ON CHORTSIS. 377 



of very iincommon occurrence, in which two flowers are combined into 

 one from their origin, owing to their buds having been adjacent. I can 

 now distinctly recall examples in two or three species of Iris, and in at 

 least three species of Oenothera, my cultivatioQ at one period of numerous 

 species of those genera affording me the opportunity of observing the 

 anomalies to which they are liable. 1 had various instances of circles 

 of five in the monster Iris and of seven in the Oenothera — one instance 

 of four in the Iris in a single circle and one of only three, the exterior 

 circles having five, and the tube showing sufficient marks of the union. 

 In the Oenotheras observed, which embraced several species, there were 

 uniformly seven parts in each circle, that is, seven sepals, seven petals,, 

 fourteen stamens and seven carpels. I gave some account of these 

 monstrosities to the Linnsean Society in 1 839, and it has since occurred 

 to me that they establish a law respecting the combination of circles 

 of growing parts, which may explain the tendency to the number five 

 in Dicotyledonous plants, since, when growth is carried on from a single 

 cotyledon, we find the number three in the circles, and where there 

 are two cotyledons we might expect the circle to be double, but the 

 fact of the loss of at least one part in combinations of two circles on 

 the same plane shows why the number five takes the place of six. 

 The liability of the natural numbers, five in Dicotyledonous and three 

 in Monocotyledonous, to be reduced by mere pressure or by irregu= 

 larity, is obvious from what has been already said. We find by obser- 

 vation that the number of parts in the successive circles of the flower 

 is usually equal, but that the inner circle, being exposed to greater 

 pressure, is apt to have fewer than the others — three and two carpels 

 being very common in Dicotyledonous plants. In some structures 

 the numbers in the different circles do not at all correspond, but thiSj 

 which is characteristic of particular families, is less common, and its 

 origin is one of the most obscure and dubious points in the theory of 

 the flower. When parts are absent either from pressure or irregia= 

 larity, we must remember that the fact is due to a special cause of 

 abortion, not to the total absence of the part from the structure, and 

 consequently that circumstances may occur from more abundant or 

 equally distributed nourishment, which may in anomalous examples 

 restore the missing part. Such examples are, indeed, almost needed to 

 confirm our judgment as to the causes of the ordinary absence of these 

 parts, and have therefore great interest for the philosophical botanist;. 

 In the natural family of the Onagraceze, to which the genera Fuchsia 



