1. It is shown that in the extracts of mangold leaves and stalks optically active impurities are always present which are not precipitated by basic lead acetate and hence vitiate the estimation of the dextrose and laevulose. These substances are possibly acid amides (such as glutamine and asparagine) or amino-acids (such as glutamic and aspartic acids) which form soluble lead salts.

2. These impurities occur in the leaves, but are much more abundant in the mid-ribs and stalks.

3. In the leaves the dextrose and laevulose appear to be present in approximately equal amount, as would be expected if they were formed from saccharose by inversion. When the ratio departs from unity it is probably owing to the presence of a dextro-rotatory impurity (glutamine?) which increases the amount of dextrose apparently present; but at certain times of the day a laevo-rotatory impurity seems to predominate so that the ratio becomes less than unity.