ON CHORISIS. 375 



^ach is capable of assuming any of the functions, for we have mon- 

 ftrous examples (and I quote none hut what I have seen) of carpels 

 occurring among the exterior parts of a half-transformed' bud, petals 

 and imperfectly-formed stamens being found within : of stamens with 

 anthers present having stigmas at their tips and imperfect ovaries at 

 their lower portion | of petals and stamens passing by all degrees into 

 '€ach other and of all the circles returning to leaves. Besides these 

 there are well-known intermediate conditions such as used to be called 

 nectaries, and besides the expanded or unfolded condition of an organ, 

 tubular., hoodedj and spur or horn-like enlargements are not unfrequent= 

 ly met with. The leading effects of varying development may, in ad- 

 dition to what has been already pointed out, be conveniently noticed 

 under the following heads, connection or separation of parts ; equality 

 et inequality of the parts of a circle, and influences on the number of 

 parts. As to the first of these, it is a law of vegetable structure, that 

 portions of growing plants, whether of the same, or of closely allied 

 kinds, being in contact and continuing so, for a time without agitation, 

 will form tissue so as to unite and become as one. This law prevails 

 in the parts of flowers as elsewhere. The result is coherence when or= 

 gang of the same circle unite by their edges, adherence when organs of 

 adjoining circles unite by their surfaces. Increased development of 

 the parts promotes coherence ; closeness of the circles promotes adher- 

 ence, and differences in these particulars have much to do with the 

 ';variations of the common plan of flowers. 



We need not, however, be in any doubt as to the true explanation of 

 ^hat occurs, as we are familiar with cases of degrees of coherence from 

 the slight attachment of the petals of a Flax or Woodsorrel to the 

 qomplete union of these parts in a Convolvulus or an Erica, from the 

 .connection of the petals at the base only in some cases, to its reaching 

 the very tip in others, and we may have seen a little starvation restore 

 a Bellflower or Convolvulus to five separate petals. 



It is necessary, to be able to express what happens in precise and 

 accurate language, and as the terms monosepalous, monopetalous, af- 

 firm what is well known not to be true, and are fitted to obscure the 

 ideas of students, whilst DeCandolle's terms, gamosepalous, gamope- 

 talons, are figurative and too long, and have met with little accept- 

 ance, I take this opportunity of proposing terms long used by me, 

 as a teacher, which seem fully to supply what is needed without being 

 liable to objection. Let the coherent parts be called synsepaloust 



