50 



Scientific Proceedings, Royal Dublin Society. 



same is true of tlie manganese, wliicli may not have been obtained 

 completely free from iron. As an additional explana.tion, we 

 may point to the presence of minute included crystals in the 

 garnets, which must have contributed some foreign material. 

 Under the microscope, when examined in the thin slice of the 

 rock, the garnets appear colourless and transparent, presenting 

 most commonly hexagonal sections, sometimes square ones ; thus 

 pointing to the rhombic dodecahedron as the form of the com- 

 plete crystal. The boundaries are frequently rectilinear and the 

 angles sharp ; but sometimes one or more corners are rounded off, 

 and a curvilinear outline results : in some cases, when a hexagonal 

 section is so situated with respect to the planes of foliation that a 



M.— 



Section through Garnet Hornfels, magnified 50 diameters. 

 G. Garnet ; a. Quartz ; F. Felspar ; M. Mica ; II. Ilminite. 



diameter joining two of its opposite angles lies perpendicular to 

 the plane of foliation, it may be noticed that the sides parallel 

 to this diameter retain their rectilinearity, while the angles at its 

 extremities are completely rounded off, so that the section is 

 bounded by two parallel straight lines at right angles to the 

 schistosity, and two curved ends conformable with it. Occasion- 

 ally, all the angles are removed, and the outline of the section is 

 curvilinear. Eounding off is generally observable where a crystal 

 projects against a plane of minor shearing. The crystals are 

 very generally traversed by cracks, which are sometimes straight, 

 sometimes curved, sometimes few, sometimes very numerous,^ 



