| Abstract: | A previously described method of combining measurements of numerous pebble cross-sections in three orthogonal planes into a single average ellipsoid was found to be in need of improvement. The method of reconciling redundancies and of distributing the errors can be left unchanged. The original measurements of principal half-axes, however, rather than being normalized as the ratios of larger half-axes divided by the smaller, should be normalized so that they produce ellipses of standard area. The coefficients of ellipses so normalized will produce distributions with a pronounced central tendency even in those cases that produce markedly skewed distributions of coefficients for ellipses normalized by the former method. Equations of the previous paper that must be amended are given in their new form. |
