588 PLANTS JAVANICiE RARIORES.
no] In describing the genus Aikinia (Epit/iema of Dr. Blume) I regarded Cyrtandracea, or Cyrta?idre<$, for the reasons now assigned, as a tribe merely of Gesneriacece, mi distinguishable from that portion of the order with hypogenous corolla, or Besleriea, by characters either of little importance or which required confirmation. For although, in addition to the characters referred to, Cyr- tandrece differ very remarkably in geographical distribution from the rest of the family, yet this difference is not entirely without exception, as I have already noticed in my account of Loxotis.
But whether these groups be considered as distinct families, or as tribes only, it will probably be admitted that in a natural classification Cyrtandrete must stand next to Besleriece ; while on the other hand they appear to be very nearly related to Bignoniacece, with which they are connected through Incarvillea, particularly with that sec- tion of it which in Dr. Royle's Illustrations I have de- scribed as a sub-genus, and named Amphicome. This sub-
thetical expression of the facts was not applicable even to all the families he lias cited in support of it. And I concluded that as a general rule the point most to be depended on in proving the importance of albumen in systematic botany was its relative quantity, especially when accompanied with a low degree of development of embryo ; for where the albumen forms the great mass of the seed in any known portion of a natural family, it may in most cases, though not always, be safely inferred, not only to be present, but to exist in like proportion in the whole of that family. This rule, however, I regarded as merely empirical, founded on extensive experience, but not necessarily con- nected either with uniformity or even apparent importance of function ; for while in some families in which its proportion to the whole body of the seed is the greatest, it constitutes the early nourishment of the embryo, in others, where it exists in equal quantity, it is either not at all or but slightly acted upon in germination. I stated also that there were cases in which this character was of reduced importance, existing only in certain tribes of one and the same great natural family, as in Rubiacece ; nor are there wanting instances in which it is only of generic value. 1 And, lastly, I noticed that in several families, in which the constancy of the character was very general, exceptions occurred, dependent on the apparent necessity for an unusual development and increased energy in the embryo, connected either with the unfavorable circum- stances in which it was destined to vegetate, as in plants growing in or exposed to the action of salt-water; or where great resistance, arising from the structure of the pericarpium, or even from the texture of the proper integu- ments of the seed itself, was to be overcome in germination.
1 'Linn. Soc. Trans., 5 vol. x. p. 36 [vol. i. p. 23], et ' Prodr. Flor. Nov. IIoll.,' vol. i. passim.
�� �� �
