CONTRIBUTIONS TO BOTANY. 
177 
ble at this stage, the second being a little shorter and more 
jiointed than the first or outer one, the third integument, en- 
closing the rudimentaiy nucleus, being still shorter than the 
others. It is worthy of notice that at this stage the outer in- 
tegument exhibits tlie peculiar appendage or duplicature so 
conspicuous in the rijie seed. 
The Tricuspidariea' may be divided into two sections : the 
first, having a baccate fruit, mostly indehiscent, comprises the 
genera Aristotelia^ Friesia^ and Vallea ; the second, with a 
capsular fruit, consists of Tricuspidaria and Crinodendron ^ — 
in all, five genera. Having studied their structure mth much 
care, I proceed to describe each genus separately. 
1. Aristotelia. 
This genus, established by L’Hdritier upon the well known 
^laqui of Chile, was refen-ed by the late Mr. Brovni and De 
Candolle to the HomaUnerB ; by Endlicher it was made into a 
suborder attached to Ternstraemiacece ; Yon Martins also made 
it a suborder {Maqinnce)^ which he placed near PittosporecB ] 
Lindley considered this suborder allied to Philadelphiacece 
Reichenbach placed it in EscalloniecB ; but Don was the first 
who rightly indicated its affinity with the Elccocarpeoe, to 
which alliance it unquestionably belongs. The genus for 
many years Avas confined to its single typical species ; but 
Dr. Hooker, in his ‘ Flora Zelandica,’ associated with it the 
genus Friesia. There are so many points of structure in the 
Chilean plant at variance with the several species of Friesia^ 
that it a])pears to me this genus cannot remain amalgamated 
Avith Aristotelia, for the folio Aving reasons. 
In Aristotelia the petiole is always 2-glandidar at its apex ; 
and the teeth of the leaves have each a glandular termination. 
In Friesia no indication of any gland is seen either on the 
leaves or petioles. 
In Aristotelia the flowers are usually 5-merous, with a 3- 
celled ovary, and always heiTnaphrodite or else polygamous, 
with only a partial depauperation of the male or female organs. 
In Friesia the flowers are 4-merous, Avdth a 4-celled ovary ; 
and they are described as being dioecious in most instances. 
In Aristotelia the petals are not divided at the apex, being 
only slightly emarginated or truncated ; and the absence of 
this feature led botanists to believe, for a long time, that it 
could not belong to Elceocarpece ; they are white, with a slightly 
yellowish tinge. In Friesia, on the other hand, they are more, 
or less deeply 3-lobed at the apex, as in other Triciispidarieo', 
and are generally marked Avith a deep puqile colour. 
In Aristotelia there is a large, deep, cup-shaped disk fixed 
2a2 
