236 THE FLOWER. 



stamens, &c. of a pentamerous flower are each a depressed spiral or 

 cycle of the f mode of phyllotaxis, and those of the trimerous flower 

 are similar spirals of the ^ mode. But then the parts of the suc- 

 cessive cycles should be superposed, or jsLiced directly before each 

 otlier on the depressed axis, as leaves are ; whereas, on the contrary, 

 they almost always alternate with each other in the flower. 



442. To reconcile this alternation with the laws of phyllotaxis in 

 alternate leaves, Prof. Adrien de Jussieu has advanced an ingenious 

 hypothesis. He assumes the f"^ spiral arrangement as the basis of 

 the floral structure both of the trimerous and pentamerous flower, 

 (at least when the envelopes are imbricated in the bud,) tliis being 

 the one that brings the successive parts most nearly into alternation, 

 either in tlirees or in fives ; as will readily be observed on inspection 

 of the tabular projection of that mode, given on page 139. The dif- 

 ference between the position of parts in regular alternation, wliether 

 in threes or fives, and that assigned by an accurate spiral projection 

 of the -jSj mode, is very slight as respects most of the organs, and in 

 none does the deviation exceed one thirteenth of the circumference ; 

 — a quantity which becomes nearly insignificant on an axis so small 

 as that of most flowers. Moreover, if the interior organs of a regular 

 and sj-mmetrical flower were thus to originate in the bud nearly in 

 alternation Avith those that precede them, they would almost necessa- 

 rily be crowded a little, as they develop, into the position of least pres- 

 sure, and thus fall into these intervals with all the exactness that is 

 actually found in nature. For in living bodies, endowed as they are 

 with plasticity and a certain power of adaptation to circumstances, 

 the positions assumed are not mathematically accurate ; and the 

 effect of unequal pressure in the bud in throwing the smaller parts 

 more or less out of their normal position may be observed in almost 

 any irregular flower. Moreover, in all the forms of phyllotaxis 

 from y^j onwards, it is doubtful whether what we term vertical ranks 

 are exactly superposed. In tracing them upward to some extent, 

 we perceive indications of a curviserial arrangement, where the 

 superposition is continually approximated, but is never exactly at- 

 tained (248). Lestibudois* has revived the older hypothesis of 

 Jussieu, and others ; viz. that a second spiral is introduced with the 

 petals and continued in the pistils. And Schimper and Braun im- 

 agine a change of half the angular divergence (prosenthesis) to occur 



* lu Annales des Scietices Naturdles, ser. 4, Vol. 2, p. 226. 



