£22 ALTERATION OF POSITION. 
free central placenta, this malformation has been 
found. 
So also among plants with indefinite inflorescence, 
prolification seems very frequently to affect those 
wherein the axis is normally prolonged; thus it is 
common in Dictamnus, which plant has an internode 
supporting the pistil; it is frequent among Umbelli- 
fere, where the carpophore may be truly considered an 
axile production; it 1s common among /tosacew and 
Ranunculacee, in many of which the axis or thalamus 
is well-marked, and it is by no means infrequent in 
the flowers of the Orange, where the floral internodes 
are also slightly elongated; on the other hand, there 
is no case on record in Magnoliacee, and some other 
orders where the floral part of the axis is at some 
point or other elongated; still, on the whole, there 
can be but little doubt that there is a real relation 
between prolification and the normal extension of the 
floral internodes. 
Under these circumstances, those instances wherein 
the parts of the flower become separated one from the 
other by the elongation of the internodes (apostatis), 
constitute a lesser degree of the same change, which 
operates most completely in the formation of a new 
bud at the extremity of the prolonged axis. Some 
specimens of Gewm rivale (a plant very lable to become 
prolified) in my possession show this very clearly. In 
the wild plant the thalamus is elevated on a short 
stalk; in the abnormal ones the thalamus is simply 
upon a longer stalk than usual, or in a more advanced 
stage of the deviation the lengthened thalamus takes 
the form of a branch provided with leaves and termi- 
nated by a flower; it is noticeable, also, in these speci- 
mens, that the sepals of the lower flower have assumed 
entirely the dimensions and appearance of leaves. 
Median prolification has occasionally been recorded 
in flowers that have, in their ordinary condition, but 
one carpel, as in Leguminose and in Santalacee. In 
Leguminose, as also in Anvygdalus, it would seem as if 
