168 APPENDIX. 
family whose principal feature is to possess anisomerous flowers ; 
but in the third case we avoid these difficulties and ensure con- 
sistency, preserving at the same time the peculiar characteristic 
features both of the Solanacee and Scrophulariacee : we should 
then have thus, 1. Solanaceae, offering isomerous flowers with a 
valvate or induplicato-valvate zestivation ; 2. Atropacee, isomerous 
flowers, or nearly so, with imbricate or a peculiar zestivation ; and 
3. Scrophulariacee, anisomerous flowers with imbricate estiva- 
tion. In any of the three modes of distribution above indicated, 
it matters little which we adopt, in regard to the absolute ar- 
rangement of the various genera, for in every case they remain 
alike, in exactly the same linear order of position. The value of 
the Atropacee, as a distinct order, must now rest entirely on its 
own intrinsic merits: its adoption seems the only course by 
which a large amount of inconsistency can be removed, and it 
appears to me a far less objectionable plan to call up a new 
family, than to destroy the great landmarks that serve to discri- 
minate the limits of two of the most natural families in the 
system. 
Having shown the arrangement proposed for the distribution 
of the Atropacee, I must offer the following explanation. The 
division into the suborders Rectembryee and Curvembryee, as 
proposed by Endlicher, and followed by me in the arrangement 
of the Solanacee formerly given in ‘ Lond. Journ. Bot.’ v. 148, 
offers by far too inconstant and doubtful a character to be main- 
tained there, or be adopted here; for among the Salpiglossidee, 
some species of Petunia possess an embryo nearly straight, and 
more curved in others, while in Salpiglossis it is often spiral 
bent into more than a complete gyration. I have preferred rather 
to follow the estivation of the corolla, as it gradually verges from 
the plicato-valvate of the Solanacee into the imbricate mode of 
the Scrophulariacee : thus in the tribes Nicotianee and Daturee 
we have the contorto-conduplicate, a form by no means valvate, 
but the first departure from it : in the Duboisiee we have another 
advance, where the lobes of the border are seemingly valvate, but 
on examination their margins will be found convolutely inflected, 
a form which I have named volutive: in the Salpiglossidee it 
assumes the next step here denominated reciprocative : in the 
Petuniee we have again another degree, which is only a modifi- 
cation of the imbricative, and which I have termed replicative : 
and finally, in the Hyoscyamee, Atropee, Solandree and Bruns- 
Felsiea, it becomes decidedly imbricative and quincuncial, as in 
the Serophulariacee, with which natural order the latter tribe most 
closely osculates. In the Atropee the amount of imbrication is 
small in extent ; in the genera Brunsfelsia and Solandra it is ex- 
cessive im amount, the lobes wholly enveloping one another in 
. 
