44 COLLOIDS IN BIOLOGY AND MEDICINE 



loids, and, on this account, the danger of considerable error is always 

 present. Moreover, in solutions with ordinary molecular weights, the 

 osmotic pressures are so great that very bulky apparatus would be 

 required. Thus, for example, the osmotic pressure of a 1 per cent 

 aqueous solution of sugar at 15.5 C. is actually 0.685 atmosphere. 



The difficulties in the direct measurement of osmotic pressure of 

 crystalloids do not exist however, in the case of colloids. Almost 

 any membrane keeps back colloids and the small rises are easily 

 measured. In order to remove the sources of error due to the pos- 

 sible presence of crystalloids, we employ membranes which are per- 

 meable for crystalloids instead of semipermeable ones (collodion 

 sacs, animal membranes). 



These physical methods for determining the molecular weight rest 

 on the assumption that the substance in solution is really broken up 

 into molecules (colloids cannot be vaporized). This condition does 

 not always exist in the case of crystalloids and only exceptionally 

 with colloids. With crystalloids these methods yield figures that 

 are either too low or too high. The former will occur if the sub- 

 stance is incompletely split up if two, three or more molecules 

 continue to be united in solution. Under these circumstances we ob- 

 tain one-half, one-third, etc., the osmotic pressure that a molecular 

 subdivision would show. The ultramicroscope and ultrafiltration 

 reveal in many solutions of bio-colloids particles of such a size as 

 show no molecular subdivision by other methods; we may assume 

 therefore that colloidal solutions for the most part contain molecular 

 groups, and that there is no prospect of determining the true molec- 

 ular weight by osmotic methods. 



The osmotic method yields a deceptively low molecular weight 

 if, for example, the substance is dissociated further than into mole- 

 cules. This occurs in the case of electrolytes. A very dilute NaCl 

 solution that has dissociated into Na and Cl ions shows twice the 

 true osmotic pressure, so that the molecular weight might be set at 

 half its real value. In this respect we may also make mistakes with 

 colloids whenever they are ionized. The osmotic method shows us only 

 into how many fragments a molecular complex breaks up in the par- 

 ticular solution. It may give either minimum or maximum values for 

 the molecular weight. Even in the case of crystalloids, the method 

 must be employed with due consideration of all the conditions in- 

 volved; for colloids it may be exceedingly deceptive. We know at 

 the outset, because of the enormous molecular weight of colloids, that 

 the lowering of the freezing point and the elevation of the boiling 

 point must be very small indeed, requiring most delicate measure- 

 ments. The matter becomes still further complicated by the fact 



