108 



BRIDGMAN. 



but is higher than that given by Guntz and Broniewski/^ 29°. The 

 value of Boyer is the mean of a number of values determined with 

 extreme care, and is seen to fit perfectly with the values which I have 

 found at higher pressures. Considering the extreme purity of this 

 sample, there is every reason to give this value the preference. 



4 5 6 7 

 Pressure, Kg. / Cm/ X 10' 

 Gallium 



Figure 8. The melting curve of gallium. 



The melting curve has the same form as that of water and bismuth, 

 the only other substances I know Avith curves of this abnormal type. 

 It is characteristic of these curves that the slope becomes numerically 

 greater at the higher pressures. The slope of the melting curve of 

 gallium increases numerically by about 15% in the pressure range of 

 12000 kg. 



The slope of the melting curve may be combined with other data 

 to gixe the latent heat on melting. Boyer's values for the densities 

 of the solid and liquid at the melting point are 5.90 and 6.09 respec- 

 tively. This gives 0.00529 for the change of volume per gm. on melt- 

 ing. Taking from the above curve the value 0.00203 for the initial 

 slope of the melting curve, and substituting in Clapeyron's equation, 

 we find 18.5 cal. per gm. as the latent heat of melting. Boyer has 

 recently determined this to be 19.1. The agreement is probably 

 within the error of the density determinations. The variation of the 

 latent heat along the melting curve cannot be found until the varia- 

 tion of the change of volume along the melting curve is also known, 

 and this will require a larger sample. 



The ratio of the resistance of the liquid to that of an equal volume 



