AFRICAN SPECIES OF AMOMUM. 291 



^ 



suppose^ because the samples which have been imported for centuries 

 have, all the same sensible properties, that therefore the species that 

 produce them do not vary in these respects. 



There is another risk to which the medical botanist is often exposed, 

 in his attempts to identify his imperfect specimens of drugs with the 

 imperfect descriptions of authors, which is, the too great confidence 

 placed in the absolute value of taste. I do not allude to its relative 

 value amongst different samples of drugs of one kind, but to the com- 

 parison of the tastes of these and of novel ones with other objects. 

 Thus I find the seeds of one species of Amomum (still unrecognized by 

 pharmacists and botanists) described as having the flavour of lemons ; 

 I am far from saying that this is not correct, but I have so often seen 

 two or more persons give such widely different comparisons for the 

 taste of a novel substance, that I should myself place little confidence 

 in one individual's opinion upon a subject of this kind. 



With regard to the correct reference of the detached leaves, flowers, 

 and fruits of the plants described below, to the different species to 

 which they belong, I am, as indicated above, wholly indebted to Dr. 

 Daniell. I have no reason to suppose any mistake to have occurred m 

 the ticketing or collecting ; but, as such often do occur, and as I have 

 made such myself in tropical jungles, even when most sedulously en- 

 deavouring to avoid the possibility of errors, I cannot attach implicit 

 faith to these. In almost all cases there is sufficient botanical evidence 

 that the flowers in spirits are certainly those of the specimens with 

 fruit and leaves attached, as in the case of A. Mdegueta, A, Granum- 

 Paradisic and -^. DanielU; but there is not such good evidence that 



the flowers of A. longucapum and A. cereum belong to the leaves and 

 fruits included in the descriptions of them ; I have no reason, however, 

 to doubt that they do. In these cases it is to be understood, that if 

 an error bas occurred, and fragments of two species have been described 

 under one specific name, that name should be continued to the plant 

 which bears the character suggested by the specific name, or, if the 

 name be of indefinite signification, it should be given to the flowering 

 specimen in preference, for in Amomum the flower affords the best 



specific character. 



"With regard to the specific names retained here, some would perhaps 



have preferred that the name A. Granum-Paradisi should be suppressed, 



or transferred to the A. Meleyuekt, which alone bears the seeds now 



