at higher temperature, so that —— passes through zero at the point R, 
after which it assumes again positive values. 
This peculiar phenomenon can hardly be explained in another 
way than by assuming a retrogression of the hydration in solution 
with rising temperature, which causes the heat of mixing to become 
smaller negative, so that the denominator of the fraction first becomes 
zero and then again positive. 
This continues like this up to the transition point F, where another 
remarkable phenomenon is met with. In this point, where the rhombic 
modification passes into the monoclinic one, which is a transformation 
without splitting off of water, arises a strongly retrograding melting* 
point line too. Now, as the difference of state between rhombic and 
monoclinic Na 2 S0 4 disappears in the liquid state, we know with 
certainty that the heat of mixing of the two melting-point lines, 
is exactly the same in the point F, from which follows that the 
reversal of sign of — - must be attributed to the circumstance that 
Qsl now refers to the monoclinic modification, the heat of melting 
of which must be considerably smaller than that of the rhombic form. 
If we now proceed to higher temperature, complications appear, 
as we now approach the critical state, in which Vg—Vl and &<? 
become continually smaller, till they finally become zero in the criti 
In this point = — oo, so that becomes = 
In conclusion we may still remark that it follows from this 
consideration that the hydration continues to exist even attiecn 
of the system etber-anthraquinone that a partly retrogressive me ^ 
point line must exist also there, the supposition suggests 1 se ^ 
the whole phenomenon is in connection with the °^ ma . e 
compounds in solution, but it is exactly the case ether-ant raq ^ * 
that admonishes us to caution, because nothing is ”° ^ geeins 
compound between these two substances in solid state. ‘ 
better to me to assume that the molecules of the two 
approach each other very closely in solution, leaving 
whether this state can be called a compound or not. 
Anarg. Chem. Laboratory of 
Amsterdam, Sept. 1909. f/ ie University • 
