164 
IOWA ACADEMY OF SCIENCE 
When wie come tio consider the arrangement of the flowers on the 
stalk, we find a corresponding tendency. The blossoms are at first 
scattered spirally over an extremely elongated axis. (Many of the 
Sympetalae have chosen to preserve this arrangement, but show a com- 
pensating] y high development of the corolla, thus adapting themselves 
to entomophily by modification of the individual flower, rather than by 
grouping of the flowers.) The shortening of the axis may result (a) in 
a corresponding broadening and flattening, as in the Composites. In 
this case, the spiral origin is usually evident in the arrangement of the 
blossoms on the flattened axis, but a sufficient reduction of the number 
of blossoms would lead to a perfectly cyclic arrangement; and often 
the outer ray flowers are cyclically arranged. ( b ) The axis may be 
greatly shortened without being broadened or flattened, the individual 
peduncles being reduced but not eliminated, and the leaves in whose 
axes they stand may persist as subtending bracts. This gives rise to 
ail the varying sorts of heads, spikes, and catkins, (c) The flower 
stalks and. subtending bracts may be completely eliminated, and the 
flowers may become densely crowded on a greatly shortened axis — the 
spadix. This results in the closest possible grouping of blossoms;, the 
only thing comparable to it in this respect being found in the Corn- 
posit ae. A single bract — the basal one— persists, and often becomes 
petaloid. (d) Whorl ed leaves, in whose axes blossoms may be borne, 
and floral whorls (ex: umbels) must be regarded as examples of extreme 
reduction of the axis, and are of course cyclic. The general tendency, 
therefore, in the grouping of flowers, seems to be (1) to pass from a 
loose to a dense spiral, and ( 2) from the spiral to the cyclic habit, or to 
something approaching it. But it is to be noted that;, whereas the organs 
making up the flow r er have attained in most cases a perfectly cyclic 
arrangement, the arrangement of blossoms on the stalk is usually com- 
pletely or partially spiral, and seldom shows more than a faint approach 
to cyclicism. Apparently therefore the tendency to axis reduction and 
cyclicism as applied to the position of individual flowers on the stalk 
must be very recent as compared with the same tendency in the arrange- 
ment of the parts of the flower. 
Presumably, at some point in the ancestry of the present Sperma- 
tophytes, the individual sporophylls vdiich now compose the flower were 
spirally arranged along an elongated axis. The very evident advantages 
to be obtained by closer grouping sufficiently explain the progressive 
reduction of the axis. The closer grouping would make possible a 
reduction in number, and when the floral whorls consist each of a 
limited number of members standing in the same plane, the climax of 
