( 174 ) 



At the Iransilion temperature of tlie hydrate Na^ SO^ . 10 H^O 

 intersection takes place of the melting-point line of this hydrate and 

 that of the anhydrons salt and now the remarkable fact occnrs, that 

 the melting i)oint line of the anhydrous salt is a retrogressive one. 



If we now for the present fix our attention on the temperature of 

 the transition point, the hydrate and the anhydrous salt are in equi- 

 librium with the same solution and with the same vapour: so lo,/ 

 refers to the same tluid i)hase for these two solid substances, and 

 now it is the question whether this quantity iVgf can have another 

 sign for the anhydrous salt than for the hydrate. 



It is natural to assume that the solution in equilibrium with the 

 solid hydrate, contains hydrated molecules, and when this is so, the 

 quantity lOsf may greatly deviate for the anhydride from that for 

 the hydrate, for then as was said before, («,ƒ),, can become i)ositive 

 in consequence of the chemical reaction, on account of which zt;,/ too 

 can assume a positive value. 



In this case the numerator is negative, whereas the denominator 



dxi 

 has remained positive, in consequence of which / ~-^ has become 



negative 



So the fact that the melting point line of the anhydrous salt is at 

 first really retrogressive is an indirect proof for the presence of 

 hydrated molecules in the solution discussed here. 



Now we see further that the melting-point line of the anhydride 

 bends again to the right at higher temperature, and so it takes a 

 course to the right. 



This too, might be anticipated, as was set forth in the intro- 

 duction, for at rise of temperature the hydration will decrease, 

 and with it (v),, will become smaller and smaller positive, pass 

 through zero, and finally assume again a negative value. So w .,/ 

 too will pass through -zero, and become negative again, and a 

 consequence of this will be that for a certain ?Z(?^a^iW value of w;^/ the 



dxi 

 numerator becomes zero, and so also T —- , so that the meltmg-point 



at 



line gets a vertical tangent. 



dx 



Above this temperature T— is again positive, and it would have 



dt 



been possible that as far as the sign of T— was concerned, no 



more change had occurred up to the melting point of Na, SO,. If 

 however, like here, we are dealing with a system of the type of 

 ether— anthraqiunone, we must get the case already mentioned in 



