glass, it was again bent at a large angle, and descended perpendicularly into the damp sand. When, as in the above cases, radicles encountered an obstacle at right angles to their course, the terminal growing part became curved for a length of between .3 and .4 of an inch (8–10 mm.), measured from the apex. This was well shown by the black lines which had been previously painted on them. The first and most obvious explanation of the curvature is, that it results merely from the mechanical resistance to the growth of the radicle in its original direction. Nevertheless, this explanation did not seem to us satisfactory. The radicles did not present the appearance of having been subjected to a sufficient pressure to account for their curvature; and Sachs has shown[1] that the growing part is more rigid than the part immediately above which has ceased to grow, so that the latter might have been expected to yield and become curved as soon as the apex encountered an unyielding object; whereas it was the stiff growing part which became curved. Moreover, an object which yields with the greatest ease will deflect a radicle: thus, as we have seen, when the apex of the radicle of the bean encountered the polished surface of extremely thin tin-foil laid on soft sand, no impression was left on it, yet the radicle became deflected at right angles. A second explanation occurred to us, namely, that even the gentlest pressure might check the growth of the apex, and in this case growth could continue only on one side, and thus the radicle would assume a rectangular form; but this view leaves wholly unexplained the curvature of the upper part, extending for a length of 8–10 mm. [1] ‘Arbeiten Bot. Inst. Würzburg,’ Heft iii. 1873, p. 398.