So back to his point ...... The data is perhaps sketchy at best in the earlier part of the graphs he showed and on which we rely for such "precise" temperature changes
