LOU TER 
TRUE (ed ous a P ET ES 
we 
GENERAL REMARKS. 731 
The additional lobes of the corolla: alternating. with the 
true lacineze are mere processes of the sinuses or rather each 
petal may be esteemed as being 3-lobed, the lateral lobes of 
two distinct petals being frequently completely connate. 
This latter view is pointed out by their occasional emar- 
gination, the former by the fact that they are completly 
evascular. 
I once had an idea, that in all truly monopetalous corollas, 
the veins would run both in the axes of the component 
lacinez, and in the axes of the sinuses, but Gentiane at least 
the species alluded to, is an exception to the contrary, to such 
a degree that no vascular fascicle is even furnished to the 
stamina, while the lacineze are provided with three distinct 
fascicles, which have in the lamina, the ordinary distribution. 
In this respect it agrees with Villarsia. The locule of the 
anthers are after dehiscence completely opened out and 
placed back to back, they are in this state much contracted. 
They recover their original dimensions and form by immer- 
sion, and what is curious, lose immediately their red colour. 
The last point is, that the whole surface of the ovaria may 
bear ovula, of this the species alluded to affords an example 
of baccate fruit. One Cucurbitaceous plant is indeed so like a 
Vitis (Cissus) that it is impossible to perceive any difference 
until after a close examination. 
To return to Ranunculacez, the tendency of the order to 
produce petals passing into stamina is borne out by the study 
of the other nearly allied orders, such as Magnoliaceze 
Nymphezacez, etc. but the transformation may be assumed in 
these to be retrograde. It will be a curious corroboration 
if retrograde metamorphosis be found to be more complete 
than progressive. 
