CONTRIBUTIONS TO BOTANY. 
175 
vaginant bractiform leaflets round each node, and a similar sti- 
puloid sheath round the base of each ramification. They are all 
of a dull reddish orange-colour. The male spikelets are ovate, 
2 lines long*. 
On the Tricuspidarie^, a Subtribe of the EUiEOCARPEAi. 
The Ekeocarpece^ as a natural order distinct from Tiliacece, 
was proposed in 1808 by Jussieu, who united with it the 
Tricuspidaria and Vallea of the ‘ Flora Peruviana.’ Kunth, 
in 1821, followed this example; but, in a note, he suggested 
that it might well form a distinct tribe of the Tiliacem. De 
Candolle, in 1821, adopted the view of Jussieu, adding to the 
list Friesia and others now' subgenera of Elceocarpus. Lindley, 
in 1836, in his ‘ Nat. Syst.,’ followed a similar course ; but in 
1845, in his ‘ Veget. Kingd.,’ he adopted the hint suggested 
by Kunth, uniting the family with Tiliacece as a distinct tribe. 
The authors of the new' ‘ Genera Plantarum,’ in 1862, followed 
this arrangement under some modifications, excluding Vallea 
upon very insufficient data, and amalgamating Friesia w'ith 
Aristotelia and Crinodendron with Tricuspidaria upon slender 
grounds. After a careful examination of these several genera, 
I am led to follow the views of Endlicher in maintaining the 
Tricuspidarioe as a subtribe distinct from Elceocarpeoe, which 
are distinguished from one another by very salient and con- 
stant characters. In the former the petals, though three-lobed 
at the apex or nearly entire, are never fringed as in the Eloeo- 
carpece ; in the latter the fruit is a drupe, with a single thick 
osseous endocarp, assuming the shape of an indehiscent tuber- 
culated nut, which, by abortion, is seldom more than 1- or 2- 
celled, each cell producing a single seed (not suspended from 
the summit, as generally stated, but) appended by the middle 
of its ventral face. On the other hand, the Tricuspidarur ^ 
besides the difference in the form of the petals, have a fruit 
always 3-5-celled, with two or more superposed seeds in 
each cell, and either capsular and dehiscent or else baccate 
with a membranous endocai'p. But a still more forcible dis- 
tinction exists in the nature of the integuments of the seeds. 
In the Elceocarpeoe the outer integument is chartaceous, thin, 
and brittle, the second tunic being submembranaceous ; but 
there is no osseous coat. In the Tricuspidariece the seeds in- 
variably have three tunics : the outer one is thick and fleshy, 
in which the chord of the raphe is imbedded ; the second coat 
is thick, osseous, obpyriform, truncated at its base, where, 
* A representation of this plant is given in Plate 79 b. 
2 A 
VOL. II. 
