156 cruger's organographical observations 



the vegetable kingdom. Among these are the irregularities of 

 the parts of the flower, which have occupied the attention of 

 morphologists for some length of time. To these must be added 

 the high degree of development of the leaves, which in the Mu- 

 saceae and Scitamineae are separated from the petiole by a very 

 evident, swollen articulation, and which goes so far in most of 

 the Orchideae that the leaf becomes detached from its stalk at 

 this place. If, with Lindley, we exclude the Dictyogens as a 

 transitional class, this structure recurs in the Endogens only in 

 the Aroideae, — in some of which the leaf falls off its stalk when 

 old, while others present real compound leaves, and in the 

 Grasses, which approach not a little to the Scitamineae in this 

 particular in such genera as Pharus, and in which the petiole 

 and lamina are most evidently distinct. These forms of leaf, it 

 may be remarked in passing, prove how mistaken those bota- 

 nists were who ascribed di folium petiolaneum to the Endogens 

 generally. 



In most of the plants of the three families above named, with 

 which I am acquainted, it is seen at once that in those genera 

 in which the stem produces branches, the inflorescence is very 

 compound, and on the other hand, that where branches are only 

 produced on a subterraneous stem, the inflorescence likewise 

 remains pretty simple. To what extent this may be generalized 

 is a question I must leave to the decision of those who have 

 access to a large number of species, in collections, &c. In the 

 following pages will be found, where requisite, a short descrip- 

 tion of the parts of the flower of certain plants of these groups, 

 together with a fragmentary contribution to the history of their 

 development. 



Musaceae. Heliconia Bihai, Sw. — The accompanying diagram 

 (PI. V. fig. I) illustrates the relative position of the parts of 

 this flower. I must remark, that the generic character, as I find 

 it given by Endlicher and others, does not agree accurately with 

 the species of Heliconia with which I am acquainted here, and 

 we possess five in Trinidad. The tw^o lateral, external, and the 

 three inner petals are more or less coherent, the outer petal next 

 the axis is free, and the stamen in front of this is abortive. 



The inflorescence of this species is composed of a number of 

 abbreviated flowering branches which are attached to the rachis 



