The pattern of vascular bundles in the stamens of Nymphaea lotus L. and its bearing on stamen morphology

In 1969 I published, together with P. G. Heinsbroek, a paper on the anatomy of the stamens of Victoria amazonica. The flowers used in that study came from plants which were cultivated in the green-house of the Leiden Botanic Garden. Because the possibility could not be excluded that the structures then observed were partly the effect of greenhouse conditions, I subsequently took the opportunity to study flowers of well developed plants, which were cultivated in the open air under the tropical conditions of the Botanic Garden at Bogor*). The results were exactly the same. Apart from the set of central vascular bundles, normal in laminar structures, there proved to be a peripheral sheath of bundles consisting of abaxial bundles, terminating half way up the stamen, as well as adaxial bundles. All bundles run parallel and the central ones branch upwards into the fertile region of the stamen. The adaxial bundles anastomose at the lower end of the pollen sacs and one of the resulting anastomoses pursues its course in the middle between the thecae. In literature the last mentioned vascular bundle had been named the ‘auxiliary vein’ (Moseley, 1958). Its position is opposed to the normal median vein, and its xylem pole is inverted. This vein played a role in diverse morphological opinions on the flat stamens. American authors (Eames, 1961), who advocated the primitiveness of laminar stamen structure, disposed of the auxiliary vein by considering it as an insignificant vein. Schneider (1976) thinks the peripheral bundle system is explicable in functional terms. On the other hand, the discovery of the opposed median auxiliary vein was welcomed by authors like Leinfellner (1956), who thought, mainly on the ground of teratology, that the stamens are diplophyllous structures, that is consist of a dorsal and a ventral blade fused medianly. By this view the existence of apparently homogeneous laminar stamens in Ranales had been difficult to explain. However, now the auxiliary vein could be considered as the median vascular bundle of the fused ventral blade, as requested by the theory. Meeuse (1972) took up the suggestion brought forward in our paper of 1969, namely that the stamen vasculature consisted of a bract component (the abaxial bundles) and a flattened axis component (the central and adaxial bundles), in analogy with the Coniferous female cone scale. According to Meeuse this is proof of his thesis that the structure of the stamen in the Ranales is a bract amalgamated with an axial system. However, careful comparison with the results of Cécile Lemoine-Sebastian (a.o. 1972) show that the vascular patterns are different from a situation as described above.