266 Mr. Griffith on the Ambrosinia ciliata of Roxburgh. 



fruit being unilocular and the placenta central, attached only by its base and 

 apex. 



In his MS. Synopsis Roxburgh says of the seeds of C. ciliata: — "The seed 

 generally vegetates in the capsule, and is as completely polycotyledonous as 

 any Pinus, or even Domheya, the Norfolk Island Pine, itself." In his 'Flora 

 Indica*' he describes the embryo as erect, furnished with a perisperm, and 

 many subulate cotyledons as in Pinus. 



I subjoin the character of the genus taken from the ' Meletemata Bo- 

 tanica' : — 



" Spatha tubo brevissimo, limbo elongato. Spadix spatbee plica tubum 

 claudente (appendicula) conjunctus. Antherse confertse, loculis amplis cel- 

 lulaeformibus, marginatis, septo vald^ distincto separatis ; poro (?) dehiscen- 

 tibus. Ovarium pluri- (6) loculare, ovulis diversa altitudine axi affixis, peri- 

 tropis. Styli plures (?),; stigmata radiata. Seniina albuminosa, testa spon- 

 giosa (?). Embryo cotyledonibus (protophyllis ?) pluribus. — Asiaticse; rhizo- 

 mate stolonifero ; foliis vagina petiolari dilatata, lamina Integra uninervi ; 

 floribus suaveolentibus." 



M. Schott has referred Caladium ovatum of Ventenat, Karin-Pola of 

 Rheede's ' Hortus Malabaricus,' vol. xi. p. 45. t. 23, to this genus. The struc- 

 ture of the fruit as described and figured by Rheede appears however to be 

 somewhat difFerentf. With this, Arum aquaticum of Rumph's 'Herbarium 

 Amboinense,' vol. v. p. 312. t. 108, has no apparent affinity. 



The roots are cellular and vascular in the centre, cellular towards the cir- 

 cumference ; the intermediate part being occupied by a number of cavities 

 (containing apriform fluid), the walls of which are formed of a single series of 

 cells, and which radiate from the centre. This structure seems not uncom- 

 mon among Monocotyledonece. The petioles and the membraniform sheaths 

 are arranged alternately, but corresponding in direction with each other, the 



* Vol. iii. p. 492. 



t la Rheede's plant there would seem to be an increased number of ovaria. From some degree of 

 subdivision existing in the tissue of the spadix between and within the apices of the ovaria in C. ciliata, 

 and in a higher degree in C. spiralis, as represented by Dr. Wallich's eirtists, I shall not be surprised 

 if species be found to exist with ovaria disposed in two or more series ; in such case their direction will 

 most probably become more horizontal. 



