Szekessy Vilmos (szerk.): A Magyar Természettudományi Múzeum évkönyve 59. (Budapest 1967)

Radics, F.: A revision of the Nymphaea material in Hungarian Natural History Museum

indicated by the slightly confused venation, the considerable difference between the thickness of the petiole and the peduncle, and the longer innermost stamens. Accordingly, we may name this plant, following GLTNNARSSON'S nomenclatoral concept, N. alba x Candida x tetragona f. subalba (Plate I, Fig. 6). On this Nymphaea specimen, the most conspicuous ones are the aZ6a-features, and since the plants under discussion are not recent hybrids, they could not have been named anything else owing to the phytosystematical knowledge of that time, but relegated, of all known species, to N. alba. SIMONKAI himself indentified the respective specimens as Castalia alba LINK, as given on the label of the sheet. Finally, the N. Candida specimens, or the specimens regarded hitherto as one of its varieties, of the Departmental herbarium should also be examined. Of the two waterlilies collected by F. HAZSLINSZKY and identified as N. biradiata (Nos. 58061 and 58079), the first originates from the Szenna marshes (Carpathian Ukraine), the second from the environs of Csap (Csop). None of them carries the date of the year, only an indication of the month of collecting (July and August respectively). Both plants agree with SOMMEEAUER/S description of N. biradiata (23) insofar as the floating leaves are alongate, green above but marginally violet reddish below, while the lamella is divided by a completely straight midrib. On the other hand, the lamellar lobes are not strikingly divergent, indeed, the interior margins of many sinuses overlap, and none have a slightly impressed apex. SOMMERAUER'S diagnosis fits also the petals, the violet reddish colour of the stigmatic area, and the tridentate termination of the base of each style. The torus is covered with the staminal stumps even higher than in alba. However, the number of carpels and the carpellary styles respectively is not 5—10 but considerably more. HASZLINSZKY must surely have observed, on the fresh plants or in the habitat, the agreement of even more features, while probably failing to consider the differences as essential ones. Aside of the manifestly candida-features (wider innermost staminal filaments, obtusely pointed, weakly arcuate and distally overlapping lobes, convergent lobal primaries, leaves green on the upperside and not paler in characteristical areas, prominent venation on the underside, etc.), the lower lobal pair of primaries are divergent on one or the other leaf, not all are plate in the astomatic area and on the primaries, the number of pistillar filaments is higher, the torus is covered with stamens up to the axile process, the margins are reddish brown on the underside of the leaves, — all definitely aZ&a-characteristics, aside of this intermixture of traits being concurrently also a proof of their hybrid origin. And more than that, the quadrangular torus (about four times wider than the thickness of the peduncle), the small lines appearing on the backside of the sepals in exsiccates, the long anthers of the innermost stamens, etc., testify to a tetragona strain. On the basis of the above considerations, these allegedly biradiata plants are again N. alba X Candida X tetragona f. subcandida. The dominancy of the cawt^a-traits is shown also by the shape of the pollen (Plate II, Fig. 7). The Nymphaea specimen (No. 58119), collected near Szeged by THAISZ and identified as Castalia Candida (PRESL) SOHINZET THELL, refers, by its paler astomatic area and primaries and the reddish brown hue at the base of the backside of the sepals, definitely to alba. On the other hand, the staminal filaments wider than the anthers, the convergent lower (lobal) primaries, and the obtusely attentuating apex of the lobes are Candida characters. And the heteromorphous lobes, the small-sized astomatic area, the shape of the stamens, the thin petioles and peduncle, the fre-