152 ME. J. A. DOUGLAS ON CHANGES OF PHYSICAL [May I907, 



ance equal to the difference between any two of the contact-studs 

 on the large coils, and thus could be used in more detailed obser- 

 vations. These wires were connected at one end to springs which 

 prevented them from sagging when they became heated. 



The head of the micrometer-screw, used in measuring the 

 expansion of the ribbon, could be read to -0005 of a millimetre. 

 The platinum-ribbon was similar to that employed by Mr. Cusack (2), 

 a length of about 10 centimetres being used for each experiment. 

 When in use, it was surrounded by a trough of asbestos to exclude 

 draughts ; a thin film of mica was also used as a cover, which in no 

 way hindered the view of the microscope. 



The appearance of minute fragments of the glasses on the ribbon 

 was not conducive to exact determinations of their melting-points ; 

 a considerable range has therefore been recorded, the highest points 

 being those at which the glass was undoubtedly molten, the 

 lowest, those at which it seemed to become viscid. The variation 

 in their melting-points is hardly as great as the difference in their 

 composition would seem to suggest • however, it must be recog- 

 nized that the melting-point of the glass must be considerably 

 lower than that of the crystalline rock, and this difference may 

 vary, as do the differences in specific gravity, with the composition 

 of the rock. 



The following is a list of the melting-points of the rock-glasses, 

 as determined by the meldometer : — 



Rhyolite 1260° C. 



Shap granite 1235-1255° C. 



Peterhead granite 1215° C. 



Plauen'scher syenite 1165-1175° C. 



Tonalite 1150° C. 



Markfield diorite 1147° 0. 



Guernsey diorite 1125°C. 



Quartz-diabase 1085-1105° C. 



Whin-Sill dolerite 1107° O. 



Eowley-Rag dolerite 1100° C. 



Andesite (14 on p. 151) 1095-1125° C. 



Andesite(13 do. ) 1095-1120°C. 



Andesite (15 do. ) 1097-1100° C. 



G-abbro 1085° C. 



Clee-Hills dolerite 1070° O. 



The refractive indices of the rock-glasses were determined by 

 the microscopic method employed by Becker in the identification 

 of the felspars. The nicol-prisms are removed, and a diaphragm 

 put under the stage of the microscope ; then, on focussing down 

 with a high-power objective, a halo of light is seen to pass from 

 the substance of higher refractive index to that of the lower. 

 Thus, if the glass be of higher refractive index than the fluid in 

 which it is immersed, the halo will pass outwards. By immersion 

 in a series of fluids of known refractive index, the refractive indices 

 of the glasses may readily be determined ; the greater the number 

 of fluids employed, the more exact will be the results. 



The fluids used in the determinations here recorded ranged by 

 intervals of '0005 from aniline (1*588) down to copaiba balsam 



