THE TONGA PLANT (EPIPREMNUM MIRABILE, SCHOTt). 333 



thickened upwards until it became about one inch in diameter, and 

 produced leaves, which, with the gradual thickening of the stem, by 

 degrees passed from their small entire state to a larger and more 

 and more divided form, until finally they became pinnatisect, and 

 including the petiole were from two to three feet in length. In 

 June, 1878, the plant flowered, and a specimen of it was brought 

 to me to name. I did not dissect it with any special care at the 

 time, but a rough dissection showed that there were but about two 

 basal ovules in each ovary, and therefore it would be either a 

 species of Epiprenmiini or of Monster a, most probably of the 

 former ; there was, however, no specimen in either genus in the 

 Kew Herbarium that would match with it, and it was not compared 

 with the genlis Bhaphidophora, since that genus, as at present under- 

 stood, has a more or less completely two-celled ovary, with 

 numerous ovules in each cell Had I compared it with that genus 

 I should doubtless have discovered its identity with RhapJiidophora 

 vitiensis, Schott, a typical specimen of which is in the Kew 

 Herbarium, and with R. pinnata, Schott, which is the same plant. 

 Though Eugler, in his monograph of the order (DC. Monog. 

 Phanerog. ii., p. 244), places R. pinnata as a synonym of Ft. pertusa, 

 Schott, with which, of course, it has nothing whatever to do, 

 whilst it. vitieyisis is made a variety of the same species. At the 

 time it was thought that Mr. Bull's plant might be Monstera 

 dilacerata, Koch, with which the leaves of a young stage of growth 

 agree well, even to the small pellucid dots and perforations 

 scattered along the region of the midrib ; but according to Koch's 

 description of that species in the ' Wochenschrift,' xiii., p. 88, it 

 cannot be the same plant. 



I have above indicated that the Tonga plant does not belong to 

 the genus Rhaphidophora as at present understood — that is, as it has 

 been understood and characterised by Schott, Engler, and others ; 

 I have also said that it is identical with R. pinnata, as has ah-eady 

 been declared by Mr. Bentham in the ' Flora Austrahensis,' vii., 

 p. 156 (and I do not doubt that Mr. Bentham is quite correct in 

 considering the Tropical Australian R. Cunninghami, Schott, to be 

 the same species, though complete specimens of this are still a 

 desideratum ; perhaps some Australian readers of this will kindly 

 take the hint by sending full and complete specimens to Kew). 

 The leaves, &c., of the two so-called sj^ecies are quite the same, 

 and on dissection I find the ovary of R. pinnata is just as in the 

 Tonga plant, being one-celled with two anatropous ovules, which are 

 seated one on either side at the base of a parietal placenta that 

 projects about one-third across the cell. Now this structure is 

 quite at variance with the characters assigned by Schott and by 

 Engler to the genus RhapJiidophora, but it is quite identical with 

 that of the genus Epipremnnni, Schott. Turning to this genus, 

 under E. niirabile we have quoted as synonj^ms, Scindajisiis decur- 

 sivus, Zollinger (No. 569), and Rhaphidophora lacera, Husskarl, in 

 part. Zollinger's specimens (^No. 569) I find I had placed (in the 

 Kew Herbarium), in the genus Rhaphidophora, having found them 

 to be identical with ii. pinnata, and thus it happened that when I 



