328 MORPHOLOGY. 



foliar organs of the flower are equally various in their duration. 

 Perianth, corolla, and accessory corolla commonly perish soon after 

 the perfecting of the flower ; either they are cast off by true dis- 

 articulation, or they wither upon the parent plant. The epicalyx 

 and calyx frequently share the fate of the axial organs supporting 

 the rudiments of the fruit ; the carpels almost invariably. The 

 carpels are rarely destroyed before the perfecting of the seed, as 

 in Leontice and, according to Robert Brown, in Peliosanthes Tlieta. 

 The stamens die away mostly immediately after the dispersion of 

 the pollen ; either they are cast off, or they dry up and die away 

 within the flower. 



The terminology in use is as follows : those parts which fail 

 away immediately, when their perfect formation is but scarcely 

 completed, are termed caducous or fugacious (paries caducce); 

 those which endure somewhat longer are termed, if they are cast 

 off by disarticulation, deciduous (partes deciduce) ; if they retain 

 their position and die by gradual withering and drying up, mar- 

 cescent (partes marccscentes) ; those parts which remain long, still 

 vegetating, are termed persistent (partes persistentes) ; if they 

 change their forms by further growth, they are termed excrescent 

 ( partes excrescentes). 



In what has just been said there are three points to which I must draw 

 especial attention, because they have most important influence on the 

 mode of observation of the entire flower. They are by no means new 

 facts, but their importance has not hitherto been properly recognised. 



a. The first relates to the position of the parts of the flower. I will 

 not enter here into the very acute theories of Schimper, but confine 

 myself singly to the true observation of Nature. She gives us two dis- 

 tinctly separate conditions ; namely, the origin of the foliar organs of 

 the individual members in closed circles, in which all the individual 

 parts appear simultaneously and at the same elevation upon the axis.* 

 Where this condition obtains, the individual parts of the several circles 

 alternate with each other, without one established exception that I know 

 of; and where this does not occur in the perfect flower, an inter- 

 mediate circle has always failed to develope, or parts have been esteemed 

 independent which are not so in fact, as in Potamogeton, where the 

 stamens are said to be opposite to the leaves of the perianth ; but the 

 so-styled perigonial leaves are only crest-like expansions of the con- 

 nectives of the anthers, and not independent foliar organs. Probably 

 exactly the same condition occurs in the Proteacece, an account of the 

 development of which I have been unable to obtain. But I must observe 

 that many of my Investigations do not lead in the same direction. I 

 may not pass over here the condition that in few- (two-) membered circles, 

 as in the Tkymelacece, every two-and-two circles approach together, and 

 alternate in pairs, although observation proves that here originally four- 

 membered circles are by no means present. All these plants belong to 

 those which in descriptive botany are spoken of as definite parts (partes 

 definite) ; and it is easy, by means of the equal number of members of a 



* Examples of this are furnished by all Lih'acece, IridetE, Palma, Grasses with 3- 

 merous ; the Lnbiatce, BoraginecE, Composite, Campanulacece, with 5-merous ; many 

 ScrophularinitB with 4-merous ; the Berberidete with 3-merous ; and the Thymeleacece 

 with 2-merous circles. 



