20 DR. J. D. HOOKER ON THE STRUCTURE AND 



tion of Ilaloragefe, of which I regard them to be reduced forms, it cannot be disputed that 

 amon"st Phaniogams tliere are few groujjs so xiniformly incomplete as regards the normal 

 complement of organs, or the arrested development of those organs which are present. 

 The value of this consideration is however much diminished hj the fact, that there are no 

 limits to the su])pression of organs in the individual genera of Orders which are, never- 

 theless, typically highly developed. 



In a systematic point of view, the value of these suppressions in the Vegetable Kingdom 

 diminishes to a great extent in ascending from the root towards the ovary : thus, the 

 al)seucc of a root of the ordinary structure, and the adaptation of the lower portion of the 

 stem to a parasitic attachment, occur in six or seven natural families of Exogens which 

 are normally terrestrial, and perhaps in many more. A total absence of leaves, or a 

 reduction of them to minute scales, occurs in many natural families. A reduction of the 

 whole plant to a leafless, single- or few-flowered stem, is found in many parasites, and in 

 Oi'chidece, Burmanniacece, EricecB, Scrophularince, Trkiriclea, liafflesiacece, Gentianece, 

 and other families which have no mutual affinities ; whilst the reduction of the inflo- 

 rescence to a single flower, and the parts of the latter to its essential organs, is too 

 frequent to need specification. That of the embryo to a homogeneous mass is found in 

 various genera, as indicated by Brown and Griffith : the reduction of the ovule to an 

 embryo-sac is. however, in the present state of oui- knowledge, almost peculiar to Bala- 

 aophorea. 



As regards the including the Monostyli and Bistyli under one Natural Order, these 

 are so manifestly different, that it is a theoretical question how far, were there more 

 genera of each, or had they a widely different geographical distribution, they would 

 l)y common consent have been imited into one natural family : and the same argument 

 might indeed be applied Avith equal force to the removing Ilystropetalon and even Cy- 

 nomorimn. 



It is difficult to indicate any particular genus of Balanopliorece which can be considered 

 typical of the Order, though Cynomorluiu may be taken as such for the Monosiyli, and 

 Helosis for the Distyli. 3Iystropetalon, though in many respects the most perfect 

 genus of the Order, cannot in any degree be considered typical of it ; for it departs far 

 more widely from the prevalent structure of its allies than any other genus does. Our 

 ideas of what is or is not typical, are, however, vague and ai'bitrary; the ideal type 

 being either the prevalent form of the group, or that which unites most of the pecu- 

 liarities which distingTiish it, or that which possesses the fullest complement of organs 

 united in one individual, or that in which these are most complex, as well as specially 

 adapted to the functions they perform. 



Classification of Balanophobe^. 



In the following arrangement of Balanophorece, I have been chiefly guided by the 

 structure of the female flowers, which are generally found to afford the most important 

 characters for systematic purposes. 



The primary division into Monostyli and Bistyli was proposed by Griffith (Linn. Soc. 



