270 Day and Shepherd — Lime-Silica Series of Minerals. 



this method were to be applied, for example, to the determina- 

 tion of the melting temperature of orthoclase or albite, or even 

 pure quartz, which have been shown to possess a viscosity 

 entirely comparable in magnitude with the rigidity of the solid 

 crystalline mineral ; almost any conclusion could have been 

 reached under these conditions. 



It is stated by M. Boudouard, for example, that all mixtures 

 of lime and silica between 30 and 90 per cent of lime melt 

 below 1500°. This certainly cannot be the case. Pure ortho- 

 silicate of calcium when heated in a platinum crucible will stand 

 without showing the slightest trace of melting while the plat- 

 inum containing vessel melts down. The temperature must 

 therefore be at least as high as the fusion point of platinum 

 (1720°). "We found no lime-silica mixtures richer than 60 per 

 cent in CaO which could be melted in platinum vessels. 



Apart from the uncertainty in the temperature measurements 

 offered by Boudouard, we shall undertake to show in its proper 

 place that there is no pure lime-silica compound corresponding 

 to akermanite and no tricalcic silicate. We are, therefore, 

 somewhat at a loss to explain in any satisfactory way how the 

 published curve which has attracted so much attention in Eng- 

 land was really obtained. 



Part II. Uxperimental. 



In this kind of investigation it is always desirable to begin 

 with a careful determination of the physical properties of the 

 pure components, although in the present case it must be 

 admitted that this was the most inaccessible and difficult por- 

 tion of the field over which we worked. 



amorphous condition. The use of subjective methods has misled Prof. 

 Doelter into picking out points upon these curves to which he attaches great 

 importance in the determination of eutectics. 



Apart from this general criticism, the particular optical methods which 

 Doelter employs appear to us rather limited in scope for the work they will 

 be required to perform. In the case of a mineral combination which is 

 neither an eutectic nor a pure compound, they are open to all the objections 

 of Boudouard's method (see above) and furnish no trustworthy information 

 whatever. It is also difficult to see how they can be used effectively to 

 determine unknown conditions of equilibrium. Doelter's method 4 (p. 6, 

 loc. cit. ), in which he places greatest confidence, appears to us to promise 

 immediate and serious difficulties of another kind. It consists in observing 

 directly with the microscope tiny grains of the substance to be studied as 

 they lie upon a tray of amorphous silica (quartz glass) in the furnace. The 

 glass tray, which is, of course, also heated, is in very unstable equilibrium 

 and therefore ready to enter into solution with almost any oxide or silicate 

 in contact with it at relatively low temperatures, and to produce what may 

 appear to be a melting point but which of course has no necessary relation 

 to that of the pure substance examined. Joly's old method of examining 

 mineral fragments on a platinum strip was much more trustworthy for 

 melting point work, although inversions in the solid state could perhaps 

 be advantageously studied in the new apparatus. 



For these reasons it does not seem to us wise to employ the subjective 

 methods when others which are reproducible by any observer are available. 



