A NGIOSPERMS. DICOTl'LEDONS. 



455 



four whorls, constituting calyx, corolla, androecium and gynaeceum, and any 

 multiplication of the whorls is almost entirely confined to the androecium ; in 

 trimerous and dimerous flowers, the number of the whorls is much more variable, 

 and two or more whorls may be devoted to one series of organs. 



The corolla is not unfrequently wanting, and in that case the flower is said to be 

 apetalous. If the calyx and corolla are both present, they almost always have the 

 same number of parts (Papaver is an exception); but the number of their whorls varies; 

 the calyx for instance may consist of two dimerous decussate whorls, the corolla of one 

 whorl of four members, as in the Cruciferae. If the androecium and the perianth, 

 whether consisting of calyx only or of calyx and corolla, are present in the same flower, 

 they generally have the same number of parts, and the flowers are isostemonous, but 

 there may often be more, seldom fewer stamens than the parts of the perianth, and 

 then the flowers are amsosiemonous 1 . In flowers with pentamerous and tetramerous 

 whorls the number of the carpels is usually less than five or four ; in those with 

 dimerous or trimerous whorls or with the parts arranged in spirals there is often a 

 larger number of carpels. 



It will be seen from these few remarks that the conditions of number and 

 position in the flowers of Dicotyledons are very various, and cannot, as in most 

 Monocotyledons, be referred to a single type; even the attempt to assign different 

 types to as many larger groups is attended with considerable uncertainty, for in many 

 cases we do not possess the knowlege of development necessary for referring 

 particular floral formulae to more general ones; moreover the excessive application 

 of the spiral theory of phyllotaxis to cyclic flowers has hindered the true under- 

 standing of the latter, and raised doubts where without that theory there would have 

 been none. 



The formula Sn Pn Stn ( + n+ . .) Cn { m} may 

 be given for a large majority of Dicotyledons ; it is 

 true for most pentamerous and truly tetramerous 

 flowers and for octamerous flowers like Michauxia, so 

 that = 5 or 4 (or 8); in the androecium an indefinite 

 number of (alternating) whorls is assumed Stn 

 ( + +..) in order to include the large number of 

 flowers in which the androecium contains more than 

 one whorl (Fig. 392); the symbol Cn ( m) is meant to 

 indicate that there are often less than 5 or 4 (or 8) 

 carpels present, m meaning any value from o to n. In 

 the larger part of the Gamopetalae and in other species 

 also there are very often only two carpels, which are placed in the .median line one 

 posterior and one anterior; on the supposition that the gynaeceum has typically five 

 alternating members and has only become dimerous by abortion, if one is median and 

 anterior, the other must be obliquely posterior; a similar difficulty occurs sometimes in 

 the trimerous and monomerous gynaeceum. It would require too much space to explain 

 the reasons which determine many botanists to consider the formula given above as 

 valid also for the gynaeceum of such flowers as these ; we will only say that in the 



FIG. 392. Flor.il diagram of Aqiti- 

 legia (Ranunculaccae) 



1 See what was said above on page 418 on diplostemonous and obdiplostemonous flowers. 



