198 OF THE INVOLUCUUM. 



This part is of an infinite variety of forms in differ- 

 ent ^e'wc'm, bcine^ cither simple or compound, divided 

 or undivided, regular or irregular. In some instances 

 it is permanent till the fruit is ripe, in others it falls 

 even before the flower is well expanded. 



Some genera have a double perianthium, as Malva^ 

 t. 671, or even a triple one, as Scabiosa, t. 1311. 



2. Involucrum, f. 143, Involucre of Professor Martyn ; 

 but I generally retain the Latin termination. This is 

 remote from the flower, and can scarcely be distin- 

 guished clearly from a Bractea. The term was first 

 adopted by Linnasus, at the suggestion of his friend 

 Artedi, in order to distinguish \he genera of umbellif- 

 erous plants, for which purpose the latter deemed the 

 part in question very important. But according to 

 the laws which Linnaeus had laid down, the parts of 

 the flower and fruit alone were to aflford generic char- 

 acters, and the most sound botanists have ever since 

 kept to this rule, with infinite advantage over less cor- 

 rect ones, however ready to derive ideas respecting 

 the natural habit, and secondary characters, of a ge- 

 nus, not only from the inflorescence and bracteas, but 

 even from the leaves, stipulas, or other parts. Lin- 

 n^us and Artedi, therefore, were obliged to consider 

 the involucra and involucella, the former accompany- 

 ing the general and the latter the partial nnibels, as a 

 sort of calyx, and the umbel altogether as one aggre- 

 gate flower, composed of florets united by a common 

 radiated receptacle. Consequently a cyme must be 

 considered in the same light ; nor are reasons wanting 

 in support of this hypothesis, which we shall consider 



