184 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



Hence those which one calls "hereditary structures" are only such 

 as have originated as varietal, but have become fixed in the constitution 

 of the being. 



It may be somewhat difficult to see this going on in nature, but it 

 may be abundantly witnessed under cultivation, and proved experi- 

 mentally ; but in nature the proof is not less convincing, though it be 

 based on inductive evidence alone. 



I will now consider the alterations in the structure of flowers brought 

 about by adaptation, i.e. evolution in detail. We may group them as 

 follows : — 



1. Freedom of the parts of whorls precedes their cohesion and adhesion. 



2. Freedom, associated with a spiral arrangement of an indefinite or 

 many parts in a floral whorl, precedes their reduction to a definite 

 number (1 to 12) which becomes verticillate in arrangement. 



3. An ordinary floral receptacle precedes a receptacular tube, or any 

 lateral expansion of the floral axis. 



4. The freedom of the receptacular tube from the ovary precedes its 

 adhesion to the latter. 



Obs. — Reversion may take place, in which an " inferior " ovary may 

 become " superior." 



5. Regularity of a whorl precedes irregularity. 



Obs. — Reversions to regularity may take place in ii regular flowers. 



6. Symmetry * precedes asymmetry by arrest of parts. 



7. All arrest of parts is of later occurrence in flowers and often the 

 cause of asymmetry. 



8. Enations are of later origin than other parts in the structure of 

 flowers. 



9. Unisexuality among normally bisexual]species is'a later production ; 

 though unisexuality was (as far as is known) the primary condition of 

 flowers. 



These " principles," if we may so call them, must now be considered 

 more in detail. 



1. Freedom. — Commencing with the class Dicotyledons, its orders 

 have long been arranged in four Divisions at least. The first is called 

 Thalamiflorce. In this the petals are free from one another, and arise 

 directly from the floral receptacle. The corolla and stamens are said to 

 be " hypogynous," as arising from under the ovary. 



Banunculacece has been placed the first in the series. This order 

 illustrates entire freedom among all the parts of the flower. One genus 

 only, Nigella, has its five carpels coherent into a capsular fruit. Some 

 have no corolla at all, as Clematis, Anemone, and Caltha, thereby 

 indicating a more primitive condition. Lastly there are usually many 

 stamens and carpels spirally arranged.t 



The question now arises, How did a corolla appear ? 



That petals arose out of stamens is obvious from numerous instances, 



* " Symmetry " in botany signifies that two or more whorls have the same or 

 multiples of the same number of parts. 



The term would be more useful in application to regular (radial symmetry) and 

 irregular liowers (bilateral symmetry). 



t The following are some other orders having the petals free and numerous parts 

 often spirally arranged : Nymplucacece, Magnoliacece, Rosacea, AlismacecB. 



