In the first experiment, several cords were put in a cage and held in water at a depth of 5-15cm. To set air bubbles free from the cords, the cage were agitated a little while at the beginning. The temperature of water ranged 6-8°C during the experiment. The amount of absorbed water was determined from the increase of weight of the cord, w, and from the decrease of buoyancy upon it. The diameter, 2r, was measured with a micrometer, and the ength under constant loading, l, with a scale.
From the experiment we know that: (1) w-w0/w0, l0-l/l0 and r-r0/r0, where w0, l0 and r0 represent the initial values of w, l and r respectively, vary in a similar way with t, the duration of immersion of the cord in water (Figs. 1, 2 and 3); (2) The relation between w-w0/w0 and t follows the formula
w-w0/w0=Q0'(1-e-a0t)+Q1'(1-e-a1t)+Q2'(1-e-a2t), Q0', Q1', Q2', α0, α1 and α2 being certain constants. Q0', Q1' and Q2' depend upon r0, but the ratio, Q0':Q1':Q2' is independent of r0 α0, α1 and α2 are proportional to r0-n, where n<1 for α0 and n=1 for α1 and α2 (Tab. 1, Figs. 4 and 5).
In the second experiment, cords were immersed in coloured water and at intervals the coloured area was observed on a cross section of the cord. Thus we know that: (3) The colour-ing begins to take place on the surface layer of a cord and extends gradually toward the centre of it. The time T, in which the colouring reaches the centre of cord varies with r02 (Fig. 6).