67 



or a crude suspension of colloid matter in a dispersion medium, 

 increasing in uniformity of distribution as the subdivision of the 

 suspended particles is increased. Such a system allows almost 

 complete separation of the dispersion means by filtration through 

 an ordinary filter paper. The ordinary filter paper will hold back 

 particles having a diameter greater than about 5ju,, though, according 

 to the method of preparation, filters can be obtained which will stop 

 the passage of particles down to an approximate diameter of 2 ju,. 

 According to E. F. Armstrong (Brit. Assocn., 1909), the smallest 

 wheat starch granules vary from 3 to 5 ju., and the largest from 30 to 

 35 /A, so that the filtration of the water from the granules should be 

 almost complete. This, in practice, has been observed to be the 

 case. 



The simplest form of starch has been given (Brown, H. T., and 

 Morris, G. H., Jour. Chem. Soc, 53, 610, 1888) a molecular formula 

 [(Ci2H2oOio)2o]5- Lobry de BrujTi and Wolff (Bee. Trav. Chim. des 

 Pays. Bas., 23, 155, 1904) estimated the size of the starch molecule 

 to be approximately 5 /u,/i., as compared with hydrogen gas 0-067- 

 -159^11111, and water vapour 0-113 ju/x. Soluble starch prepared by 

 the action of ozone on common starch was examined by Friedenthal 

 {Physiol. Zentralbl., IS, 849, 1899). who obtained a molecular 

 weight of 9,450 as against 32,400 in the formula above. The product 

 obtained in this way was clearly more highly dispersed than ordinary 

 starch, a fact borne out by the definite depression of the freezing point 

 of water containing it in contrast to a suspension of ordinary starch 

 or to a starch paste. 



Depression of Freezing Point of Soluble Starch. 

 Concentration per cent. Depression of the Freezing Point. 



2-5 0-005 



5-0 0-01 



10-0 0-02 



On the figures of Lobry de Brujm and Wolff, " if a cubic centimeter 

 of dry starch could be subdivided into its molecules or dissolved in 

 the ordinary sense of the word, the starch would present a total surface 

 of several thousand square metres towards the solvent," and, in doing 

 so, would pass from an average size of its individual particles of 20 ^ 

 through the value - 1 ^, which represents the limit of microscopic 

 visibihty, to a value of 1 ixjx, a figure somewhat smaller than that 

 of a particle hitherto observed with an ultra-microscope. 



It is between the last two values that colloid chemistry has to deal, 

 according to Zsigmondy's system of classification. (Zsigmondy, R., 

 Zur Erkenntnis der Kolloide, XXII., Jena, 1905.) 



With regard to filtration of colloids through various papers and 

 diaphragms, a considerable amount of work has been done, the 

 most interesting for the immediate purpose being that by Bechold 

 {Zeitsch. physik. Chem., 64, 328, 1908). Bechold's results, from 

 his experiments with the pores of filter papers, cannot be taken 

 as absolutely accurate during prolonged filtration of colloid solutions, 



E 2 



