“…We further agree with Harvey 4 that for very precise comparisons, care must be taken to compensate for the very small temperature shifts resulting from revisions in the accepted standard for the International Temperature Scale over the years. However, although Harvey 4 specifically refrains from commenting on the validity of the mixture model description of liquid water, focusing on whether the uncertainty in the Tilton and Taylor 3 data is best represented by the authors' stated value, 3 1ϫ10 Ϫ6 , or by the average standard deviation of their smoothing equation, 3 1.2ϫ10 Ϫ6 , or by a ''reasonable estimate'' of the result of a modern uncertainty analysis, 4 2 ϫ10 Ϫ6 , suggests that the mixture model analysis can be thought of as another empirical fitting function comparable to others 3,[5][6][7][8][9] in the literature. Such a conclusion loses sight of the physical motivation underlying the mixture model analysis which leads to physically meaningful fitting parameters not found in other treatments.…”