Responses to specific sweet stimuli are specific for any one individual, but for that individual only. Comparisons as to degree of sweetness of different stimuli may vary from individual to individual. Rigid data are probably unattainable; it is only possible to determine the average opinions of a group of individuals.The following solutions were found to be equally sweet by the group of individuals employed:(i) 5.0% sucrose and 8.3% glucose; 10.0% sucrose and 14.4% glucose (which modified the figure previously found, after proper weighting, to 14.6%).(ii) 5.0% galactose and 2.5% sucrose; 10.0% galactose and 6.0% sucrose; 15.0% galactose and 10.0% sucrose.(iii) 5.0% mannitol and 2.47% sucrose; 10.0% mannitol and 5.54% sucrose; 15.0% mannitol and 9.8% sucrose.(iv) 2.0% sucrose arid 0.0049% 'soluble saccharine' (approximate only).(v) 0.0125% dulcine and 3.35% sucrose; 0.025% dulcine and 4.1% sucrose; 0.050% dulcine and 4.6% sucrose (all only approximate).(vi) 5.0% glycine and 3.25% sucrose; 30.0% glycine and 5.5% sucrose.(vii) 3.0% dl-alanine and 3.0% sucrose.While it has been shown earlier that the sweetnesses of sucrose and glucose, sucrose and lactose, glucose and lactose, sucrose and glycerol, and glucose and glycerol are additive when calculated in terms of glucose (but not in terms of sucrose), and it is now shown that this is true also for mixtures of galactose and glucose, and (probably) of dulcine and sucrose, this relationship does not hold for mixtures of glycine and sucrose, alanine and sucrose, and alanine and glycine. In such mixtures the total sweetness is less than that calculated in terms of glucose and usually less than when calculated in terms of sucrose.The relative sweetnesses of the group of compounds glucose, galactose, lactose, glycerol, and mannitol, compared with any one of them, do not change with concentration, within the limits of the experimental method used. On the other hand, the relative sweetnesses of these, and of fructose and dulcine, when compared with sucrose, change with concentration, the relationship being a log-log one. There is a similar log-log relationship when comparisons are made with fructose.A method is described for determining the value of iso-sweet iso-caloric mixtures of glucose and sucrose.A type curve can be constructed relating sweetness to concentration for glucose and the other compounds of its group. The relationship is log-log. A different log-log relationship between sweetness and concentration holds for sucrose.