Measures of roundness for individual detrital particles need to be readily applicable and reasonably representative of what they purport to depict. A review of better known methods shows that the use of intercept measurements such as length, width, etc. introduce a flatness/sphericity factor into the computation. Use of curvature measures involving consideration of only the curvature of the sharpest corner is also not without defect: roundness values are understated and even mis-stated. Wadell's method of relating the mean curvature of n corners to the maximum inscribed circle is deemed the most representative. However, the complexity of Wadell's technique may be reduced without serious disadvantage by considering curvature only in regard to the two sharpest corners. It is also postulated that curvature measurements are most usefully made only along a single (the principal) plane.