SIZE OF PARTICLES 



45 



that crystalloids are adsorbed by colloids and cannot be completely 

 removed by dialysis. Each crystalloid molecule or each crystalloid 

 ion may thus falsely represent the osmotic pressure of a colloid mole- 

 cule having perhaps a thousand times its mass. 



The coefficient of diffusion may be employed in the determination 

 of the molecular weight of crystalloids, but in the case of colloids 

 it gives information concerning only the average size of the particles. 

 The method is not much impaired by the increase in the size of the 

 molecule, because it is only the square of the coefficient of diffusion 

 which diminishes proportionately to this increase. The adsorption 

 of crystalloids, on the contrary, is also in this case a source of error 

 because every crystalloid molecule or ion acts as a team-mate of its 

 colloid particle and hastens its rate of diffusion. The objection to 

 the principles governing the calculation are mentioned on page 53. 



Before we come to concrete examples we shall mention one other 

 method which may enlighten us concerning the particle content of 

 a solution the conductivity. In a solution the electric current is 

 carried only by the electrically charged particles (ions). In an NaCl 

 solution this is done by the Na and Cl ions; in a Na^SC^ solution, 

 2 Na ions and 1 S0 4 ion, that is 3 ions, take part. The assumption 

 is that many molecules are completely or almost completely split into 

 ions; this actually occurs in the case of strong electrolytes when in 

 great dilution. The conductivity thus affords us fractions of the 

 molecular weight: minimum figures (values less than actual). 



My chief purpose in making these statements is to show what 

 facts may be deduced from the various methods used in determining 

 the molecular weight; they give only limiting values, so that no con- 

 clusion is to be drawn from any one method. 



The following remarks will show what difficulties stand in the 

 way when we try to learn the size of the colloid molecule. 



Among the colloids whose chemical composition is best known 

 are the soaps. 



As was found by F. KRAFFT and A. SMITS, very dilute soap solu- 

 tions showed a well-marked rise in boiling point; but this did not 

 rise in proportion to the concentration of the soap, as may be seen 

 from the following table by A. SMITS for sodium palmitate: 



