Fig. 1) with 

 crystals of both orthorhombic and monoclinic pyroxene, and olivine, in 

 a colorless groundmass. This groundmass appeared to be a completely 

 homogeneous colorless glass. Pieces of this glass, heated three days 

 at the temperature of red glow without melting, acquired a trachytic 

 crystalline habit of rough surface, and lost their original glassy luster. 

 The groundmass by this heating, developed a crystalline mixture of 

 tiny plagioclase and augite microlites, showing that long continued 

 application of heat to a supersaturated solution, even in solid condi- 

 tion, could bring about crystallization. In order to test the tempera- 

 ture necessary to produce crystallization, the following experiments 

 were devised. Six crucibles were filled with fragments of this slag and 

 placed in a row between the hottest part of the interior of the furnace 

 and the middle of the entrance chamber. At the end of a month it 

 appeared that the innermost four crucibles contained only glass, which 

 had strongly corroded the crucible" walls ; while in the fifth and sixth 

 crucibles (those within the chamber) crystalline products had formed. 

 An investigation of preparations from these crucibles showed that the 

 order of crystallization of the component minerals was the same 



