into hard little beads, and these acted like any other hard object, causing the radicles to bend to the opposite side.  Even extremely minute beads of the shellac occasionally acted in a slight degree, as will hereafter be described.  But that it was the cards which chiefly acted in our many trials, was proved by coating one side of the tip with a little bit of goldbeaters’ skin (which by itself hardly acts), and then fixing a bit of card to the skin with shellac which never came into contact with the radicle:  nevertheless the radicle bent away from the attached card in the ordinary manner.

Some preliminary trials were made, presently to be described, by which the proper temperature was determined, and then the following experiments were made.  It should be premised that the beans were [page 134] always fixed to the cork-lids, for the convenience of manipulation, with the edge from which the radicle and plumule protrudes, outwards; and it must be remembered that owing to what we have called Sachs’ curvature, the radicles, instead of growing perpendicularly downwards, often bend somewhat, even as much

Fig. 65.  Vicia faba:  A, radicle beginning to bend from the attached little square of card; B, bent at a rectangle; C, bent into a circle or loop, with the tip beginning to bend downwards through the action of geotropism.

as about 45o inwards, or under the suspended bean.  Therefore when a square of card was fixed to the apex in front, the bowing induced by it coincided with Sachs’ curvature, and could be distinguished from it only by being more strongly pronounced or by occurring more quickly.  To avoid this source of doubt, the squares [page 135] were fixed either behind, causing a curvature in direct opposition to that of Sachs’, or more commonly to the right or left sides.  For the sake of brevity, we will speak of the bits of card, etc., as fixed in front, or behind, or laterally.  As the chief curvature of the radicle is at a little distance from the apex, and as the extreme terminal and basal portions are nearly straight, it is possible to estimate in a rough manner the amount of curvature by an angle; and when it is said that the radicle became deflected at any angle from the perpendicular, this implies that the apex was turned upwards by so many degrees from the downward direction which it would naturally have followed, and to the side opposite to that to which the card was affixed.  That the reader may have a clear idea of the kind of movement excited by the bits of attached card, we append here accurate sketches of three germinating beans thus treated, and selected out of several specimens to show the gradations in the degrees of curvature.  We will now give in detail a series of experiments, and afterwards a summary of the results.

[In the first 12 trials, little squares or oblongs of sanded card, 1.8 mm. in length, and 1.5 or only 0.9 mm. in breadth (i.e. .071 of an inch in length and .059 or .035 of an inch in breadth) were fixed with shellac to the tips of the radicles.  In the subsequent trials the little squares were only occasionally measured, but were of about the same size.