ON CERTAIN EPIGYNOUS MONOCOTYLEDONS. 169 



yielded me any result which can serve as evidence in interpreting 

 the nature of the labeilum and the column. The segments of 

 the perianth generally present themselves just as they appear 

 in the full-grown flowers, with a few exceptions which are 

 readily explicable, and do not differ from known modes of deve- 

 lopment of foliaceous organs. That what is called an epigynous 

 flower first presents itself as a simple body, as observed in all 

 these plants, will not be found strange when we recollect that 

 the same has long been established in gamopetalous flowers, 

 and that in the majority of cases there is no essential distinction 

 between flowers of the two kinds, except the radial confluence. 

 I may add that I am not in a position to cavil at or to under- 

 value observations and opinions which have emanated in part 

 from distinguished philosophers ; on the contrary, those theories 

 are so much the more worthy of admiration, since in some cases 

 they were necessarily set up without knowledge of the history 

 of development. Rightly and by universal agreement have the 

 physical sciences been ruled since the middle of the last century 

 by direct experience, but far too much power would be claimed 

 for this, if we condemned everything which had not yet been 

 confirmed by it. 



If now, in reviewing these four families, we direct our at- 

 tention to the mode of arrangement, we are in the first place 

 struck by the fact that the calyx and the corolla stand in an 

 unchangeable relation to the bract and to the axis. Lindley 

 appears to wish to make the Orchideae an exception to this, 

 ascribing to this family an additional outer circle of floral 

 envelopes, occurring in Epistephium, of which organs one must 

 therefore fall next the axis and two towards the bract. And he 

 uses this as a support for his opinion of the morphological com- 

 position of the flower of the Orchideae, to which we shall return 

 in the sequel. Disregarding the objection that so rare a case as 

 the appearance of a calyculus, even if it should occur in certain 

 other genera, would scarcely prove anything for the entire 

 family, unless it were accompanied by other conditions, and 

 further, that this calyculus has not yet been investigated in its 

 earhest conditions, &c., — such a mode of arrangement is a very 



