28 ART. 12. — II. HIROBE : ON THE FUSION SURFACES OF THE 



appearance of the eutectic points. Were the observations extended 

 to lower temperatures, doubtless the curves would become concave. 



As the curves (/. x) coincide so closely, the isothermals are 

 nearly straight lines parallel to the line JV C. From the devia- 

 tion between the curves D and E we are led to tlie conclusion 

 that the isothermal must be slightly concave toward P. 



The fusion surface of phenol is therefore almost a plane, 

 having a very small negative curvature, ])c'ing sliglitly concave 

 in the horizontal and convex in the vertical direction in the 

 region observed. 



As the melting point of chlorobenzene is too low for con- 

 venient measurement, its fusion surface has not been determined. 



F. The Eiiiectic Curve. 



Of the three eutectic curves of the system only one has 

 been studied, i.e. the line of intersection of the fusion surflxces 

 of naphthalene and phenol. On this curve six points have been 

 determined. The binary eutectic point is already given in Tables 

 2 and 9. Two other points have been determined by following 

 up the fusion curves of naphthalene, and the other three by 

 following up those of phenol. 



Id the observation of a point on the eutectic line the 

 temperature of the bath was kept some ten degrees below the 

 fusion point of the solution, which was steadily stirred in order 

 to have the liquid and the crystals in a state of equilibrium. 

 The temperature sank continuously while the crystals separated 

 out more and more. At a certain point t — Jt the temperature 

 began to rise owing to the separation of the other component as 

 the solid phase, reached the maximum point t in a short time, 



