68 REPORTS ON THE STATE OF SCIENCE.—1917, 
and Colourists,’ 1916, 32, 40-44), the composition of the precipitate varies 
with the concentration of the barium hydrate according to the laws of 
adsorption. According to Tollens, starch combines with Na or K in the 
proportion of 1 atom to 4 (C;H,,0;) ; on addition of alcohol amorphous 
precipitates with alkaline reaction are formed; these are represented as 
potassium starch C,,H3,0.)K and sodium starch C,,H3,0.,Na. 
On heating starch with caustic alkalies a change takes place with the 
scission of acetic acid, the amount of alkali neutralised when N strength of 
solution is employed being equal to 8-33 per cent. of KHO. A similar 
change takes place with all carbohydrates, but the amount of acid liberated 
varies very considerably, pointing to differences in structural arrangement. 
The stiffening power of starch appears to depend upon the amount of 
swelling to which individual starches are liable on heating with water ; in 
other words, to the extent of the dispersion; thus, by heating to a few 
degrees above the temperature of swelling and centrifuging the sol, W. 
Harrison (‘ Jour. Soc. Dyers and Colourists,’ 1911, 27, 84-88) found the 
volume of the swollen granules to be directly proportional to the stiffening 
power, the experiments agreeing closely with the formula 
Vol. of granules 
Viscosity = 1 (4:75 x total vol. of solution) ; alkalies if present cause 
high results, as might be expected. 
Starch gel on standing separates into two phases, a more solid gel and a 
liquid phase, the latter containing very little starch ; several other colloids 
behave in a similar way, notably cellulose xanthate or viscose; some 
colloids, such as gelatine, do not behave in thismanner. The phenomenon 
is probably due to the aggregation of the molecular complexes, the colloid 
becoming less dispersed. Starch shows practically no osmotic pressure 
(/.e., only 2 m.m. at 13° for a 1 per cent. solution) in a parchment paper 
diaphragm (Moore and Roaf, ‘ Biochem. Jour.’ II. 39), neither does it show 
any appreciable depression by the freezing-point method of Raoult, pointing 
to a very high molecular weight (E. Fouard, ‘Compt. Rend.’ 1908, 
146, 978-981). On the other hand, pure soluble starch, according to 
L. Macquenne (‘ Compt. Rend.’ 1908, 146, 317-318), passes through filtering 
material, even the Chamberland filter, and E. Fouard (‘Compt. Rend.’ 1908, 
146, 285-287) found that a solution containing 2°74 per cent. of pure 
starch could be filtered through a collodion membrane. The molecular 
solution volume is lower than that calculated by Taube’s method (‘ Ber.’ 
1895, 28, 410), 7.¢., 92-6-93-3 instead of 102-6 (Cross and Bevan, ‘ Ber.’ 
1909, 42, 2,198-2,204), which the authors regard as evidence of ring- 
formation. Other colloids show a similar result. 
With regard to the action of reagents, other than those mentioned, upon 
starch, it is capable of forming a xanthic derivative in the same way as 
cellulose by action of caustic soda and carbon bisulphide (C. F. Cross, J. F. 
Briggs, and Société Francaise de la Viscose, Fr. Pat. 370,505, 1906), 
while a nitro-derivative of starch has been suggested as a substitute for 
celluloid (G. E. Arnold, A. 8. Fox, A. C. Scott, and H. E. V. Roberts, Eng. 
Pat. 3,450, 1906). 
Oxidising agents have a marked effect upon starch (see soluble starch). 
F. G. Perkins (U.S. Pat. 1,020, 656, 1912) proposes to prepare a glue from 
cassava starch by acting upon it with sodium peroxide and caustic soda. 
Starch and tannic acid mutually precipitate each other, this being also 
