THE FLOKAL STEM. 739 



rachis and evident flower-stalks; the spike (spica) with elongated rachis and 

 extremely reduced flower-stalks; the umbel (umbella) with an extremely reduced 

 rachis and elongated flower-stalks; and the capitulum (capitulum) with a very 

 short thick rachis and exceedingly reduced flower-stalks. All these inflorescences 

 are connected together by intermediate forms, of which the corymb (corymbus) — 

 especially characteristic of Cruciferse — forming a link between the umbel and the 

 raceme, deserves special mention. The capitulum exhibits the greatest variety, 

 but this is produced less by the difl'erent forms of the floral stem than by the 

 shape of the floral leaves, especially of the numerous crowded bracts which 

 collectively surround the flowers as a cup-like envelope. A form of spike with 

 thickened rachis, called a spadix (spadix), is also worthy of note, and also the 

 spike known by the name of catkin {am^entum), the flowers of which are devoid 

 of perianth-leaves, and spring from the axils of scale-like bracts. The whole 

 catkin falls oflT after flowering, or after the ripening of the fruit, a separation of 

 the tissue and a detachment of the cells having previously occurred at the base 

 of the rachis. 



When spikes are themselves arranged in a spicate manner, the whole 

 inflorescence is called a compound spike (spica composita); racemes grouped 

 into larger racemes form a com,pound racerne (racem,us compositus); and umbels 

 when arranged in larger umbels form a corapound umbel (umbella compiosita). 

 The first two occur very often in grasses, the last in umbelliferous plants. The 

 term panicle is also often applied — rather loosely — to any compound raceme. 



Various combinations of the above simple inflorescences have been distinguished, 

 particularly combinations of centripetal with centrifugal inflorescences. Capitula 

 and compound umbels which are arrangeu in cymes, and cymes which succeed 

 one another in a spicate or racemose manner are of very common occurrence. In 

 these inflorescences the order of blossoming becomes altered. Of the many umbels 

 which are grouped together in an extensive cyme, the central umbel is the first 

 of the series, but it is the flowers on its periphery and not the central flowers 

 which open first. If cymes are arranged like a spike, the lowest, i.e. those on the 

 periphery of the whole inflorescence blossom first, though in each individual cyme 

 the central flowers are always the first to open. 



The order of blossoming, which is determinate for the flowers of every given 

 species, is related to the transmission of the flower-dust or pollen to the stigma, and 

 therefore with the processes of fertilization. When in one and the same flower 

 the organs in which the pollen and those in which the ovules are developed stand 

 closely side by side, it might be thought that the pollen would be certain to reach 

 the adjoining stigma. But this opinion is not confirmed by experience. It has 

 been demonstrated, on the other hand, that it is of advantage to the plant that 

 the pollen of one flower should reach the stigma of another, indeed of the flower 

 of quite another plant often some distance away; thus we find that cross-fertili- 

 zation is aimed at, at any rate at the commencement of the flowering period. 

 I purposely say "aimed at", and avoid saying that crossing of difi-erent plants 



