. GOETHE ON THE METAMORPHOSIS OF PLANTS. . 15 
56. Still more striking and remarkable alterations are produced in 
the petals of different plants. A small cavity, filled with a honcy-like 
liquid, occurs in the inner base of some flowers. This cavity is much 
deeper in some families and species than in others, and is elongated at 
the back of the petal in the shape of a spur or horn ; the rest of the 
petal being also more or less modified in form. The genus Aguilegia 
is a good example of this.* 
57. The nectary is most disguised in Aconitum and Nigella; but 
even here its similarity to the ‘ leaf-form may be perceived by a little 
attention. It has a strong tendency in Vigella to become petaloid, the 
flower becoming double from the altered nectaries. In Aconitum the 
resemblance of the nectaries to the helmet-shaped sepal beneath which 
they are concealed, is evident. 
58. Having observed above that the nectaries may be considered as 
transitional organs between petals and stamens, we may here introduce 
a few remarks on irregular flowers. In Melianthus, the five outer divi- 
sions may be described as true petals, and the five inner ones as an 
accessory Corolla consisting of six nectaries ; of which the superior one 
is most like the petals, whilst the inferior one, commonly called the 
nectary, most differs from them. In the same sense the keel of papi- 
lionaceous flowers £ might be called a nectary ; since of all the petals 
it is nearest in form to the stamens, whilst it differs widely from the 
leaf-like form of the standard (vexillum). Thus also the brush-like 
from the ordinary stamens except in the breadth of the filament.” Anthers so placed 
are commonly met with in some of the double Nareissi. 
Schleiden also asserts that the nectaries of Ranunculus and Parnassia, the scales 
of Sileneæ, and the erown of the Passion-flower, are secondary productions from the 
‘petals, and not independent foliar organs; but, on the other hand, some of the rays 
of the crown of the Passion-flower have been observed to be replaced by anthers 
(Moquin-Tandon, “Tératologie Végétale,' p 220), while in Passiflora Murucuja the 
rays are combined iuto a cup, like that of Narcissus or like that of Melia, except 
that it does not bear anthers. In Saponaria and some others of the Si/enee we 
have remarked the scales of the corona bearing anthers as though they were re- 
ferable to the adhesion of two stamens, the anthers of which are us y wanting 
(Journ. Linn, Soe, i. 1857, p. 159). 
* In Angraecum sesquipedale, an Orchid native of Madagascar, the nectary mea- 
sures nearly a foot in len 
+ The parts called by Goethe nectaries, in the Aconite, Nigella, etc., are now con- 
sidered as petals ; the outer pieces as sepals, in spite of their colour and form. In 
the Winter Aconite, Brantäis hyemalis, a transition may sometimes be seen between 
the large, flat, coloured sepals, and the small, tubular, greenish petals (nectaries). 
See Plate, Fig, 3. 
t The ET in papilionaceous flowers is evidently formed by the junction of two 
