290 Messrs. C. H. Burgess and A. Holt, Junr. [Oct. 27, 



The melting point curve for the crystals is more easily explainable 

 than that of the glass. 



We were unable to obtain any point on it between pure boric 

 anhydride and a mixture of composition Na 2 0.4B 2 O 3 . This was 

 because the crystallisation of the small mass of substance in the beads 

 we employed for the melting point determinations took an extremely 

 long time, and also because the devitrification was never complete 

 except in the neighbourhood of the mixture Na 2 0.4B 2 3 . 



From a mixture of composition Na 2 0.4B 2 3 the melting point falls 

 nearly uniformly till 5Na 2 0.8B 2 3 is reached, at which point it begins 

 to rise rapidly. The mixture 5Na 2 0.8B 2 O 3 would then appear to be a 

 marked eutectic point. 



The summit of the curve is reached with composition 5Na 2 0.4B 2 3 , 

 when it falls again to a mixture which may be considered to 

 have the composition 5Na 2 0.4B 2 3 + 4Na 2 CO 3 , and which represents 

 the eutectic point between 5Na 2 0.4B 2 3 and Na 2 C0 3 . 



The further addition of sodium carbonate causes a gentle, almost 

 uniform rise in the melting points. 



As the summit occurs with composition 5Na 2 0.4B 2 3 , and as 

 this very nearly agrees with the analysis of the borate richest in 

 sodium, which can be obtained on fusion of boric anhydride and 

 sodium carbonate, this would seem to indicate a compound. 



The results of these melting point determinations of the crystalline 

 mixtures indicate that borax (Na 2 0.2B 2 O 3 ) is not a definite compound 

 in this state, but is almost a eutectic mixture of the borate with com- 

 position 5Na 2 0.4B 2 3 and one of composition Na 2 0.4B 2 O 3 , for the 

 existence of which we have other evidence. 



This melting point curve for the crystalline forms of these various 

 mixtures cannot be regarded as truly representing either the solidus or 

 liquidus. It seems probable that in this case they are situated close to 

 each other, and that our melting points really represent temperatures 

 close to the liquidus. The fact that a mass of crystals would apparently 

 melt almost completely at an almost constant temperature certainly 

 indicates that there cannot be any very great difference between the 

 liquidus and solidus, and, as in the cases studied by Heycock and 

 Neville, that the actually determined melting points lie probably very 

 near to the liquidus. 



The following table gives the melting points for the various glasses 

 and crystals : 



