104 Mr. J. Miers on the genus Liriosma. 



where its central placentation divides at its base into pseudo-dis- 

 sepiments, leaving the summit of the internal space always free 

 and unilocular, and the ovules suspended in that free space from 

 the common apex of these incomplete divisions. This structure 

 is decidedly marked in Liriosma, where the ovarium is 3-locular 

 at base, with one ovule in each half- cell, suspended from the in- 

 ternal angle of these incomplete divisions, and here the apices of 

 the ovules, rising above the points of suspension, are seen conni- 

 vent in the perfectly unilocular summit of the ovarium. Of these 

 ovules, as in the Olacacem and Santalace<2, only one is perfected 

 in the fniit, which becomes an oval crimson -coloured drupe, con- 

 taining a putamen covered by pulp and inclosing a single albu- 

 minous kernel that fills its cavity ; this exhibits externally a di- 

 stinct raphe-like thread, extending from the base to the summit, 

 as if the seed were suspended by a funicular support ; but it will 

 be seen that this thread partakes in no degree of 'the character 

 of a true raphe, but is merely the remains of the pseudo-disse- 

 piments, extended with the growth of the seed, and forced into 

 a groove formed by pressure along its sides, there being seen at 

 its summit a small cruciform extremity, resulting from the 

 abortive ovules, similar to the sti-ucture that forms so remarkable 

 a character of the Santalacece, as first pointed out by Mr. Robert 

 Brown. This same appearance is well defined, and a true expla- 

 nation of its origin is given by Mr. Bentham, in his excellent 

 memoir upon the Olacacea, in the 18th volume of the ' Linnsean 

 Transactions ' : he had observed the same structure in the genera 

 Olax, Heisteria and Schopfia, but its true nature had been mis- 

 taken by other botanists. The development of the seed in Liri- 

 osma is therefore identical with that of Schopfia, a genus referred 

 by Mr. Brown to Santalacea, but which Mr. Bentham first 

 placed among the Olacaceat with great reason ; and this proves 

 that the views of the former, in regard to the close affinity ex- 

 isting betwen these families, are founded upon truth, and I will 

 presently adduce other proofs of their validity. The stamens, 

 both fertile and sterile, in Liriosma, in form and structure 

 resemble those of Olax : and L. Gardneriana and the typical 

 species have their three fertile stamens placed in an alternate 

 position, between each second petal and their six sterile stamens, 

 intermediate between them, which are of course placed opposite 

 the six petals. With the exact number and position of the sta- 

 mens in the other species we are not informed, but I believe that 

 in this genus, as in Olax, the real normal number of petals is 

 six, with six fertile and six sterile stamens, the former being 

 always alternate with, the latter opposite to, the segments of the 

 corolla. 



The following is an amended generic character of Liriosma, 



