INTRODUCTION. 9 



former condition longer than the others; for when ir- 

 regular flowers become regular or peloric, they are apt 

 to be central; and such peloric flowers apparently owe 

 their origin either to arrested development that is, to 

 the preservation of an early stage of development or 

 to reversion. Central and perfectly developed flowers 

 in not a few plants in the normal condition (for in- 

 stance, the common Eue and Adoxa) differ slightly in 

 structure, as in the number of the parts, from the other 

 flowers on the same plant. All such cases seem con- 

 nected with the fact of the bud which stands at the 

 end of the shoot being better nourished than the others, 

 as it receives the most sap.* 



The cases hitherto mentioned relate to hermaphro- 

 dite species which bear differently constructed flowers; 

 but there are some plants that produce differently 

 formed seeds, of which Dr. Kuhn has given a list.f 

 With the Umbelliferse and Composite, the flowers that 

 produce these seeds likewise differ, and the differences 

 in the structure of the seeds are of a very important 

 nature. The causes which have led to differences in the 

 seeds on the same plant are not known; and it is very 

 doubtful whether they subserve any special end. 



We now come to our second Class, that of monoecious 

 species, or those which have their sexes separated but 

 borne on the same plant. The flowers necessarily 

 differ, but when those of one sex include rudiments 

 of the other sex, the difference between the two kinds 

 is usually not great. When the difference is great, 

 as we see in catkin-bearing plants, this depends 



* This whole subject, including mesti cation,' chap. xxvi. 2nd edit, 

 pelorism, has been discussed, and vol. ii. p. 338. 

 references given, in my ' Variation f ' Bot. Zeitung,' 1867, p. 67. 

 of Animals and Plants under Do- 

 3 



