THE HERBARIUM AMBOINENSE. 379 
to be that of the Bilacus minimus or Kitsiil; but no such 
species is mentioned in the text : and the third kind de- 
scribed there is the Bilacus Pissang, which, as Humphius 
says, " nomen obtinuit ab oblonga forma instar Musse Pi- 
sangh dictae but the fruit C is shaped hke a pear, and 
cannot represent such a Bilacus, as is described. These 
three fruits, I think, represent three varieties in form of 
the Bilacus tellor, which may indeed be called maocimus, 
ovalis, and minimus ; but these are merely such varieties 
as occur in all cultivated plants ; while the Bilacus tauriiius, 
if it be represented in plate 82, and Bilacus Pissang-, are 
probably distinct species. On this subject, however, I must, 
for a farther account, refer the reader to the Commentary 
on the Bilacus taurinus. 
The Bilacus tellor and its varieties were early noticed by 
botanists under the name Marmelos, corrupted from the 
Marmeleira of the Portuguese, given to it because this 
people seem to have prepared a marmalade from its fruit ; 
on which account the older botanists compared it to the 
Cydonia, a tree resembling it in no other respect. In the 
Hortus Malabaricus (iii. 57, t. 37) it was described under 
the name Covalam ; but the commentator, Commeline, 
still adhered to the slight resemblance with the Cydonia. 
Plukenet seems to have been sensible of the absurdity in 
. this comparison ; but he was little more fortunate in calling 
it " Cucurbitifera trifolia, spinosa, Indica, fructus pulpa 
Cydonii semula," (Aim. 125; Phyt. t. 170, f 5). Now 
his Cucurbitiferas include Crescentia, Strychnos, and other 
plants equally dissimilar. The elder Burman (Thes. Zeyl. 
84) continued to class it with the Cydonia, but gives the 
synonyma of preceding writers with sufficient accuracy. 
In the Flora Zeylonica, Linnaeus was equally unfortunate 
with his predecessors, and united this plant with the Tapia 
of Margrave, which he supposed to be the Niirvala of the 
