1. The omission of the tension in the air-water interface has introduced an erroneous factor into Haines’ formulae; certain additional factors have also crept into his expressions for average stress.

2. With these corrections, the stress due to moisture varies comparatively little with changing water content, though falling slightly throughout the range. The energy needed to cause rupture rises continuously in a manner not unlike Haines’ measurements, and should more probably be associated with them than should the tensile stress.

3. The geometrical approximation used by Haines gives a close geometrical representation of the figure, but a mechanical approximation which is less satisfactory. Since neither the formulae connected with the true curve, nor the tables needed to use them, are readily, accessible, sufficiently exact numerical data have been here put on record.