TERMINOLOGY IN MONOCOTYLEDONS. 493 



rideae are very different from any thing observed in the essentially 

 apocarpous Alismaceae and Naiadeae. 



The Scitamineae (including Musaceae and Canneae) are placed 

 next, not with the intention of implying any affinity to Hydrocha- 

 ridese, which stand, as it were, alone, but because they could not be 

 conveniently placed between any two of the following orders 

 without breaking through more natural series, which they do not 

 here ; for although Hydrocharideae and Orchideae are artificially 

 connected by their exalbuminous seeds, there is no natural bond 

 of union beyond the general one common to all Monocotyledons. 

 Scitainineae are, on the other hand, notwithstanding their very 

 different seeds, in some measure connected with Orchideae by the 

 peculiar irregularity of their flowers, all but one of the stamens, 

 or at least one out of the six, being suppressed or replaced by 

 petal-like or variously shaped staminodia. 



Orchideae form one of the very few Monocotyledonous orders 

 that are absolutely defined, surrounded on all sides by a gap which 

 is nowhere bridged over. It is here placed as approaching, in 

 some measure, on the one hand the Scitamineae by the above- 

 mentioned irregularity in their flowers, and on the other the 

 Burmanniaceae by their minute exalbuminous seeds with an ap- 

 parently homogeneous embryo. 



Burmanniaceae constitute a small order, also very fairly limited, 

 connected, on the one hand, w r ith Orchideae through the ovary 

 and seeds, and, on the other, with Irideae by their centrifugal in- 

 florescence, the equitant leaves of several genera, &c, but without 

 any intermediate genus on either side. Some affinity has also 

 been suggested with 'Triurideae, founded, however, merely on 

 adaptive resemblances of some genera of less importance than 

 those which, as above mentioned, seem to connect Hydrocharideae 

 with Alismaceae. 



With Irideae we commence a long series of orders extending to 

 the Cyperaceae inclusive, nowhere separated by definite boundary 

 lines which are not occasionally crossed over from the one side or 

 the other. In the lineal series we are obliged, it is true, some- 

 times to make sudden breaks ; but that is merely owing to the 

 impossibility of representing complicated cross connexions in one 

 continuous line. Having followed out one branch of affinities to 

 its apparent end, we have suddenly to go back to take up another 

 branch which we had been compelled to leave behind. 



Irideae are for the most part distinguished from the other Epigy- 



