560 MR. J. PARKIN ON THE 
(3) In the first place, two main classes of flower-clusters appear capable of 
being distinguished, which may be named the apical and intercalary. The 
majority of inflorescences belong to the apical class, which comprises the two 
types long recognised by botanists, viz., the eymose and racemose, An inter- 
ealary inflorescence is characterised by the fact that the main axis, after 
emitting a number of flowers laterally, continues its apical growth vegeta- 
tively ; thus the flower-cluster is inserted on the axis between two sets of 
foliage leaves. 
(4) Reasons are adduced for regarding these two main classes of inflor- 
escences as not being strictly comparable to each other, since their respective 
evolutions commence differently. An apical inflorescence arises from a 
single axis bearing a solitary terminal flower. An interealary inflorescence 
requires two or more of such axes for its evolution. An apical inflorescence 
is produced by the pushing out of new flowers alongside the terminal one. 
An interealary inflorescence, on the other hand, results when a number of 
adjacent lateral axes bearing terminal flowers lose their foliage and shorten 
their internodes. The one is a new production: the other a segregation. 
Regarding apical inflorescence :— 
(5) The first flower-cluster to arise from the solitary terminal flower is 
eymose in character. This is formed by two or more of the leaf-axils below 
the terminal flower pushing out lateral floral shoots, each of which usually 
produces a pair only of leaf-struetures (braeteoles), and then ends in a 
flower. If only two of these shoots be emitted—perhaps the commonest 
first step in the evolution of inflorescences—a dichasium (a group of three 
flowers) will be formed. If three, then a trichasium: and if more, a 
pleiochasium. 
(6) From the beginning two main tendencies are observable in flower- 
clusters arising in this way, viz. (1) the number of secondary (lateral) floral 
axes tend to increase, (2) the branching tends to be carried further, so that 
tertiary and higher axes are foreshadowed. These two tendencies are to a 
large extent opposed to each other. 
(7) The first tendency, when followed, leads eventually to the formation 
of true racemose inflorescences by the following steps :— 
(ij) A considerable increase in the number of secondary floral shoots 
results in the terminal flower no longer blooming first, the flowers 
generally inclining to an acropetalous order of opening. 
(ii.) Next, the uppermost part of the inflorescence is arrested in its 
growth, and its flower-buds do not expand. 
(iii.) This part finally never develops, so that the terminal flower dis- 
appears. The inflorescence may now end in a mere filament or 
a slight protuberance. 
