CHANGE OF DIRECTION, 209 
like torus are bent downwards towards the stunted 
carpels instead of being, as they usually are, straight. 
Amongst orchids, where the pedicel of the flower or 
the ovary is normally twisted, so that the labellum occu- 
pies the anterior or inferior part of the flower, it fre- 
quently happens, in cases of peloria and other changes, 
that the primitive position is retained, the twist does 
not take place, and so with other resupinate flowers. 
In Azaleas a curious deflexion of the parts of the 
flower may occasionally be met with. Fig. 112 shows 
an instance of this in which the corolla, the stamens 
and the style were abruptly bent downwards : as 

Fre. 112.—Flower of Azalea, showing the corolla reflected. 
young flowers of this singular variety have not been ex- 
amined it is difficult to form an opinion as to the cause 
of this variation. In one plant the change occurred in 
connection with the suppression of all the flowers but 
one in the cluster, or rather the place of the flowers 
was occupied by an equal number of leafy shoots. 
Moquin' mentions a flower of Rosa alpina in which 
two of the petals were erect, while the remaining ones 
were much larger and expanded horizontally. The 
same author quotes from M. Desmoulins the case of a 
species of Orobanche, in which a disjunction of the 
petals constituting the upper lip took place, thus libe- 
1 Loe. cit., p. 315. 
14 
