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



it is absolutely necessary to prevent this evaporation. The following 

 method was found effective : — A short length of broad rubber-tubing 

 was passed over the head of the separator, so as to fit closely round 

 the base of the neck, where it was secured by an elastic band. This 

 formed a cup, which could be filled with mercury to above the level 

 of the stopper. The lower limb of the separator, carrying the small 

 stopper, was fitted into a wooden tube also filled with mercury. 

 Thus evaporation from either end was completely prevented. 



The apparatus was swung from a horizontal wire, and the slight 

 movement necessary for separation was obtained by passing one end 

 of a string through the elastic band mentioned above, and causing 

 the other end to be jerked at intervals by clockwork. The strength 

 and periodicity of the jerks could be regulated by increasing or 

 diminishing the length of the string. Thus, by causing the lower 

 end of the separator to tap against a wooden upright, the fluid 

 could be gently agitated for an indefinite period. 



The purified mineral, of known specific gravity, was removed 

 from the separator, washed with benzol, and dried in a water oven ; 

 it was then ground down to the finest powder in an agate mortar, 

 and melted in the platinum-furnace, as described above. The 

 specific gravity of the homogeneous glass thus formed was measured 

 in a diffusion-column, and the percentage -increase in volume 

 calculated. 



The following list gives the result of these determinations, and 

 it will be seen that the increase of volume becomes gradually less 

 in passing from the soda- to the lime-end of the series. This, 

 however, is to be expected from a study of the change of volume 

 in the igneous rocks. The Roman numerals refer to analyses in 

 Prof. Carl Hintze's * Handbuch der Mineralogie ' vol. ii (1897) 

 pp. 1470, 1492, 1531, & 1549. 



Sj). Gr. Sp. Gr. Chavge Percentage 



of of of increase 



Crystal. Glass. Sp. Gr. of volume. 

 XXV. Albite (Pfitschthal, Tyrol)... 2625 2-373 -252 10*61 



LXXXVI. 01igoelase(Tvede- I 2 . Q5Q 2 . m , m ^ 



st-Vcind ) •.«••.••••••... J 



Andesine (separated 1 ^ ^^ .^ 



from tonahte) J 



CLI-OLXII.^Labradorite(Lab-j 2 . m 2 ^ Q , m 5 . gg 



XIX-XXL Ancr^ite (Monte j ^-^ 2mb .^ ^ 



The following minerals were also experimented on : — 



CXLIV. Adularia(St.Gotthard).. 2'575 2-370 '205 8-65 



Leucite (Vesuvian lava).. 2-480 2-410 '070 2-90 



Apatite (St. Gotthard)... 3-197 2-972 -225 7'57 



Tremolite 2*990 2-780 '210 7"55 



Actinolite 3'040 2-810 '230 8*18 



Pargasite 3'109 2'790 '319 1P43 



Quartz, infusible. 



Andalusite is not fusible in sufficient quantities to obtain a zone of 

 glass. 



