Effect of salts. In studying flocculation of starch pastes Samuel investigated the effect of different metallic salts and found their action to be varied. Some lowered the gelatinization temperature considerably.

Viscosity of starch pastes. Viscosity of pastes made from the same concentration of starch vary (1) with different starches, (2) with its phosphoric acid content, (3) with the phosphorus salt content, (4) with the temperature to which it is heated, (5) with the degree of injury or grinding of the granule, (6) possibly the proportion of beta- and alpha-amylose, and (7) with the length of the starch chain. Chain length may in turn be affected by application of dry heat, by boiling with water, or by acid and enzyme hydrolysis.

Different starches. It has long been known in food preparation that starches from different sources need to be used in different concentrations, if the product to be served is stiffened to the same degree.

Woodruff and Nicoli determined the stiffness of starch gels from 5 per cent concentrations of starch. The results, arranged in order of decreasing stiffness are: corn, wheat, rice, potato, arrowroot, and cassava. Other differences noted were: Of the three cereal starches, rice was the most translucent and tender when cut; corn was the firmest, the cold gel being chalky white; and wheat was intermediate between these two. For the other starches, the potato paste was ropy and too gummy to leave the mold well so that it gave a poorly formed gel, although quite transparent; arrowroot gave a still softer and more transparent gel; and cassava starch gave only a viscous fluid. It was necessary to heat the gels to 90°C. for maximum stiffening, but heating of any one starch to 90°, 95°, 99.5°C. gave gels indistinguishable from each other.

Phosphoric acid. Samec's results showing the increased viscosity of phos-phated starch have been given in the discussion of phosphoric-acid content of starch. From these results he concluded that the viscosity of starch pastes could be reduced in two ways: (1) by removing the phosphoric acid and (2) by reducing the size of the molecule. In general, paste formation increased with increased phosphoric acid content, but this did not hold for wheat starch. Samec believes this discrepancy may be explained on the basis that phosphoric acid in wheat starch may be present as a compound with the proteins.

In general smaller granules gave richer or stiffer pastes than larger granules, which Samec ascribes to the relatively higher phosphorus content of the smaller granules.

Phosphorus salt content. In separating starch by electrodialysis Samec states the amylophosphoric acid isolated in this way proved to be a diabasic acid. The solutions of various amylophosphoric acid salts not only gave pastes with differences in viscosity but also affected other properties. Thus copper and iron amylophosphate paste separated badly, the calcium amylo-phosphate tasted like chalk, magnesium and ammonium caused a slight burning and pricking sensation on the tongue, respectively, and sodium and potassium were reminiscent of wheat flour.

Temperature to which the starch is heated. Although the starch may swell to a certain extent at lower temperatures, it may not reach its maximum stiffening power until a higher temperature is reached. Thus Alsberg and Rask found the maximum viscosity of wheat starch occurred at 95 °C. Woodruff and Webber have shown that a 5 per cent wheat starch paste must be heated to at least 90°C. before the resulting gels will form mold-able gels, such as would be served for puddings. Furthermore, rapid heating to this temperature gave a better-formed gel than did slow heating. But temperatures of 95°C. or above gave the firmest gels obtained with this concentration.

For cornstarch paste Alsberg and Rask found the maximum viscosity at 91 °C, but Richardson, Higginbotham, and Farrow's results indicate that maximum swelling of some cornstarch pastes does not occur until a temperature of 100°C. is reached.

Degree of grinding. Samec states: "The capacity of a starch to form paste decreases with continued grinding of the starch and the grinding also peptizes the starch to such an extent that it can pass through ultra-filters." Alsberg found that in flour ground until the granule walls are broken a larger percentage of the starch is dispersed. A higher percentage of such flour is required to produce a paste of definite thickness. His explanation of this is that starch from the broken granules passes into solution and does not thicken as much, hence a greater quantity is needed to produce a paste, for there are fewer whole granules to swell and occupy the volume of the suspension and thus produce stiffening or thickening. During gelatinization starch may absorb about 30 times its weight in water. The smaller the number of granules broken or dispersed, the smaller the percentage of starch required to produce a paste of definite stiffness.

The proportion of alpha- and beta-amylose. Samec says that paste formation depends on having a proper balance between the tendency to associate and to hydrate. With potato starch, and starches from similar species, this relationship is attained by the presence of both amylo-amylose and erythro-amylose, and by esterfication of the polysaccharid with phosphoric acid.

Length of the starch chain. Richardson, Higginbotham, and Farrow state that Staudinger has shown that for many types of long-chain polymers there is close correlation between the chain length and the viscosity of the solution. This relation was found to hold with sago starches but with other starches the relation was not found, so that it was suggested that other factors, such as distribution of length, rate of gelatinization, and degree of hydration of the chain molecules are as important as chain length in determining viscosity. Natural starches contain chains of varying lengths and it may be that the average chain length is greater for some species of starches than for others. However, it is known that dextrins have shorter chain lengths than starch, and they have little or no paste-forming properties.

When starch is dextrinized by heat or broken into shorter chain lengths by hydrolysis, either by diastase, acid, or other means, the stiffening power is decreased. Thus more of browned flour is required for thickening gravies and sauces than of the unbrowned flour.

Acid. When starch pastes are treated with acid, hydrolysis of the gluco-side linkages takes place and shortening of the molecular chain occurs. Richardson, Higginbotham, and Farrow suggest that the attack occurs at random along the starch chains, so that in any paste of acid-modified starch there is a wide distribution of chains of varying length. As hydrolysis proceeds an increasing number of shorter chains is produced, the reducing power is increased, and the stiffening power decreased. Hydrolysis with acid is more rapid as the temperature increases, which gives an explanation for adding lemon juice to the partially cooled gel of water, cornstarch, sugar, and egg yolk. Hydrolysis is also speeded by increasing the concentration of the acid.

Boiling and enzyme hydrolysis. Hydrolysis is also brought about by diastatic enzymes and by boiling with water, the last more rapidly under pressure at temperatures above boiling.

Commercial soluble starch. The commercial soluble starches should not be confused with soluble starch occurring naturally in food products. Commercial soluble starches have been treated, usually with acid, and although they are more soluble, their properties, such as stiffening power, are also modified.

The effect of acid upon the thickening power of starch has been known for a long time. If cornstarch, lemon juice, and water are boiled together, the resulting paste is not as thick as one made with the same amount of cornstarch and liquid, but without the acid.

Sugar and starch. Woodruff and Nicoli have reported that pastes made from 5 grams of corn, wheat, rice, potato, arrowroot, or cassava starch and 95 grams of water required heating to 90°C. or higher before the cooled gels were strong enough to retain the shape of a mold. When 10, 30, 50 or 60 grams of sucrose were added to the above quantities the three root starches formed increasingly softer gels with increasing amount of sugar, the 50 and 60 grams giving a sirup. Although the three cereal starches showed increasing transparency and tenderness with increase of sugar, a gel that would not mold was not obtained until more than 50 grams of sucrose were added. A sirup was obtained with 60 grams of sugar and the cereal starches.