CONTRIBUTIONS TO BOTANY. 
231 
some distance from the hypogynous disk, which in Colletia 
assumes a perigynous form : in this respect we meet with only 
one exception, in Adolphia, where, owing to the extreme short- 
ness of the calycine tube, their point of attachment is drawn 
close to the disk ; but, although thus brought into proximity, 
they are not actually inserted upon its margin, as they are ge- 
nerally in the Rhamnece. In the tribe Gouaniece, so different in 
most other respects, the insertion of the petals and stamens is 
like that of the CoUetiece. The stamens are placed opposite to 
as many scale-like petals in all the Rhamnacece •, they are gene- 
rally concealed and hooded by them in the CoUetiece, and are 
inserted upon their claws, between the lobes of the calycine 
border ; the petals are, however, sometimes wanting, and in a 
single apetalous species the insertion of the stamens is below 
the middle of the tube. The CoUetiece, besides, are generally 
distinguishable at a glance by their ])eculiar habit : the branch - 
lets, half abortive, often assume the form of spines, which are 
either quite bare, or they are foliiferous, or else they become 
more elongated, bearing both leaves and flowers, but sometimes 
they are destitute of both, with a broom-like appearance ; the 
leaves, when present, are always small and opposite, or, by the 
approximation of the axils, almost fasciculated, and are either 
entire or denticulated, and sometimes are reduced to almost 
obsolete proportions. Most of the species at present known are 
natives of South x\merica ; there are a few in Mexico, one from 
the Galapagos, one from the island of Juan Fernandez, one from 
Australia and Tasmania, and another from New Zealand. 
There has hitherto been much confusion among the genera, the 
* characters of many of them being still undefined : thus Trevoa 
and Talguenea, which I proposed above thirty years ago, have 
been confounded together by most botanists. My observations, 
I hope, will enable me to define the characters and limits of 
these as well as the other genera of the tribe. 
When Jussieu published his ‘Genera Plantarum^ (in 1789), 
the only genus then known of the present tribe was Colletia ; it 
was there placed in his order Rhamni, a heterogeneous group 
composed of Rhamnacece, Celastracece, Aquifoliacece, Bruniacece, 
Staphijlacece, and some other genera. From this association 
Mr. Brown (in 1814) separated the Celastracece, including the 
Aquifoliacece, and (in 1818) the Bruniacece-, and thus circum- 
scribed, DeCandolle (in 1825) published an enumeration of the 
order, wherein Colletia appears as the only genus belonging to 
the tribe since constituted. We owe to Brongniart (in 1826) 
the best monogi’aph of the family that has yet been published, 
where its afiinities are well discussed, and where ample charac- 
ters of the genera are given : here he first separated Retanilla 
