STRUCTURE OF FLOWERS. 107 



folded or rolled backwards, as seems to happen in the 

 petals of some Umbelliferae. 



4th. The parts of a verticil may be disposed in an 

 exact circle as to their position, but each of them 

 slightly twisted upon its axis, so that by one of its sides 

 it overlaps one of the neighbouring ones; and its other 

 side, being a little more interior, is overlapped by the 

 other one which is next to it. This disposition, which 

 is called Twisted or Contorted Estivation, (PI. 19, 

 figs. 2p, 4p, 5p,) is rare in envelopes which have their 

 pieces perfectly free ; it is seen in the petals and sepals 

 of the Flax, in the petals of the Pink and of the Mal- 

 vaceae ; but it is much more frequent in the free 

 portions or lobes of gamopetalous corollas, as in the 

 Apocyneae and Rubiaceas. 



When the parts of a regular verticil are in two or 

 more rows, i. e. when the same verticil is double, &c, it 

 may also present several cases : — 



1st. If the parts, placed exactly in the same direc- 

 tion as regards the axis, be alternate with one another, 

 there results the Alternate ^Estivation, (PI. 19, fig. 

 14,) where the pieces of the second row exactly alter- 

 nate with those of the first, and those of the third with 

 those of the second, &c. ; this is seen in the tepals of 

 the Liliaceae, the petals of the Nymphaeaceae, &c. Each 

 of these rows may itself present one of the preceding 

 dispositions ; but as the pieces are more distant, we are 

 rarely able to recognise it with precision. 



2d. Under the name of Imbricate ^Estivation, 

 (PI. 19, figs. 9, 11,) are generally confounded all the 

 cases where the envelopes, being in several rows, have 

 no determined order, and where the pieces overlap each 

 other as the tiles on a roof. We see this in the involucra 

 of most Compositae, in the petals of most double flowers ; 

 but it is probable that we confound here, in the same 



