|
|
Oxalis obtusa Jacq. Oxal. (1794) 106, t. 79, f. 1 (A Group species) (Salter #33).
O. cuprea Lodd.
O. piottae Colla*
O. cruentata Jacq.
O. thermarum E. & Z.
O. lacunosa E. & Z.
O. ciliariflora E. & Z.
O. fimbriata Phillips
O. pearsonii R. Knuth (non Bolus f.)
O. cuneiformis Salter & Exell
O. framesii L. Bolus
O. membranacea Weintroub
O. fugax Schltr. Ex R. Knuth
O. schultzii R. Knuth
O. rondeboschensis R. Knuth
O. obtusa Jacq. Vars. subvillosa Sond., glabrata Sond., hantamensis
R. Knuth
A rather lax plant, usually 5-15cm high.
Bulb: oval with attenuate points at the base and apex, often fusiform, with hard blackish-brown tunics deeply pitted and irregularly angled, glabrous or rarely sparsely pilose with simple and short capitate hairs.
Rhizome: short, rarely up to 10cm long, producing bulbils freely in the axils of the scales.
Stem: either not exserted, forming a short stipe or, particularly in shade and mountain forms, elongated to 10cm or more long, leafless, slender, always pilose with soft reversed hairs.
Leaves: 10-20, rarely up to 60, basal or congested at the apex of the stem, or very rarely imbicating on a short stipe : petioles 2-4cm long or longer in shade forms, articulated to a scale like and sometimes pseudo-stipulate base, pilose with soft reversed hairs, or very rarely with patent, ascending or short glandular-capitate hairs : leaflets 3 (or in Var. atrata, 5), shortly petiolulate, 0.5-2.5cm long, obcordate, cuneate-obcordate or oblong-obcordate, sometimes triangular in outline, bilobed to the middle, often heterophyllous, the early leaflets obcordate, the later elongating to narrow-cuneate or linear emarginate, varying from closely silky-pubescent on both faces to entirely glabrous, more or less impresso-punctate when dried, variously callus-dotted or ecallose, occasionally purplish beneath and rarely purple-mottled above.
Peduncles: about twice as long as the leaves, softly pilose with reversed hairs, rarely with short capitate hairs admixed : bracts 2, opposite, at an articulation above the middle.
Sepals: oblong or oblong-lanceolate, obtuse or subacute, 1/3 to 1/4 the length of the corolla, softly pillose with ascending hairs, always with some reversed hairs at the base, ciliate, ecallose or very rarely with apical calli.
Corolla: 1.5-2.5cm usually about 2cm long, pink, brick-red or palish yellow : tube short, more or less widely or rarely narrowly funnel shaped, yellow : petals subcuneate, with variable attenuation into the short claw, obliquely subtruncate at the apex, glabrous or very rarely pilose near the outer nargin, often purple veined beneath.
Filaments: often glandular-pubescent, the longer either edentate, gibbose or even with teeth over 1mm long, the longest (or styles) well exserted from the corolla tube.
Ovary: pubescent on the upper part, ecallose, the chambers many-ovuled : styles usually with simple and short capitate hairs admixed.
Capsule: oblong, not or scarcely exserted beyond the calyx. Seeds endospermous. Sometimes producing double flowers.
Flowers: June-October.
Form B.
Flowers much larger, white, pale rose or yellow. Leaflets always heterophyllous.
Flowers: September
Var. atrata (Weintroub) Salter O. atrata Weintroub.
Stemless.
Leaflets: 5, heterophyllous, obcordate, cuneate, oblong or linear up to 1.6cm long. Corolla pale yellow, veined, 1.1-2cm long with a widely funnel-shaped tube.
Filaments: the longer 8.5mm long, well exserted from the corolla tube, with rather acute teeth about 0.5mm long.
In some specimens the leaflets are all obcordate. The variety only differs from Namaqualand specimens in having 5-foliate leaves.
This species is perhaps the commonest and most widespread of the whole genus in South Africa. The characters by which it may be distinguished are as follows:
(i) Bulb irregularly pitted and sharply angled
(ii) Longest stamens (or styles) well exserted from the corolla.
(iii) Ovary without calli
Almost invariably:
(iv) Reversed hairs on the stem, petioles and peduncles and at the base of the sepals.
(v) The presence of flavone in the petals, affected to a greater or less degree by anthocyanin staining.
And there is a marked, though not invariable tendency to heterophylly the early leaflets being obcordate or cuneate-obcordate, the later lengthening to narrow cuneate, oblong or linear. Some forms are certainly not heterophyllous.
In other respects it assumes a bewildering number of forms in which the variable characters (vide revised description) are combined in so many different ways, though never apparently in constant groups, that, at the present juncture, it seems inadvisable to attempt to create an infinite number of varieties. It is, indeed, doubtful whether extended exploration and intensive study of many thousands of plants, both in the field and in cultivation, will ever lead to a workable taxonomic classification and certainly not until more detailed field observation of the heterophylly of the species has been made. Both large and also dwarf forms occur and the fact that these retain their approximate size in cultivation shows that these variations are not epharmonic. In some colonies the production of a stem appears to be confined to shade specimens, but other colonies, with the plants all caulescent, may be found in the open.
The typical form, that illustrated by Jacquin (tab. 79), is the commonest in the extreme south west, the corolla being usually brick-red. Chance specimens gathered in outlying ocalities would often be unlikely, on their own evidence, to be recognised as belonging to the same species, especially in the dried state, and there is no doubt that they would be definite variations from the type. Moderate extensive collecting and observation, has, however, shown that almost every such gathering is a new form. In the past, on the variations over a limited quantity of herbarium material (sometimes without bulbs) a number of species and varieties have been created and, with the exception of the 5-foliate variety atrata, in none of these, when weighed in conjunction with the mass of intermediates, can I find characters which are sufficiently individual to warrant taxonomic separation.
The names of the various species amalgamated here will be found in the list of synonyms given above. No useful purpose would be served, nor does space permit of pointing out their separate characters, for they merge into one another and are no more worthy of separate notice than all the other variants. Some of them owe their origin to the unawareness of their authors of heterophylly in the species, for it is only in the description of O. framesii L. Bolus, that there is any reference to it.
O. framesii, with very large pale rose or white flowers, without any trace of flavone n the petals, is perhaps worthy of special notice. The discovery of a form of this, exactly similar in all respects except that the flowers are pale sulphur-yellow, shows that it would be inconsistent to give this form varietal status. O. rondeboschensis R. Knuth, though the colour of the corolla in the type is not now discernible, is near this form and it has evidently been so named through an accidental interchanging of Schlecters ticket in Berlin Herbarium. It cannot have come from the suburbs of Cape Town.
Some forms found in northern Namaqualand flower in June-July, considerably earlier than the normal flowering season, viz. Aug-Oct. These forms have very thick leaflets and the appear in other characters to be extreme variants. Owing, however, to the discovery of a number of intermediates with other forms it is not considered feasible to separate them. Others from the same region are characterised by the presence of capitate hairs, particularly on the peduncles, but a marked tendency has been noticed for such hairs to become more numerous and spread to the whole plant in cultivation and they are not accompanied by any other exceptional character.
This species is one of the few in which I have found the teeth on the longer filaments to be variable and unreliable as a character. Many forms are entirely without teeth, while others vary from gibbose swellings to teeth over 1mm long, a fact which seems to suggest that extensive mutation is taking place.
* Artificial hybridisation between O. piottae and O. obtusa is recorded by Hildebrand, but they are probably conspecific.