skip to content

Darwin Correspondence Project

From Fritz Müller   [12 and 31 August, and 10 October 1865]1

In your Paper on the “Movements and Habits of Climbing-Plants,” you say that you have seen no tendrils formed by the modification of branches, and you even seem to entertain some doubt whether such tendrils exist.2 In the genus Strychnos, the tendrils are called by Endlicher ramuli cirriformes, and I have now satisfied myself that they really are of this nature.3 On the branches of upright shoots of a Strychnos which grows here, the tendrils are disposed in a very regular manner. On the branches, the leaves of the first, third, fifth, &c. pairs are horizontal, those of the second, fourth, and sixth pairs are vertical in relation to the main axis; and it is from the angles of every under leaf of these latter pairs that the tendrils spring. Now, on the points commonly occupied by tendrils, true branches are sometimes developed. The leaves from the angles of which the tendrils spring are often much reduced in size, while in other cases they are but little or not at all changed. Each tendril bears near its tip a pair of rudimentary leaves; and whilst very young the tendrils are straight, but soon become curved downwards and rolled into a helix, whether they have clasped a support or not. This Strychnos is a very inefficient climber; the short stiff tendrils but rarely catch anything.

A member of the Hippocrateaceæ, probably a Tontelia, is likewise a branch climber. One of its branches, three feet in length, had not as yet developed leaves, and resembled a gigantic tendril, with most of its lateral branchlets already grasping neighouring objects. From the angles of the tendril-branches, other branches arise, which as far as I have seen, are not sensitive, and never clasp anything. This latter arrangement must be serviceable to the plant; for such branches grow upright without being arrested in their course, whilst the plant is secured by the tendril-branches.

Caulotretus, one of the Leguminosæ, offers another case of tendrils being formed from modified branches.4 In the species which I observed, the branches bear tendrils only in the angle of their first leaf, and this leaf is always rudimentary. In young shoots it might, at first sight, be thought that tendrils spring from the axils of all their leaves. In this plant every tendril appears to consist of two parts, separated by a small swelling—the inferior being straight, the superior curved, with its end rolled into a helix. But what appears to be the inferior part of a tendril is in fact the first internode of a young branch, the swelling being its terminal bud, and the tendril really springs from this young branch, from the angle of its first squamiform leaf, but nevertheless accompanied by two stipules. The end of the tendril very soon rolls up into a helix; but it does not lose by this the faculty of catching a support; on the contrary I know of no other tendrils which become entangled with small objects so easily as these rolled-up, highly elastic tendrils of the Caulotretus.

By far more interesting than the tendrils of Strychnos and Caulotretus are those (Pl. IX. figs. 1 and 2) of a climbing Papilionaceous plant with a woody stem, which from its general aspect I suppose to belong to the Dalbergieæ, Benth.5 They consist of thin, slender, flexible, leafless branches, with numerous (12–25) internodes, armed with sharp, hard, hook-like stipules. The young, soft, herbaceous shoots of this plant which rise from the ground are leafless. I saw one, seven feet high, which in its lower half was naked, while the upper half bore about a dozen tendrils stretched out in every direction. The oldest of these tendrils were from nine to twelve inches long, and armed with from twelve to sixteen pairs of sharp hooks: at the sides of the younger tendrils there were large, foliaceous, deciduous stipules, and at their bases very small bract-like leaves. The hooks of the tendrils are evidently stipules, which so often in this family assume the form of hooks or spines; in fact, while in the older tendrils they are strongly curved, and have a hard, sharp, darkly coloured apex, at the summits of the younger ones they are straight, soft, and green, resembling in this early state the much larger stipules at the bases of the tendrils. Afterwards, on the summit of the shoot, true leaves are developed at the bases of the tendrils instead of the small rudimentary ones; and finally, when the plant has reached the light, and spreads over the upper surface of a thicket or tree, the tendrils disappear. The inverse may be observed when the plant sets out on the conquest of a new dominion, a neighbouring tree for instance. Then a branch bearing only leaves begins to produce on its tip tendrils supported by leaves, and finally, growing rapidly to a long slender shoot, it produces only tendrils, the leaves being replaced by small squamæ. Thus in this plant, the branches assume four different shapes:—1st, tendrils, leafless, armed with hook-like stipules; 2nd, long, slender, leafless shoots, bearing tendrils and broad deciduous stipules; 3rd, branches with leaves, from the axils of which tendrils spring; and 4th, branches bearing only leaves without tendrils. Between the leaf and the tendril there is an accessory bud (fig. 1 b), which often developes into a branch; these branches issuing from the accessory buds seem never to produce tendrils. The tendrils, after having clasped a support, thicken partially where they are in contact with it (fig. 1a). Tendrils which have caught nothing behave in different ways. Some wither and fall after contracting irregularly. Others likewise become flexuous, or contract into a spire, or occasionally into a helix, but remain, thickening somewhat and becoming ligneous and rigid. Others produce branches from one or some of their internodes: this also occurs, and perhaps more frequently, with tendrils which have found a support; in this case the tendrils thicken much, and sometimes attain a diameter of more than one inch (fig. 2a, a thickened tendril clasping a branch of a Psidium; b, branch issuing from the tendril; c, tendril-bearing branch; d, branch from an accessory bud, without tendrils). Lastly, the tendrils even transform themselves into true branches: in this case they may remain nearly straight or become but little flexuous, and at their ends they produce leaves; the first of these leaves have sometimes hook-like persistent stipules, like those of the tendril, while the stipules of the following leaves are deciduous like those on other branches. These tendrils often become much elongated. I saw a shoot, almost all the tendrils of which were developed into serpentine branches; and under each of these branches there was a straight branch from an accessory bud. One of the tendrils was thirty inches long; it had twenty-five pairs of hooks, and at the tip three short internodes with leaves and destitute of hooks; from its seventeenth internode a branch arose. Excepting their hook-like stipules, by which they may be easily recognized, the branches formed by tendrils resemble in almost every respect the ordinary branches; but, as far as I have seen, they never produce tendrils, nor do the branches which spring from an internode of a tendril or (as I have already stated) from an accessory bud.

If we restrict the name of tendrils to filamentary organs used exclusively for climbing, those of the present plant would be excluded;6 for after having done their work as tendrils, they may be transformed into, and do all the work of, branches.

While in this plant the highly modified tendrils may be changed again into true branches, in two other plants which I have seen, the branches themselves, without having suffered any modification, act as tendrils. One of these plants belongs to the Dalbergiæ. Many of its branches had clasped small branches of a tree. These tendril branches, as they may be called, had not continued to grow beyond the support; and where they touched it, most of them had thickened: some showed a tendency to spiral contraction, forming a semicircle between the support and the stem. The plant does not twine. I may add that another genus, belonging to the same section of the Leguminosæ, namely Hecastaphyllum,7 is also a branch climber.

The second plant above referred to is a Securidaca (Polygalaceæ), and a most powerful climber (fig. 3). Its branches often curve in a very odd and complicated manner. Thus I saw a thin branch, which with its lateral twigs had become curved like ribs into semicircles (about four inches in diameter), imitating the bones of the thorax; from the twigs sprang secondary branchlets, which were very regularly curved, twisted together, and formed into a sort of network around the middle hollow space. When the branches wind round a support, they thicken and become more rigid, like true tendrils; but even these thickened parts may bear leaves or secondary branches. In the preceding plant the branches seem to be arrested in their longitudinal growth when they clasp a support; in the present plant they continue to grow, and the same branch may successively catch different objects. The branches which project freely from a thicket are rather thin and slender: with their twigs spreading all in the same horizontal plane and diminishing in length towards the extremity of the branch, and with their leaves arranged in two horizontal rows, they apparently form gigantic bipinnate leaves; and when covered with their bluish-purple flowers, this Securidaca is one of the most elegant and magnificent plants of our flora.

From the last two plants it is but one step to the primordial and simplest condition of branch climbers, exhibited by the numerous species which scramble up a thicket without twining and without the aid of rootlets, hooks, or tendrils.

Thus we can trace in the development of branch climbers the following stages:—8

1. Plants supporting themselves only by their branches stretched out at right angles—for example, Chiococca.

2. Plants clasping a support with their branches unmodified—Securidaca (Hippocratia according to Endlicher, Gen. Plant. No. 5700, “arbores v. frutices, ramis contortis scandentes”).9

3. Plants climbing with the tendril-like ends of their branches. According to Endlicher (Gen. Pl. No. 5745), this is the case with Helinus (“ramulorum apicibus cirrhosis scandens”).10

4. Plants with highly modified tendrils, which may, however, be transformed again into branches—for example, the above-mentioned Papilionaceous plant.

5. Plants with tendrils used exclusively for climbing—Strychnos, Caulotretus.

I will here add a few miscellaneous observations. You describe some species of Bignonia in which the tips of the tendrils become enlarged and adhesive after remaining for a short time in contact with some object;11 but the trifid tendrils of Haplolophium, one of the Bignoniaceæ, terminate (without having come into contact with any object) in smooth shining disks, which, however, after adhesion, sometimes become considerably enlarged.12 In Cardiospermum you state that the common peduncle which bears the subpeduncles with the flower-buds and the pair of short tendrils, although it spontaneously revolves, does not bend on contact or contract spirally;13 hence it may be worth mentioning, as showing a difference in the action of the tendrils in related genera, that in Serjania the common peduncle contracts spirally when the single tendril which it bears has clasped, as frequently happens, the plant’s own stem.14

With respect to spirally twining plants, you state that though the Hibbertia dentata sometimes revolves in one direction and sometimes in the other, yet it invariably twines from left to right.15 But in another genus belonging to the same family, namely the Davilla, the stem twines indifferently from left to right or from right to left; and I once saw a shoot, ascending a tree about five inches in diameter, reverse its course in the same manner as so frequently occurs with Loasa. 16 Although individuals, as we have just seen, in some few cases twine in opposite directions, and you are able to give only two cases of species within the same natural order thus twining.17 But a Mikania growing here twines from right to left, whilst the Mikania scandens described by you twines in an opposite direction;18 and I believe that there are species of Dioscorea which twine in opposite directions. Lastly, with respect to the thickness of the support which can be ascended by spirally twining plants, I have lately seen a trunk about five feet in circumference which was thus ascended by a plant apparently belonging to the Menispermaceæ.19


The text is composed of parts of three letters edited by CD for publication in the Journal of the Linnean Society (Botany) (see letters to Fritz Müller, 17 October [1865] and 9 December [1865]). The texts of the original three letters cannot be distinguished in this version; the dates are established by the relationship between these letters and the letters to Fritz Müller, 17 October [1865] and 9 December [1865]. See letter from Fritz Müller, 12 August 1865, n. 11.
In ‘Climbing plants’, pp. 110–11, CD had concluded that although all tendrils performed identical functions, they consisted of modified leaves, or modified flower-peduncles, or possibly modified branches and stipules; he also questioned whether the difference between foliar and axial organs could be ‘of so fundamental a nature’ as was generally supposed (ibid., pp. 113–14; for CD’s discussions on the derivation of tendrils, see Correspondence vol. 12, letters from Daniel Oliver, [28 January – 8 February 1864], 12 March 1864, and [before 31 March 1864], letter to Daniel Oliver, 11 March [1864], and letters to J. D. Hooker, [27 January 1864], [8 February 1864], and 3 November [1864]). Müller may be referring to CD’s remark on page 112 of ‘Climbing plants’, that he knew ‘nothing of any transitional states’, if such occurred, between branches and tendrils; however, CD gave as an example Lophospermum, in which branches spontaneously revolve and are sensitive to contact, to suggest that such modifications were possible; see also ibid., p. 10. In Climbing plants 2d ed., p. 196, he added that such transitional states had ‘been fully described by Fritz Müller’. CD also revised his description of tendril formation. In ‘Climbing plants’, p. 48, he wrote, ‘True tendrils are formed by the modification of leaves with their petioles, of flower-peduncles, perhaps also of branches and stipules’; in the second edition the sentence ends, ‘… flower-peduncles, branches, and perhaps stipules’ (see Climbing plants 2d ed., p. 84 and n.).
Müller refers to the description of Strychnos in Stephan Ladislaus Endlicher’s Genera plantarum (Endlicher 1836–42, p. 575), ‘ex axilla ramulem simplicem, cirrhiformem exserente’ (putting forth a simple, tendril-like branch from the axil). CD mentioned Strychnos in Climbing plants 2d ed., pp. 84–5.
CD mentioned Caulotretus in Climbing plants 2d ed., pp. 84–5 n.
Müller included sketches of the plants he described with his letter of 31 August 1865 (see letter to J. D. Hooker, 3 October [1865] and n. 3). When CD informed Müller that he would submit Müller’s observations on climbing plants to the Linnean Society he wrote, ‘As the Soc. is rather poor I fear that it will give only a few wood cuts from your truly admirable sketches’ (letter to Fritz Müller, 17 October [1865]). The printed version included one woodcut (Plate IX) of Dalbergia and Securidaca, which is reproduced above. The original sketches have not been found.
In ‘Climbing plants’, p. 48, CD had given the following definition: By tendrils I mean filamentary organs, sensitive to contact and used exclusively for climbing. By this definition, spines or hooks and rootlets, all of which are used for climbing, are excluded.
The correct spelling is Hecastophyllum; these plants are now included in the genus Dalbergia.
CD reproduced this list of stages in the development of branch climbers, with slight modification, in Climbing plants 2d ed., pp. 84–5 n. Müller’s list, tracing the gradation from branch to tendril, echoes CD’s description of the possible stages of development from leaf to tendril and from flower-peduncle to tendril in ‘Climbing plants’, pp. 111–12.
Müller refers to the description in Endlicher 1836–42, p. 1091, of Hippocratea, ‘Arbores v. frutices, ramis contortis scandentes’ (trees or shrubs climbing with twisted branches).
Müller refers to the description in Endlicher 1836–42, p. 1102, of Helinus, ‘ramulorum apicibus cirrhosis scandens’ (climbing by the tendril-like (curled) tips of branchlets).
In ‘Climbing plants’, pp. 57–9, 103–4, and 113, CD discussed the development of adhesive discs in Bignonia capreolata. For more on CD’s work on this genus and on others with adhesive discs, see Correspondence vol. 12, letters to J. D. Hooker, 4 December [1864] and n. 13, and 10 December [1864] and n. 6. CD’s notes on the development of adhesive discs in Bignonia capreolata are in DAR 157.1: 140.
CD added Müller’s observation on the discs in Haplolophium to Climbing plants 2d ed., p. 102 n., and amended his discussion of adherent discs on page 179 to take Müller’s observation into account.
CD’s observations on Cardiospermum halicacabum are in ‘Climbing plants’, pp. 87–9, 92, 112, and 113. CD’s notes on the species are in DAR 157.2: 68.
CD added Müller’s comment about spiral contraction in the peduncle of Serjania to Climbing plants 2d ed., p. 152 n.
Hibbertia dentata is described in ‘Climbing plants’, pp. 14, 16, 20–1, and 116. CD’s notes on the species are in DAR 157.1: 31, 49, and 57.
Müller’s observations on the twining of Davilla are given in Climbing plants 2d ed., p. 36 n.
In ‘Climbing plants’, pp. 18 and 20, CD had described the reversal in the direction of twining of Loasa aurantiaca and Scyphanthus elegans. His notes on these species are in DAR 157.1: 49 and 54; other notes on Loasa are in DAR 157.1: 55–6.
CD had noted in ‘Climbing plants’, p. 18, that Mikania scandens (climbing hempvine) moved ‘against the sun’; in Climbing plants 2d ed., pp. 33–4, he cited Müller’s remark on the Brazilian species of Mikania twining in the opposite direction. CD’s notes on Mikania are in DAR 157.1: 2 and 48.
The final sentence was evidently a late addition by CD to the printed version of the paper, which was published in November 1866, from information contained in a later letter from Müller (see Correspondence vol. 14, letter from Fritz Müller, 13 February 1866). In ‘Climbing plants’, p. 21, CD had noted: ‘Our English twiners, as far as I have seen, never twine round trees, excepting the Honeysuckle’, and on page 22 he remarked: ‘The tropical twiners, on the other hand, can ascend thick trees.’ CD had previously discussed the topic with Joseph Dalton Hooker (see Correspondence vol. 12, letter from J. D. Hooker, 2 December 1864 and n. 10, and letter to J. D. Hooker, 4 December [1864]). In Climbing plants 2d ed., p. 37 n., he added Müller’s observation on the thickness of the trunk around which a twining plant might climb.


Climbing plants 2d ed.: The movements and habits of climbing plants. 2d edition. By Charles Darwin. London: John Murray. 1875.

‘Climbing plants’: On the movements and habits of climbing plants. By Charles Darwin. [Read 2 February 1865.] Journal of the Linnean Society (Botany) 9 (1867): 1–118.

Correspondence: The correspondence of Charles Darwin. Edited by Frederick Burkhardt et al. 29 vols to date. Cambridge: Cambridge University Press. 1985–.

Endlicher, Stephan Ladislaus. 1836–42. Genera plantarum secundum ordines naturales disposita. With 4 supplements; in 2 vols. Vienna: Friedrich Beck.


FM’s comments on Climbing Plants.

Letter details

Letter no.
Johann Friedrich Theodor (Fritz) Müller
Charles Robert Darwin
Source of text
Notes on some of the climbing-plants near Desterro, in South Brazil. By Herr Fritz Müller, in a letter to C. Darwin. [Read 7 December 1865.] Journal of the Linnean Society (Botany) 9 (1866): 344–9.

Please cite as

Darwin Correspondence Project, “Letter no. 4881F,” accessed on 12 September 2023,

Also published in The Correspondence of Charles Darwin, vol. 13