the identity of these plants; such as the erect branchlets, 
which in our plant are remarkably spreading, the smoothness 
of the leaves on both sides, in ours rugose underneath, the 
setaceous stipules, which in ours, though narrow, are toothed ; 
the solitary peduncles, which in ours grow more frequently 
two, sometimes three and even four together. It is however 
possible that they may be the same, for the country of which 
AmyepALus pumila is a native, is by no means certain ; 
Prrsoon attributes it to China, and the African origin given it 
by Linnaus, is doubtful ; being derived solely from Herman, 
who apparently speaks only from memory; for although he 
says he found it growing luxuriantly at the Cape, he acknow- 
ledges that he introduced it into the Dutch gardens, from the 
English, in which, on his return to Europe, he found it was 
not uncommon. Herman’s figure does not represent our 
plant so well as Piuxener’s, as he has represented the 
flowers to be sessile and solitary. 
There is only one specimen of AMy@paAtus pumila in the | 
Banksian Herbarium ; this agrees in every respect with our 
og and was taken from Mr. Lxe’s nursery in the year 
Our drawing was made from a specimen communicated by 
Mr. Ler of Hammersmith in May 1819; and we have been 
this year favoured by him with the ripe fruit, which very 
much resembles a small red Cherry, and abounds with an 
acid juice, not unpleasant to the taste; the stone or nut 
however is rugose, like that of the Peach ; and quite different 
from that of the Cherry or Plum. And therefore whilst this 
character is to decide the genus, it must be referred to 
Amygdalus and not to Prunus. It may nevertheless be con- 
sidered as a connecting link between the two, if any limits 
can be found to separate the genera. : 
It is remarkable in this species that there are more generally 
two ovaries or germens in each flower, and now and then 
three ; in consequence of which there are sometimes two 
drupes upon one pedunéle, but usually one of them aborts ; 
leaving however sufficient rudiments to shew the previous 
existence of two. | 
