1248 



FLOWER 



FLOWER 



are as usual inserted morphologically lower on the 

 receptacle. While in most flowers the ovary is 

 inserted on the summit of the receptacle {superior 

 ovary), in others, as in the Orchidaeese, Onagracese, 

 UmbeUiferse, Rubiaceae, and Compositae, the ovary 

 appears to oc»^:upy the center of 

 the club-shaped structure {inferior 

 ovary) below the insertion of the 

 calyx, corolla, and stamens which 

 seem to spring from the summit 

 of the ovary {epigynous). The 

 view has been held that in such 

 cases a gamosepalous calyx similar 

 to that described above in the 

 perigynous flower has grown fast 

 to the surface of the ovary, 

 and that the other organs are 

 borne on the calyx-tube at the sum- 

 mit of the ovary. The opinion is 

 now becoming general that the true 

 explanation of the phenomenon is 

 that the cup-shaped receptacle of 

 the perigynous flower, and not the 

 calyx, has grown fast to the surface 

 of the ovary. In the Onagracea; 

 and some other plants, the hollow 

 receptacle has not only grown fast 

 to the whole surface of the ovary 

 but projects beyond it so that such 

 flowers have an inferior ovary and 

 are also perigynous (Fig. 1530). 



Bracts. — The leaves on the peduncles and upper 

 parts of the stem adjacent to the flower deserve a word. 

 They are often much modified in size, shape and color 

 from the normal foliage leaves, being often much 

 reduced. They sometimes form an involucre around 

 the flower, and are calyx-like, as in hepatica and straw- 

 berry. In other cases, they form a showy coroUa-hke 

 involucre, as in Cornus and Poinsettia, and are then 

 often mistaken for a corolla. In the Arum, a single 

 huge bract (spalhe) envelopes the entire flower-cluster 

 (^spadix); these are weU shown in Figs. 1532, 1533. 



1530. The fuchsia 

 flower in longitu- 

 dinal section. 



1531. a, epigynous flower; b and c, perigynous flowers. 



Incomplete flowers. — Not all of the floral sets described 

 above are always present. The flowers may be incom- 

 plete. Thus the corolla may be wanting (flower apetal- 

 ous) as in hepatica and anemone, or both calyx and 

 coroUa may be absent {naked or achlamydeous) as in 

 willow and pepper, or the stamens may be wanting 

 {imperfect or unisexual, pistillate flower) as in willows 

 and oaks, or the pistils may be absent {stami-nate 

 flowers of willows and oaks) . At least one set of essen- 

 tial organs is necessary for a functional flower, but in 

 some cases, through specialization for other purposes, 

 both sets may be absent. Thus the marginal flowers of 

 the hydrangea are enlarged and showy for insect attrac- 

 tion, but are neutral. In the case of unisexTial flowers, 

 the stamens and pistils may be borne in different flowers 

 on the same plant {monoecious) as in the oak and birch, 

 or on separate plants {dicecious) as in the willow and 

 poplar. In some plants, as in the maple, certain 

 flowers are unisexual while others are perfect, a con- 

 dition termed polygamous. 



The plan of the flower. — If the numbers of parts in 

 each set are counted, a certain number will be found 

 to be common to many or all of the sets of the same 



flower. This is the numerical plan of the flower (Fig. 

 1534). Thus in geranium there are five sepals, five 

 petals, ten stamens, and five parts to the pistil. The 

 stamens, when numerous, are often in multiples of this 

 numerical plan. The parts of the pistil, on the other 

 hand, frequently show a reduction from the numerical 

 plan as exhibited by other parts of the flower. The 

 number of parts in some flowers is so irregular that a 



1532. The great white spathe (and the spadix) of the garden calla. 



numerical plan can be made out only with difficulty, 

 while in some flowers such a plan is apparently wanting. 

 The members of each floral set are usually inserted 

 all at the same height on the floral axis (receptacle), 

 and are therefore in whorls, although frequently more 

 than one whorl occurs in the androecium and rarely 

 in other sets. The parts of one set normally faU between 

 those of the set next outside and next inside, and are 

 said to alternate with these. In some families, as for 

 example in the Ranunculaceae and Magnohaceae, some 

 or all of the organs of the flower are inserted spirally 

 on the receptacle like scales on a pine cone. In such 

 cases there is often a marked intergrading between the 

 organs of the adjacent sets at the boundary Une. The 

 relative position of parts of the 

 flower may be graphically indicated 

 by means of a diagramatic cross- 

 sectional plan, called the floral dia- 

 gram (see Fig. 1534.). Information 

 in regard to the number and union 

 of parts may also be indicated by 

 so-called floral formute as follows: 



K 

 5 



C A G 



5 5+5 2 



In this formula, the letters from 

 left to right indicate calyx, corolla, 

 androecium, and gynoecium respec- 

 tively . The brackets over the letters 

 indicate a fusion of parts in the 

 same set, while the bracket under- 

 neath indicates a fusion of different 

 sets. The above flower would be 

 polysepalous with five sepals, gamo- 



1533. Spathe and 

 spadix of Jack-in- 

 the-pulpit. 



