292 Joty—On the Volume Change of Rocks and Minerals attending Fusion. 
Paper, the following particulars are given, details of the determinations being 
omitted. For the sake of distinguishing this curve from a subsequent curve it is 
called the A curve. 
Curve A.—GALVANOMETER SLIDE At JY. 
Sete Melting Pt. 
: Centigrade. 
Chloride of silver, . ee! Wg 451° 
Sulphate of silver, . ely Ce 654 
Potassium carbonate, a leo 834 
Sodium fluoride, : 58 ID) 902 
Copper oxide, . ! 5 Miles 1055 
In effecting these observations no difficulty was experienced in noting the 
moment of fusion. In some cases the solid bead always possessed an angular form, 
due to its crystallization. This suddenly disappeared at the moment of melting, to be 
resumed with a ‘‘ hitch” at the moment of freezing. In the case of copper oxide 
the remarkable phenomenon is presented of a densely opaque solid becoming a 
deep red translucent liquid upon its melting point being attained, a behaviour 
suggestive in connexion with the fact of the fine transparent red of this salt 
when used as a colouring matter in glass. With the foregoing data as deter- 
mining a thermometric curve, experiments were begun upon rock magmas. 
The problem first claiming attention is that of the thermal expansion of the 
slag or magma of certain important types of rock. The acid, the intermediate 
and the basic types will be sufficiently representative generally of the innumerable 
grades of chemical differences obtaining among plutonic rocks. But a difficulty 
at once presents itself. Without a preliminary fusion and intermixture of the 
constituents of a large mass of any porphyritic rock, a minute bead representative 
of its chemical composition cannot be prepared. It would be difficult thus to 
prepare a reliably representative bead of the rock. The difficulty may be sur- 
mounted by the use of the glassy representatives of the types of rocks enumerated. 
For the acid type an obsidian may be used. For the intermediate an andesitic 
lava; for the basic a tachylite or basaltic glass. In this Paper the several types 
are not thus systematically dealt with. It appeared probable that such a mineral 
as orthoclase, having percentages of silica and alumina, conforming to the acid 
type of rock, although not to the more acid types of rock, would in its behaviour 
indicate the properties of the acid magma. Less volatile matter would probably 
be present than in a pitchstone. The pitchstone of Arran contains some 5 per 
cent. of volatile matter. An obsidian experimented with also showed, by its 
boiling, the presence of a large amount of volatile matter. Prolonged heating will 
indeed expel this volatile matter, but such prolonged heating is to be as far as 
