6 AN INTRODUCTION 



which is the upper part, and usually spreading. A 

 polypetalous Corolla is cruciform, cross-shaped, 

 when it consists of four Petals that are equal and 

 spreading; and papilionaceous, butterfly-shaped, when 

 it is irregular, consisting of four Petals, of which the 

 under one resembles the keel of a ship, the upper 

 one rises, and the two side ones stand single. 



There belongs also to the Corolla a part called the 

 Nectarium, which has been but newly distinguished, 

 having been by former botanists confounded with the 

 Petals. It is by Linnaeus defined to be the part which 

 bears the honey, and belonging to the flower only. 

 This part affords a wonderful variety in the manner 

 of its appearance. In some plants it is very large, 

 as in the Narcissus and Aquilegia; in the former of 

 which the cup, and in the latter the horns, are Necta- 

 rla: in others it is scarcely discoverable, even with 

 glasses. In some plants it is united with, and makes 

 part of the Petals : in others it is detached from them. 

 Its shape and situation are also as various. Its use 

 is not known, unless the supposition of its secreting 

 the honey may be depended npon- 



Between the Calyx and Corolla nature has put no 

 absolute limits; as is plain from the Daphne, in 

 which plant they grow together, and are united in the 

 margin, like a leaf of the Buxus ; but they may be 

 commonly distinguished by their position in respect 

 of the stamina, the petals and stamina being ranged 

 alternately ; whereas the segments of the Calyx, and 

 the stamina, answer to each other. That this is 

 their natural situation, appears from the complete 

 flowers in the classes Tetrandria and Pentandria: 

 And the use of applying this rule will be found in 

 the instances of Chenopodium, Urtica, and Parieta- 

 ria ; where it decides, that the single cover in those 

 Genera is a perianthium, and that it is the Corolla 

 that is wanting. Should we infer, where only one 

 of the two covers appears, that it is a Corolla, be- 

 cause that is a more principal part, there would be 

 i^n rertaintv from such an inference ; as is evident- 



