A Synopsis of Galearia Zoll. & Mor. (Pandaceae)

Abstract

In a recent paper (Forman in Kew Bull. 20: 309-321 (1966)) it was shown that Galearia Zoll. & Mor. together with Microdesmis Hook. f. are closely allied to Panda Pierre, and that they should be transferred from the Euphorbiaceae to the Pandaceae. Panda, whichis a monotypic genus, was previously considered to be isolated in a family of its own, without obvious close relatives. It is now apparent that the closest ally of Panda is, in fact, Galearia, especially the two species G. celebica Koord. and G. maingayi Hook. f., which together form the subgenus Orthopetalum Forman (see Kew Bull. 14: 3 1-317 (i960)). These species clearly resemble Panda in general appearance, e.g. cauliflorous racemiform inflorescences (sometimes terminal in G. maingayi), fruits with thick-walled, much pitted and channelled endocarps, and leaf nervation. Anatomical data supporting this conclusion was provided as an appendix to Forman in i.c. (1966) by C. R. Metcalfe and N. Parameswaran. A noteworthy feature of the leafy shoots of Galearia (and of Panda) is their organization which in certain respects resembles that of pinnate leaves. The number of leaves per shoot is limited and there are no buds visible in their axils, although there is always a bud present in the axil of the shoot itself. The insertion of the petiole on the shoot is oblique, with one stipule much higher than the other, and the leaves are strictly distichous. The shoot as a whole thus presents a distinct dorsiventral organization. The presence of a terminal inflorescence in most species of Galearia makes the nature of the leafy shoots perfectly clear. As discussed more fully in the previous paper (1.c., 1966) a transition can be traced through cauliflorous species of Galearia (subgenus Orthopetalum) to Panda, where the leafy shoots are very similar indeed to pinnate leaves. There is considerable diversity amongst the flowers and fruits of this small genus (see Figs. I & 2, pp. 154 & 156) and they both fall into two main types corresponding to the two subgenera, as indicated in the key below. Only six species are recognized in the following account and these are easily distinguished, especially when male flowers are present. The genus has for long been in a confused state taxonomically owing to the large number of names which have been in use, all referring to the variable species, G.fulva (Tul.) Miq. The placing of Galearia in the Pandaceae seems to be gaining general acceptance: e.g. Airy Shaw in Willis, Dict. Fl. P1. & Ferns, ed. 7 (1966); Takhtajan, Sistema i filogeniya: 260 (1966); & Cronquist, Evol. & Classif. Fl. PI.: 258 (1968). Grady Webster in Journ. Arn. Arb. 48: 313 (1967) also supports this conclusion, remarking that Galearia and Microdesmis have a wood-structure unlike that of other uniovulate Euphorbiaceae. The xylem anatomy of Pandaceae, including Galearia, is briefly reviewed by W. L. Stern in Amer. Journ. Bot. 54: 673 (1967). Further support linking together the genera of Pandaceae is provided by the seed-testa according to Vaughan & Rest in Kew Bull. 23: 215-8 (1969). On the other hand, Bakhuizen van den Brink in Backer, Fl. Java 3: 649 (1968) does not agree with removing