GEOPHYSICAL LABORATORY. 147 



(45) The Geophysical Laboratory. Arthur L. Day. Trans Am. Ceram. Soc, 15, 



49-54. 1913. 



An informal address, delivered at the fifteenth annual meeting of the 

 American Ceramic Society held in Washington on February 25, 1913, upon 

 the general subject of the laboratory study of silicates. 



(46) The change in the crystal angles of quartz with rise in temperature. Fred. E. 



Wright. J. Wash. Acad. Sci., 3, 485-494, 1913. 



In this preliminary paper the results of the crystallographic measurement 

 of quartz crystals at temperatures up to 1250° are presented. The crystals 

 were measured in the electrical furnace attachment to the Goldschmidt two- 

 circled goniometer (described above in Paper No. 17). The polar angle p 

 (51° 47'3 at 20° C.) of the unit rhombohedron decreases at a constantly 

 increasing rate with rise in temperature up to 575° (p = 51°36'7) at which 

 temperature the a-quartz inverts into |8-quartz. This is accompanied by an 

 abrupt decrease of about 2' in the angle p, which then, however, remains 

 practically constant up to 1250°; it increases, possibly, very slightly, but 

 the data of observation are not sufficiently accurate to determine the exact 

 amount, which is of the order of magnitude of tenths of minutes. 



A comparison of the changes in the crystal angles of quartz on temperature 

 rise, Avith the changes in specific volume, thermal expansion, birefringence, 

 circular polarization, and also excess of specific heat of quartz over that of a 

 normal silicate without a low temperature inversion-point, shows that the 

 differential changes between these properties are practically linear. A 

 detailed correlative study of the changes in the crystallographical and 

 physical properties of different minerals with change in temperature and 

 pressure should lead to a better understanding of the forces which produce 

 these changes. 



(47) A useful type of formula for the interpolation and representation of experimental 



results. L. H. Adams. J. Wash. Acad. Sci., 3, 469-474, 1913. 



It is often desirable, either for the sake of convenience in interpolation or 

 in order to discover accidental errors in a series of observations, to express 

 the results of an experimental investigation by some sort of empirical equa- 

 tion. For such purposes a power series — usually a quadratic or a cubic — 

 is employed by many people even in cases where it is obvious that such a 

 form of equation can not possibly fit throughout the entire range of the 

 observations. It is therefore advisable always to choose an equation which 

 conforms to the general character of the graph of the observations — a con- 

 dition which in general is easily satisfied. For instance, an equation con- 

 forming to the condition that the graph in the limit approaches a straight 

 line (a case of frequent occurrence in physical measurements) reproduces the 

 thermo-electromotive force of copper-constantan couples over a large range 

 of temperature much better and more conveniently than a power series 

 containing a like number of constants. 



