44 LAWS OF FLORAL STRUCTURE 



Law 4. — Each part of a circle develops separate and disconnected from 

 all others in that and in other circles. As the mass of tissue forming 

 each of the floral parts becomes isolated and projected from the torus, 

 its margins and faces should develop completely separate from those 

 of all adjacent parts. The law assumes that growth shall continue in 

 the isolated portions, by which process they must continue separate. 

 But this form of growth of the parts does not always occur. Very 

 commonly the point of growth changes or becomes restricted to the 

 basal portion, where they have not yet separated from one another. 

 This projection from the torus of an undivided or unseparated portion, 

 and its subsequent growth, must clearly result in the development of 

 a portion of the flower consisting of more than one floral part in union. 

 The component parts are usually indicated by more or less of a separa- 

 tion of their apical portions. This principle has been already carefully 

 explained in connection with Fig. 42. 



Connaiion. — There is no other direction in which deviation from the 

 type represented in Fig. 47 is so frequent and variable as in that just 

 described, nor in which the results are so far-reaching or call for so 

 extensive a classification and terminology. The deviations are of two 

 classes. When a part is united laterally with another part of the same 

 circle the condition is called Connation, Cohesion, Coalescence, or 

 Syngenesis. When connation does not exist the parts are said to be 

 Distinct or Eleutherous. Connation will be discussed in our detailed 

 consideration of the several floral parts. 



Adnation or Adhesion. — In the second form, called Adnation or Ad- 

 hesion, one circle is more or less united with another. Adhesion may 

 affect any two or more circles of the flower, and it may affect an entire 

 circle or only one or more of its parts. Thus, Fig. 54 illustrates a 

 petal of the ^'anilla adnate to the ganaecium, while the other petals 

 are free. It is plain that when the calyx and gynaecium are adnate, all 

 the intervening circles must be included in the condition, as in the 

 lower portion of the colocynth (Fig. 5(5). Since all the parts start 

 from the torus at a, they must be adnate to the whole surface of the 

 ovary between the points a and /;. 



Epigyny. — In the last case, as in all cases where one or more circles 

 are adnate to the gynaecium, the free or ununited ends of the parts 

 must lose the appearance of emanating from the torus and must appear 

 to emanate from the gynaecium. They are, therfore, said to be Epigy- 

 nous. At this ])oint the student should not fail to impress himself 

 with an understanding of the fact that in all such cases the epigynous 



