CnAP. IV. VARIOUS OBJECTIONS. log 



stamens are ruilimcntury in Ononis columnas ; and in some spe- 

 (•i(^s of Viola three stamens arc in this state, two retaining their 

 proper fimction, but being' of very small size. In six out of 

 lliirtv of tlie closed flowers in an Indian violet (name unknown, 

 for the plants have not as yet jiroduced perfect flowers), the 

 sepals were reduced from the normal lunnber of live to three. 

 In one section of the Malpighiacea3 the closed flowers, according 

 to A. de Jussieu, are still further modified, for the five stamens 

 which stand opposite to tlie sepals are all aborted, a sixth 

 stamen standhig opposite to a j)etal being alone developed; 

 and this stamen is not present in the ordinary flowers of these 

 species ; the style is aborted ; and the ovaria are reduced froni 

 three to two. In all the foregoing plants the minute closed 

 flowers are of high service, for they yield with perfect seciu-ity, 

 !Uid with the expenditure of extremely little pollen, or other 

 organized matter, a large supply of seed ; while the perfect 

 flowers permit occasional crosses with distinct individuals. 

 Therefore, these changes may have been, and no doubt have 

 been, efTi.'cted through natural selection ; and I may add that 

 nearly all the gradations between the perfect and imperfect 

 flowers may sometimes be observed on the same plant. 



With respect to modifications which necessarily follow from 

 other cljangcs — through the withdrawal or excess of nutriment 

 — througli pressiu-e and other unknown influences — there is 

 space here only for a few brief illustrations. In the Spanish 

 chestnut, and in certain fir-trees, the angles of divergence of 

 the leaves differ, according to Schacht, in the nearly horizontal 

 and in the upright branches. In the common rue and some 

 other plants, one flower, usually the central or terminal one, 

 ojicns first, and has five sepals and petals, and five divisions to 

 the ovarium ; while all the other flowers on the plant are te- 

 tramerous. In the British Adoxa the uppermost flower gener- 

 ally has two cjilyx-lobes with the other organs tetramerous, 

 while the surrounding flowers generally have three cal^'x-lobes 

 with the other organs pentamerous ; and this difference appears 

 to follow from the manner in which the flowers are closely 

 ]> icked together. In many Compositic and Umbellifera?, and 

 in some other plants, the circumferential flowers have their 

 corollas much more developed tlian those of the centre; and 

 this is probably the result of natural selection, for all the 

 flowers are thus rendered nuich more conspicuous to those in- 

 sects which are useful or even neccs.sary for their fertilization. 

 In connection with the greater development of the corolla, the 



