ABSTRACTS OF FOUR LECTURES TO STUDENTS. 



519 



green ; hence it is regarded as the fundamental colour of all flowers, and 

 to which they often " revert," as do Chrysanthemums. 



Carpels. — These are obviously formed by a leaf folded upon itself or 

 " conduplicate " : that is, like a piece of note-paper, but with the margins 

 coherent. The ovules arise on these two coherent and greatly thickened 

 margins, called the "placentas." There is usually one row at least on 

 each margin, but they may be increased to many or reduced to one ovule. 



The origin of ovules appears to be homologous with foliaceous out- 

 growths which often arise from the ribs and veins of Cabbage leaves, or 

 as " crests " on some corollas of Begonias, Gloxinias, Primulas, and 

 Daffodils. For, when a leaf will form a carpel, the fibro vascular cord 

 entering the base divides into three branches. One corresponds with the 

 midrib of the leaf, the other two pass up the now enlarged margins and 

 send off shoots into the ovules. 



In monstrous conditions of flowers in which the carpels revert more 

 or less to leaves, then the ovules may become like minute leaflets &c, as 

 has occurred in Mignonette, Clover, and many other cases.* 



Hence every part or "appendage" of a flower is homologous with a 

 " foliar " structure. 



Even when the ovule appears to rise from the base of the ovary, as in 

 the Composite, the cord which supplies the ovule arises laterally, showing 

 that it is really marginal and not axile.f 



Number. — The number of parts of the whorls of flowers are usually 

 fives, fours, and threes, and occasionally twos (Enchanter's Nightshade), 

 or twice these numbers. 



Fives are peculiarly characteristic of Dicotyledons. This is due to the 

 fact that the arrangement of leaves in Dicotyledons is usually represented 

 by the § plan, five forming one " cycle " along the spiral line. 



Similarly ^ and J plans are characteristic of the foliage of Monocoty- 

 ledons, and these most easily give rise to cycles of three, as in the 

 flowers. 



The | plan is easily obtained from opposite leaves (derived from the 

 two opposite cotyledons) while the others are equally referable to the 

 single cotyledon in Monocotyledons. % When, however, cycles become 

 whorls, the individual members of the latter are so placed as to fall over 

 the intervals between those of the next whorl below. Hence is recognised 

 the " law of alternation " in flowers. 



If exceptions cccur, as when the five stamens of the Plumbaginece, 

 Rhamnus, and Primula stand in front of the petals, it signifies that a 

 second outer whorl of stamens is suppressed. Rudiments of filaments 

 occur in Brook-weed (Samolus), which thus reveal the ancestral numbers. 



In flowers where an indefinite number of parts occur, as of stamens 

 and carpels in Buttercups and Roses, then, as a rule, the spiral arrange- 

 ment is retained. Nature, however, has another way of multiplying the 

 parts of any whorl. This is done by a branching of the (say) original five 

 fibro-vascular cords, which would enter as many stamens ; but each branch 



* "Fasciation and Allied Phenomena," Journ. R. Hort. Soc. vol. xxvi. p. 155. 

 t Vascular System of Floral Organs &c.," Journ. Linn. Soc, Bot. vol. xxviii., pi. 

 29, fig. 6. 



X I have fully explained this in my papers on Phyllotaxis : Trans. Linn. Soc. vol. 

 xxvi. (1868), and vol. i. (Botany) (1875). 



