N A PREVIOUS paper (8) it was shown that by meas-I uring with hypoiodite the amount of material in a starch, amylose, or starch product that is attacked by hot aqueous alkali, a number could be obtained that was an index to the peculiar make-up of the sample in hand. Slight changes in the starch or starch product that cannot be detected by such well-known determinations as viscosity, specific rotation, color with iodine, and initial reducing value, are, however, immediately reflected in a change in the stability of the product to the action of hot aqueous alkali. While this has been used in a qualitative way (3) by others, the subsequent quantitative determination by iodometric titration of the extent of attack by the aqueous alkali has made the method of interest to starch chemists.Every starch and starch product contains some material which is very quickly attacked by hot aqueous alkali and a part which is relatively slowly attacked by the same reagent. The latter part can be recovered substantially unchanged after the alkali has acted. The two parts have been called, respectively, alkali-labile and alkali-stable, the one being the antithesis of the other. Various pretreatments of the starches or their amyloses cause changes in the amount of this alkali-susceptible portion. By titrating iodometrically the alkali-digested solution of noncarbohydrate breakdown material having its origin in the alkali-labile portions, a number, expressed in milligrams of iodine per hundred milligrams of sample may be obtained. This is called the alkali-labile value.A somewhat analogous procedure, called the coppernumber determination (d), has been used to some extent in cellulose chemistry for a somewhat similar purpose. Both methods depend on reactions that have for their point of departure the attack of free or easily available aldehydic groups in certain parts or fractions of complex carbohydrates. The copper-number method is not, however, useful for application to the starches or starch products, one reason being that the hot aqueous alkali that causes the decomposition and the copper that measures it are both present during the operation. In the alkali-labile determination, on the other hand, the alkali is allowed to act, the solution is neutralized, and the effect of the hot aqueous alkali measured separately by hypoiodite solution, volumetrically. The small amount of cold aqueous alkali used in the iodometric determination produces itself no further alkali-labile material.The method under discussion is a semi-micro one and very sensitive to changes in technic. After being applied to several thousand samples by several analysts, certain discrepancies in results made their appearance. Therefore all apparent variables were investigated one a t a time and a modified method was evolved which, while giving slightly higher results than the older method, gives much greater precision. The values are still admittedly empiric, but with care and a little practice they can be duplicated and for comparative purposes serve very well. Deve...