WE have been engaged for some time on a study of the condensation of 2 : 4 : 5-triamino-6-hydroxypyrimidine (I) with glucose and with fructose. As Ohle and Hielscher1 have carefully established the optimal conditions for the condensation of o-phenylenediamine with glucose and with fructose in the preparation of tetrahydroxybutylquinoxaline, we have, from the outset, applied these conditions to (I). Glucose afforded a product (II) having? approximately — 81.5° in N hydrochloric acid or in N/10 sodium hydroxide solution, while fructose afforded a substance (III) with a slightly higher specific rotation, approximately — 86.6°, under the same conditions. At this point, a paper by Karrer, Schwyzer, Erden and Siegwart2 appeared describing the condensation of (I) with a variety of simple sugars using less elaborate conditions than ours. Their product from glucose had — 68.93°, and their product from fructose had — 46.0°, respectively, in N/10 sodium hydroxide solution ; these figures are significantly different from ours, particularly in the case of the product from fructose. Although final conclusions were not reached, Karrer and his collaborators considered their products from glucose and fructose to be respectively (IV, or V?) and (V, or IV?), the constitution of their product from glucose being indicated as (IV) with some degree of certainty.