142 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



to depend entirely upon its own powers, yet, if it grows weakly, as when 

 not in the sun, or from any other cause acting to hinder strength when 

 growing — in such, if a stick is put into the ground close by it, it will twine 

 round it in loose spiral turns." 



Similar experiments have been carried out by later botanists ; thus 

 Noll found that the buckbean {Polygonu7i\ Fagopyrum) would climb. 

 I have tried the greater periwinkle {Vmca major), which, usually making 

 long shoots, which root at their ends, but does not climb, did so when 

 grown in the dark and the shoot was etiolated. 



This converse to the above also holds good. Normally climbing plants 

 can become creepers on the ground in the absence of supports. Thus our 

 bindweed may have half its shoots climbing up the iron rods of palings 

 while the rest creep along the ground, just as they do on a sunny bank, as 

 by a railway, where there is no support handy. Our maritime Convolvulus 

 soldanella does not climb, its habit having become adapted to a sandy 

 shore, like Ipomoea pes-caprcB in warmer climates. Similarly, in the 

 desert near Cairo, 0. lineatus and C. lanatus are short woody plants 

 forming little shrubs, while C. althcBoides both climbs and creeps on the 

 ground in Malta, like our own bindweed. 



As an example of the first step described by Professor Warming, I found 

 an illustration in Rtiscus hypophyllum. A number of large plants were 

 growing in pots in a very dark corridor in Malta. They had sent out long 

 etiolated shoots, which were intertwined among the green stems, but not 

 climbing spirally. 



A little more advanced case is described by Darwin in Hibbertia 

 dentata, which behaved at first just like the Busciis, but subsequently 

 climbed spirally.* 



In the lecture referred to above I mentioned the difference between 

 the two species of Ampelopais ; thus A. hederacea, the Virginia creeper, 

 does not produce adhesive pads until after contact has been made ; but 

 A. Veitchii develops them partially before any external irritation has been 

 set up. Hence this affords an excellent illustration of the acquirement of 

 an hereditary character by mere external irritation upon the soma, in this 

 case metamorphosed floioering branches. 



That the power to produce these adhesive pads is inherent in proto- 

 plasm generally is apparent from the fact that they are found in other 

 families of plants. Thus in Bignoniacece Darwin gives an interesting 

 description of the behaviour of Bignonia capreolata, which developed 

 cellular " balls " in order to envelop the fibres of cotton and flax supplied 

 to them. Darwin saw eight discs formed on the same tendril. He notes 

 also " the singular fact that a leaf should be metamorphosed into a 

 branched organ, w^hich turns from the light, and can by its extremities 

 either crawl like roots into crevices, or seize hold of minute projecting 

 points, these extremities afterwards forming cellular outgrowths which 

 secrete an adhesive cement and then envelop by their continued growth 

 the finest fibres." t 



Darwin adds in a note that " Fritz Miiller states that in S. Brazil 

 the trifid tendrils of Haplolophmm, of the same family, without having 

 * Climbing Plants, p, 35. t Ibid. p. 103. 



