EVOLUTION OF THE INFLORESCENCE. 515 
Cesulia), in which the florets of the capitulum expanded in the reverse way, 
commencing to open from the centre outwards iustead of from the periphery 
inwards, as is usual for this family. He deduced from this fact that the 
capitulum of these genera is in reality a compound one, composed of a number 
of single-flowered heads—a view now generally held. These inflorescences 
are referred to later in this paper, as being among the most evolved. 
Roeper *, eight years after, revived this view of Robert Brown and, partly 
from a consideration of it, divided inflorescences into two groups, according 
to the way in which the individual flowers successively expand. He named 
the two methods the centripetal and centrifugal respectively. The division 
of inflorescences into these two main groups has been generally accepted by 
botanists ever since. The terms racemose and cymose or indefinite and 
definite, are now more generally used than centripetal and centrifugal 
respectively. Though each pair of the foregoing terms signifies a different 
feature concerning the inflorescence ; yet, as a rule, in racemose (indefinite) 
inflorescences the flowers open centripetally (acropetally) and in cymose 
(definite) inflorescences centrifugally (basipetally). 
Guillard, in 1857, published a highly “ philosophic " paper entitled ** Idée 
générale de l’Inflorescence " +. He first calls attention to the great neglect 
of the study of the Inflorescence. For three quarters of a century he writes 
that there has been no advance on the ideas of Linnzus. This statement, 
though partly warranted, is a slight exaggeration in the light of the con- 
tributions of Robert Brown and Roeper just considered. Grillard’s paper 
is so full of fine terms and so-called laws that he seems to cloud rather than 
illuminate the subject. His “lois de progression et de recurrence” ap- 
parently fit the modes of development of racemose and cy mose inflorescences 
respectively. 
Eichler f, as already pointed out, in his great work * Bliithendiagramme,’ 
treats of the Inflorescence in a truly scientific manner. He shows that a 
sharp distinction between racemose and cymose inflorescences does not exist 
in nature, calling attention to the fact that some flower-clusters are a 
mixture of both. He also definitely states that the one type of inflorescence 
may pass into the other, but refrains from expressing which he considers the 
more primitive. 
Celakovsky §, in a paper published in 1893, discusses the nature of the 
panicle, showing that in its manner of flower-expansion, it can be partly 
centrifugal and partly centripetal. He distinguishes fourteen main forms 
* Roeper, J., Mél. bot. de Seringe, No. 5, p. 71, 1826. 
t Guillard, A., Bull. Soc. Bot. France, iv. (1857) pp. 29-41. 
t Eichler, * Bliithendiagramme,’ pp. 33 & 34, 1875. 
$ Celakovsky, L. J., * Gedanken über eine zeitgemasse Reform der Theorie der Bluten- 
stande," Bot. Jahrb. f. System (Engler), xvi. (1898) p. 33. 
