664 LECTURE xxxvin. 



between each two of which there always lie a number of turns in the same direction, 

 which is opposite to that of those between the next such points. In long closely- 

 wound tendrils there are often five or six such points of reversal. Darwin has 

 already pointed out that this is not a peculiarity confined to tendrils, and still less 

 a specific consequence of the stimulus ; on the contrary, the occurrence of points 

 of reversal is a mechanical necessity. If a body which tends to coil up is fixed 

 at both ends so that neither end can twist round, coils in opposite directions must 

 of necessity occur,, in order to compensate for the torsion inseparable from the 

 coiling up. This behaviour of fixed tendrils can be imitated by cementing a narrow 

 stretched strip of caoutchouc on to one which is not stretched; on releasing the 

 former, it contracts and forms the inner side of a spiral, the outer side of which 

 is constituted by the strip which was not stretched. On now seizing both ends 

 and extending the double strip straight out, and then again bringing the two 

 ends closer together, spiral turns will be produced to the right and to the left, 

 as in a tendril which is fixed. If one end is set free, the strip will twist up and 

 become coiled into a spiral. 



Since all the movements of tendrils here mentioned result from growth, 

 they only take place when the external conditions are favourable for growth, and the 

 more energetically the more favourable the conditions are — that is when nutrition is 

 active, the temperature high, and the plant abundantly provided with sap, caused 

 by an abundant supply of water when the loss by transpiration is small. Given 

 these conditions, tendrils can, as I have shown, . carry on their nutation and 

 irritable movements, wind around supports and become coiled up, even in the- 

 dark (e.g. plants of Cucurbita Pepo, growing with the apical portions in a 

 dark vessel, and nourished by green leaves exposed to the light). 



As regards the mechanism of the irritable curvatures induced by contact (the 

 twining and coiling up of attached tendrils), as well as the coiling up of free tendrils, 

 there can be no doubt that it depends upon the process of growth in length 

 and its modification due to transverse pressure on the side which is growing more 

 feebly. Tendrils are irritable to contact or pressure only so long as they are 

 growing in length; a passing curvature due to irritation may, it is true, be equi- 

 librated again during growth, just as for example the passive curvature of growing 

 shoots caused by vibration. If the stimulus at the support continues for a 

 longer time, however, and the tendril twines round it, the difference in length 

 of the convex and concave sides becomes permanent, and can no more be com- 

 pensated. The cells of the convex side are proportionally longer than those of 

 the concave side (just as in roots which have curved downwards, and the nodes 

 of Grasses which have curved upwards) ; in the case of thick tendrils wound round 

 their supports the difference in length is so striking, that it is at once detected by 

 the eye, without measurement, as I have convinced myself in various cases. Recent 

 experiments by De Vries, who marked tendrils which were still straight with cross 

 divisions, and measured them after they had' twined or coiled up, have shown that 

 the growth of the convex side is more pronounced and that of the concave side 

 less so, than is the case with regions on the same tendril which remain straight 

 above and below the curved portion. A tendril of Cticurbiia Pepo, for example, 

 had coiled round a support i-2 mm. thick; after the completion of the curvature- 



