68 OSTENFELD. 



and to the morphology of the plant, but apart from these errors, which 

 are but natural in such early botanical work, there is no doubt as to the 

 identity of the Philippine plant with Gaudichaud's type; moreover 

 Gaudichaud's figure is excellent. Below is given a diagnosis of the 

 species, as far as the material allows, the male flowers being unknown 

 to me. 



Halophila ovata Gaudichaud in Freycinet's Voy. Bot. (1826) 430, pi. J t 0, f. 1., 

 non H. ovata auctt., nee H. ovalis (R. Br.) Hook. f. a ovata J. B. Balfour in Trans. 

 Bot. Soc. Edinburgh 13 (1878) 335. 



Gracilis; caulis horizontaliter repens; internodia gracilia, 1-3 cm 

 longa; rami erecti nulli; nodi foliis duobus, squamis duabus radiceque una 

 instructi ; f oliorum lamina late ovato-elliptica, obtusa, circiter 0-10 mm 

 longa et 4-5 mm lata, glabra, integerrima, trinervia; nervi secundarii 

 3-7 inter nervum medium et nervos submarginales suspensi; foliorum 

 petioli laminis equilongi vel paullo longiores. Plores dioeci ( ?), solitarii, 

 spatha diphylla membranacea suffulti; masculus (sec. Gaudichaud) 

 podunculatus; stamina 3, unilocularia ( ?) ; femineus sessilis, perianthium 

 minimum, triphyllum; ovarium ovoideum, longe rostratum, circiter 2 mm 

 longum; styli 3, filiformes, circiter 6 mm longi; fructus maturus mem- 

 branaceus, globosus, 2-2.5 mm diametro; semina subglobosa, pallida, 

 0.5 mm. 



Species H. ovali affinis, praecipue differt omnibus partibus multo minor 

 graciliorque, nervis secundariis 3-7, nee pluribus (in H. ovali 12-25), 

 staminibus unilocularibus (?). 



Besides its small size, the above species is most easily distinguished by the 

 arrangement of the secondary nerves. In H. ovalis the number varies from 12 

 to 25, being most often 17 or 18, forming a very acute angle with the median 

 nerve, and the area limited by two secondary nerves is much longer than broad. 

 In H. ovata the number of secondary nerves varies from 3 to 7, the most common 

 numbers being from 4 to C, and they form nearly a right angle with the median 

 nerve, while the area between two secondary nerves is nearly as long as broad, 

 that is, rhomboidal. 



In Gaudichaud's figure the leaves are drawn with emarginate tips, but this 

 is incorrect, as I have found upon examination of the type material. In other 

 respects his figure gives an excellent idea of the habit of the plant. In habit it 

 is quite different from H. ovalis, and much more like H. baillonii Aschers. and 

 H. decipiens Ostf., but nevertheless the other characters point to H. ovalis as the 

 most closely allied species. As mentioned above, I have not found any male 

 flowers in the material at hand, and it would be very interesting to have them, 

 as Gaudichaud's description of them seems to be incorrect in some important 

 points. 



H. ovata Gaudich., has been found only in the following places: Marianne 

 Islands, Gaudichaud, specimens in Mus. Hist. Nat. Paris; and the Philippines, 

 Luzon, Loher 1595, May, 1892, in Herb. Kew.; Manila Bay, Merrill ^112, April, 

 1905. According to Mr. Merrill the species is very abundant in shallow water 

 along the shore near Manila in April and May, but disappears later in the season. 



