SYMMETRY OF THE FLOWER.—COLLATERAL CHORISIS. 359 
while the other four are placed in pairs opposite the anterior 
and posterior sepals; we have here, therefore, four stamens 
instead of two, which results from the collateral chorisis of those 
two. Insome Cruciferx, as Streptanthus (fig. 795), we have a 
strong confirmation of this view presented to us in the fact that, 
in place of the two stamens, as commonly observed, we have a 
single filament forked at the top, and each division bearing an 
anther, which would seem to arise from the process of chorisis 
Fia. 795. 
Fic. 794. 
Fig. 795. 
Flower of a species of Strveptanthus, with the floral envelopes removed, 
showing a forked stamen in place of the two anterior stamens, From 
Gray.—Fig. 796. Diagram of the flower of the Fumitory. 
Fig. 794. Diagram of the flower of the common Wallflower. 
being arrested in its progress. The flowers of the Fumitory are 
also generally considered to afford another example of collateral 
chorisis. In these we have two sepals (jig. 796), four petals in 
two rows, and six stamens, two of which are perfect, and four 
more or less imperfect ; the latter are said to arise from colla- 
teral chorisis, one stamen here being divided into three parts. 
Other examples of this form are by some considered to be af- 
forded by the flowers of many species of Hypericum ( fig. 554, 
f, f); in which each bundle of stamens is suppesed to arise from 
the repeated chorisis of a single stamen. 
Collateral chorisis may be considered as analogous to a com- 
pound leaf which iscomposed of two or more distinct and similar 
parts. Transverse chorisis is supposed by Gray and some other 
botanists to have its analogue in the ligule of Grasses (fig. 374, 
lig), as that appendage occupies the same position as regards the 
leaf as the scales of Lychnis (fig. 501, a) and other plants do to 
the petals (see page 239). 
Lindley held that the whole theory of chorisis ‘is destitute 
of real foundation, for the following reasons :— 
‘1. There is no instance of unlining which may not be as 
well explained by the theory of alternation. 
‘2. It is highly improbable and inconsistent with the simpli- 
