MICKOSCOPY AND CRYSTALLOGRAril V 171 



indistinct cleavage, is a})parentl3' etched, suggesting the prohability of 

 the lime being slowly leached away. 



Figure G, plate 18, is a camera lucida sketch of a cross-section of the 

 crystal shown in figure 1, the calcite being indicated by stij)pling. 



The preparation of such slides offers unusual difhculties because of 

 the readiness with which tlie sand and silt grains part from the matrix 

 in grinding the section. 



Crystallographicall}^ these objects yield readily to inspection. They 

 are plainh^ of the hexagonal S3^stem, but belong not to the holohedral, 

 but to the hemihedral division. This is instantly apparent to the eye, 

 for instead of presenting 6 similar faces they present three })airs of faces. 

 Accordingly, similar edges and i)lanes come to view with each turn of 

 120 degrees. 



Of the six edges, three, lettered ,9, s, s, in plate 18, are visibl}^ more 

 obtuse and alternate with three which are less obtuse. Furthermore, 

 the three edges and faces forming the more obtuse angles give brilliant 

 reflections in sunlight, while the others are dull. 



The sand crystal is simply a scalenohedron with many of its true 

 characters interfered with and obscured by the sand inclusions. For 

 instance, the zigzag is apparently lost. This is a condition far from 

 phenomenal, for ordinary scalenohedra of calcite sometimes show no 

 zigzag, save to the critical observer. , 



However, in a few sand crystals it seems to be barely discernible, and 

 in others may be inferred from the fact that a given face does not lie in 

 a continuous surface of curvature, but is warped slightly. If a crystal 

 is held as shown in figure 2, plate 18, a portion, heavily shaded, re- 

 mains in view after the rest of the face is revolved out of sight. A 

 further revolution on the axis gives the result in figure 3, showing that 

 they alternate. This points directly to the scalenohedron, the faces of 

 which, if slightly curved vertically as well as laterally, would give pre- 

 cisely this result and would efface the zigzag. These are superficial 

 signs, but they unmistakably indicate the scalenohedron. 



Measurements confirm the eye determinations, but are subject to 

 error from the fact that the faces are surfaces of double curvature and 

 must be measured tangentially. 



The crystal shown in figure 1, plate 18, when measured carefully, gave 

 the following results, read in true angles, with supplement angles fol- 

 lowing in parentheses: (1) 113° (67°), (2) 128° (52°), (3) 116° (64°), 

 (4) 125° (55°), (5) 114° (66°), (6) 126° (54°). The error in reading 

 ordinarily runs from 2 to 6 or 8 degrees, being 2 in this case, as is shown 

 by taking the sum of the supplement angles. All measurements show 

 angles less obtuse, alternating with those more obtuse, at s, s, s. 



