July 13, 1905] 



NA TURE 



259 



such minerals as the felspars the viscosity of the fused 

 substance may be of the same order as the rigidity of the 

 solid crystal approaching fusion, so that there is to the 

 eye no abrupt change. The discordance between the results 



temperature is to be regarded as a superheated solid or as 

 a liquid crystal, in which deorientation is prevented by 

 extreme viscosity. 



Specific Gravity 



r^rr;: 



Crystals of Bytownite from Middle oi Ci 

 mmorphism and Thermal Properties of the Fe 



of different experimenters is largely attributable to this 

 fact. The method followed was therefore to plot as a 

 curve the relation between temperature and time, and to 

 note the place where a change in the shape of the curve 

 indicates an absorption of latent heat. To avoid 

 the disturbing influence of impurities, the 

 several felspars to be examined were prepared 

 artificially. Thin slices of the crystallised pro- 

 ducts were studied optically by Prof. Iddings, 

 and they are illustrated in the memoir by a 

 series of beautiful plates. 



Anorthite was the felspar most easily crystal- 

 lised, and its curve gave a sufficiently sharp 

 melting point at 1532°- Other varieties ex- 

 amined had the compositions AbiAn^, AbiAn^. 

 Ab,An,, Ab,Anj, Ab^An,. These gave progres- 

 sively lower melting points ; but it was found 

 that, in passing from anorthite towards the 

 albite end of the series, viscosity rapidly in- 

 creases and obscures the phenomenon of fusion, 

 the break in the curve of heating becoming for 

 AbjAUj a barely perceptible deviation. For 

 albite, and also for orthoclase, the method fails 

 to give any result, and in a certain sense it may 

 be said that the alkali-felspars have no melting 

 point. In this connection, a special series of 

 experiments gave some remarkable results. A 

 small fragment of crystalline albite, embedded in 

 albite glass, was heated to 1200° and slowly 

 cooled. Thin slices showed that the crystal had 

 melted to a glass only along cleavage and other 

 cracks. The experiment was repeated with 

 higher temperatures of heating up to 1250°, and ' 

 it was found that, though the lanes of glass 

 encroached more and more upon the crystal, con- f-"^- =■ 



siderable relics of the latter were still left, pre- 

 serving undisturbed their original orientation. 

 It thus appears that a mineral like albite, which 

 melts to an ultra-viscous liquid, may be maintained for 

 half an hour "at a temperature well above its normal 

 melting point without being completely fused. It seems 

 doubtful whether the crystalline substance at such a 



We reproduce in tabular form the chief 

 numerical results obtained. The general con- 

 clusions arrived at are of great importance. The 

 melting point curve for the lime-soda-felspars, 

 as well as the curve of specific volume, is con- 

 tinuous, and not very different from a straight 

 line, and we have almost conclusive pi-oof that 

 this group of minerals forms a truly isomorphous 

 series. Further, it belongs to type i. of Bakhuis 

 Roozeboom, the melting point falling steadily 

 from one end of the series to the other. Here a 

 further point of interest arises. According to 

 theory, the crystals first formed from the fused 

 mass should be richer in anorthite than the liquid 

 from which they separate, and should contain 

 an increasing proportion of albite as crystallisa- 

 tion proceeds. Day and Allen, however, verified 

 ,. in several cases that their crystals had the same 



composition as the mother liquid. This can only 

 be due to undercooling, the beginning of crystal- 

 lisation being deferred until the temperature had 

 fiillen below the range proper to normal crystallisation. 

 Those natural rocks in which the felspar crystals show a 

 zoned structure (the outer zones richer in albite) must have 

 crystallised without undercooling, and, indeed, their felspars 







^' 







-Sph. 



/_ 



^ 





_Ma,imA 



-^^. 



^f" 



must have been formed within a certain range of tempera- 

 ture, which can be more or less closely determined. In 

 this and other petrological applications the work of the 

 authors affords a valuable supplement to that of Vogt. 



A. H. 



NO. 1863, VOL. 72] 



