116 Mr Jackson, Molecular Weight of Glycogen. 
The glycogen, thus prepared, was tested for proteids and no 
trace of nitrogen could be found ; a weighed quantity was ignited 
in a platinum crucible and no weighable ash was left. It may 
therefore be assumed that the sample used was free from inorganic 
salts. 
In determining the molecular weight the cryoscopic method 
was followed, 30 cc. of water were taken and successive quantities 
of 2 grams of glycogen were added until 6 grams were present 
in solution. After each addition the depression of the freezing 
point w T as noted and it was thus possible to obtain three readings 
from each experiment. 
The depressions were extremely small, and the values of the 
molecular weights deduced from them varied between 9,500 and 
10 , 000 . 
These results point to glycogen having a highly complicated 
molecule, which links it closely to the dextrins derived from 
starch. 
The application of Raoult’s method to colloidal solution has 
been criticised. Recently Lobry de Bruyn ( Rec . Tr. Chem. 1900), 
in considering the size of particles present in colloidal solutions, 
shews that their diameters deduced from purely physical con- 
siderations, such as the size capable of polarizing the light 
scattered by them, are in agreement with those calculated from 
chemical data, such as the molecular weight determination. There 
is no distinction between true solution and colloidal solution ; 
there is no criterion of the homogeneity or heterogeneity of a 
liquid, and it is possible to pass continuously from undoubted 
solutions to liquids containing obvious particles in suspension. 
The author would like to express his thanks to Mr W. A. 
Hoffmann, B.A., of Christ’s College, for assistance in this investi- 
gation, and to the Grant Committee of the Royal Society for 
funds kindly placed at his disposal. 
