THE ANGIOSPERMAE 



1 143 



difference between an involucre surrounding a single flower and one sur- 

 rounding a condensed inflorescence, such as that of the Compositae or of 

 Euphorbia, which exhibits the gestalt or pattern of a flower. There is no 

 need to suppose that involucrate single flowers are reduced capitula (Cf. 

 Scabiosa, p. 1 145). The two cases may be quite independent from an 

 evolutionary point of view, though from the aspect of gestalt morphology 

 they are equivalent. 



Fig. 1 1 19. — Poeonia delavayi. Leafy involucre. 



Involucres are very varied in appearance. In Dianthiis we have a simple 

 involucre with four bracts, in two decussate pairs, set very close to the 

 flower and resembling a calyx. The Ranunculaceae display all types of 

 involucre, from the spiral series of bracts in Paeonia illustrated in Fig. 1 1 19, 

 and the wide, leafy involucre of Eranthis, looking like a green collar below 

 the flower, to the loose, leafy involucres of Anemone, which are usually 

 separated by some distance from the perianth. The latter genus is particu- 

 larly interesting because of the series of stages to be observed between 

 species like A. nemorosa, in which there is a whorl of three palmatifid bracts, 

 each resembling a foliage leaf (Fig. 1 120), which stands about half-way down 

 the flowering stem, and A. hepatica, in which there is also a whorl of three 

 involucral bracts, but so reduced in size and so close to the flower that it is 

 difficuh to withhold from them the name of sepals. Series like this support 

 the view that the sepals are in fact involucral bracts which have become a 

 permanent part of the flower. 



The flowers of Mirabilis (Nyctaginaceae) are arranged in cymes of three, 

 of which only the middle one develops. Around its base stands an involucre 



D* 



