HAPTOTROPISM 489 



and that it describes approximately a conical curve in space, because the zone of 

 more vigorous growth affects successively new surfaces. At first the axis of the 

 cone is directed steeply upwards, but it gradually sinks lower and lower below the 

 horizontal, in which position finally the circumnutations cease. The growth 

 which now takes place is very conspicuous, for the whole tendril, under suitable 

 conditions, may elongate from 50 per cent, to 90 per cent, in one day, and indi- 

 vidual zones may show extension amounting to 100 per cent. Growth is inter- 

 calary, for no apical elongation takes place. It is most vigorous in the basal half 

 and continues in that region for a longer time than in the apical half. After about 

 3 to 5 days the tendril appears to be full-grown, but more accurate measure- 

 ments show that it is still growing though feebly, i.e. about 0-5-2 per cent, in 

 24 hours. After this feeble growth has continued for a few days vigorous 

 growth recommences, although it never reaches its original activity. The 

 growth moreover is always unilateral, the upper side growing more rapidly, and 

 hence inducing the formation of a coil or spiral, the concave side being the under 

 side morphologically. This spiral coiling begins as a rule in the middle or in 

 the basal part of the tendril and spreads towards the apex. 



The specific sensitivity of the tendril, the capacity for twining round a sup- 

 port, makes itself apparent as soon as it has reached from |^ to ^ of its ultimate 

 length, and it may be still observed when spontaneous inrolling takes place. 

 The sensitivity ceases when the tendril is full grown. In order to understand 

 how the enclosure of the support by the tendril is achieved, we must study 

 first of all the incurving movement which takes place after a brief contact with 

 some solid body. If, for instance, the tendril of Passiflora or of Cyclanthera 

 pedata be rubbed on the under side with a splinter of wood or a pencil, after a few 

 minutes, or even a few seconds in some cases {Cyclanthera, 5-7 seconds ; Passi- 

 flora and Sicyos 25-30 seconds), a vigorous incurving takes place, the under 

 side to which the friction was applied becoming concave, and this incurving 

 progresses so rapidly that it is often possible to follow the movement with the 

 naked eye. After 30 seconds, or a longer period in less sensitive tendrils, the 

 curvature ceases and soon thereafter the tendril becomes straight once more, but 

 the undoing of the curvature always takes a much longer time than its formation. 



So far as the result is concerned it is by no means immaterial to which part 

 of the tendril the friction is apphed ; we find that the under side is not equally 

 sensitive throughout its entire length. The same stimulus produces a more rapid 

 response if applied to the upper third of the tendril than if applied in the middle, 

 and no visible response is given by the base of the tendril where the growth is 

 most vigorous. On the flanks the sensitivity also decreases from the base to the 

 apex, and it is generally less there than on the under side. Further, on the flanks 

 the stimulated part is always towards the concave side, but the curvature is 

 less marked and slower than when an equally strong stimulus is apphed to the 

 under side. Finally, if the upper side be stimulated, as a rule no curvature follov/s. 

 This is not the case, however, with all tendrils ; those of Cohaea scandens, Eccre- 

 mocarpus scaber, Cissus discolor, &c., curve quite as vigorously when the dorsal as 

 when the ventral side is stimulated. We shall find later on that these tendrils 

 also are physiologically dorsiventral. Such tendrils as those of the plants just 

 named may be considered as uniformly sensitive on all sides, as contrasted with 

 the others previously referred to, which were sensitive on one side only (Darwin, 

 1876a). FiTTiNG(igo3a) had shown, however, that the dorsal side of unilaterally 

 sensitive tendrils can receive a stimulus although it does not respond by curving. 

 The facts on which he bases this conclusion are as follows : If a tendril be touched 

 with a stick on a reacting side only a short distance away an incurving 

 takes place only at the stimulated place, and this curving is propagated about 

 2-5 mm. on either side of the region of stimulation. If two places be stimulated 

 about i|-2 cm. apart, two curvings occur, the region between remaining straight. 

 If the whole of one side of the tendril be stimulated from base to apex, the tendril 



