82 BOTANICAL GAZETTE [Aucusr 
volatilized and as occupying the same volume as that possessed 
by the solution. Thus van’t Hoff showed how the laws of 
Boyle and Gay-Lussac can be applied to dilute solutions. He 
was also enabled to make the following important extension of 
Avogadro’s hypothesis: Equal volumes of all solutions having 
the same temperature and the same osmotic pressure contain an 
equal number of molecules, which number is identical with that 
contained in a gas having the same volume, temperature and 
pressure. 
When Avogadro put forth his hypothesis that equal volumes 
of all gases under the same conditions of temperature and pres- 
sure contain an equal number of molecules, facts were found 
that apparently spoke strongly against this view. Thus it was 
observed that the vapor density of the chloride of ammonium 
was only a little more than half as great as was required by the 
principle of Avogadro, or, in other words, the molecular weight 
of the chloride of ammonium as calculated from the vapor den- 
sity was found to be only a little more than one-half of that 
expressed by the formula NH,Cl. This fact at first caused much 
opposition to Avogadro’s views, which was finally cleared away, 
however, when it was shown that in the vapor of the chloride of 
ammonium there are not simply molecules of that salt, but also 
hydrochloric acid and ammonia molecules, the products into i 
which ammonium chloride in the va 
ciated. 
The theory of van’t Hoff had to contend with a similar diffi- 
culty. While the behavior of many solutions was such as t0 | 
strongly support the theory, a large number of solutions (partic-_ 
ularly aqueous solutions of acids, bases and salts) showed com 
siderable deviation in their behavior from what the theory 
required, inasmuch as their osmotic pressures were greater than 
they ought to be according to the theory. The empirical results 
of Raoult,t which led to the methods for determining molecular 
weights of dissolved substances from the diminution of the 
vapor tension (or the elevation of the boiling point) of the 
4 Compare Ostwald, I 
por state is largely disso- 
-ehrbuch der allgemeinen Chemie 1:715, 748. 
