526 



PHANEROGAMS. 



and the diagram is therefore so far a theoretical one. If the dots are removed, we get 

 the empirical diagram ; the number and position of the carpels is here determined from 

 those of the stigmas ^ 



The flowers of Orchideae can also be derived, like those of Gramineae, from the type 

 represented in Fig. 375, the empirical diagram of Lihaceae, although their external form 

 is so remarkably different. While in Grasses the perianth is especially degenerated or 

 even partially abortive, in Orchids both whorls are developed in a petaloid, and like the 

 whole flower in a zygomorphic or monosymmetrical manner. Of the androecium, which 

 consists typically of two alternating whorls, each of three stamens, only a single stamen 

 is completely developed in most Orchids (Fig. 379, A), viz. the anterior one of the outer 



FIG. 379— Diagram of the flower of Orchideae : A the ordinary structure ; B that of Cypripedium {see Figs. 341 and 363) : 

 the dots indicate stamens wliich are altogether abortive, the shaded figures rudimentary stamens which become abor- 

 tive or transformed into staniinodes. 



whorl, the others being abortive. Indications of these are however sometimes found 

 in the young bud, as in Calanthe -veratrifolia (according to Payer, cf. Fig. 363), where at 

 least the two anterior ones of the inner whorl (but not the posterior one) appear as 

 small elevations which soon disappear. In Cypripedium, on the contrary, a large 

 staminodium (see Fig. 341, p. 479) takes the place of the anterior stamen which is else- 

 where fertile ; while the two anterior and lateral anthers of the inner whorl are fully 

 developed and fertile (Fig. 379 5). In Ophrydese two small staminodes are found beside 

 the gynostemium {cf. Fig. ^S7,D,st, p. 536) in the place of the two fertile stamens of 

 Cypripedium ; while in Uropedium even all three of the inner \yhorl are completely 

 developed. (Doll.) The carpels which, by adhesion with the androecium form the gyno- 



FIG. 380.— Diagram of the flower of Fumariacere (after Eichler). 



stemium, are developed unequally, a diflference which however is usually not discernible 

 in inferior ovaries, and is therefore not indicated in the diagram. The student who 

 desires to investigate these relationships for himself must observe that the long inferior 

 ovary of most Orchids undergoes a torsion at the time of the opening of the flower, 

 which causes the posterior side of the flower to assume an anterior position ; but trans- 

 verse sections even of advanced buds show clearly the true position of the parts of the 

 flower in relation to their mother-axis. 



The flowers of most Monocotyledons, like those of Orchids and Grasses, can be de- 

 rived from a type which is actually seen in Liliaceae, and which represents a flower 



' Compaie further, Diill, Beitrtige, in the Jahresbericht des Mannheimer Vereins fiir Naturkunde, 

 1870, where an actual pentacyclic trimerous flower of Streptochaete is described. 



