IOWA ACADEMY OP SCIENCE 
165 
cyclicism has been attained. (This applies of course to floral leaves as 
to sporophylls. Their relation will be discussed later.) 
If this theory be correct, the change from the spiral to the cyclic 
habit in the arrangement of floral parts must have occurred at a very 
early period, for in the Mesozoic Benneititales we find the spirally ar- 
ranged carpels surrounded by cyclically arranged stamens. Neverthe- 
less most plants that can be recognized as primitive show the spiral 
rather than tiie cyclic arrangement,. Spirality among Angiosperms 
must be regarded as a more or less vestigial characteristic, and the 
genera exhibiting it must be regarded as representatives of the more 
primitive phyla. No such characteristics are known in the 'family 
Araceae. Incomplete flowers consisting of a single carpel, or a pair of 
stamens, are common, and it is of course impossible to definitely classify 
such imperfect representatives of floral structure 1 , but when a complete 
flower occurs it is invariably cyclic. Therefore, all other things being 
equal, the Araceae are to be regarded as less primitive than are those 
families whose flowers show distinctly spiral arrangement of parts. An 
acquaintance with the Aroid structure makes it evident (1) that the 
process of axis reduction, and reduction of floral parts has here reached 
its climax, in the formation of a perfectly cyclic flower and (2) that 
the process of stalk reduction and crowding of blossoms has been carried 
to its climax, without corresponding reduction in number, so that the 
group of flowers remains completely spiral, no trace of an approach to 
cyclicism being discernible. The Araceae are not therefore as primitive a 
group as might at first appear. While they are perhaps one of the more 
primitive groups of the Monocotyledonous alliance, they are by no means 
the most primitive, and they certainly are far less primitive than are 
those Dicotyledons the parts of whose blossoms are spirally arranged, if 
floral structure is to be taken as any criterion.* 
*A sharp distinction should be drawn between spirality of parts of the flower, and 
spirality of flowers on the stalk. The spirality of the Aroids is not to be compared, 
for instance, with that of the Ranunculacecie . In the latter case the sporophylls are 
spirally arranged in the blossom, while in the former the cyclic blossoms are 
spirally arranged on the stalk. The difference is indicated by the fact that the 
stamens are above in the Aroids and below in the Ranunculaceae. In the flower, 
the staminate sporophylls occur below the pistillate, but on the axis of a monoecious 
plant the staminate flowers usually occur above the pistillate. (Doubtless this is a 
vestige of primitive anemophily. ) 
The writer must protest against such confusion as appears in the following state- 
ments : 
According to Engler, the general tendency among Monocotyledons is to advance 
from naked flowers with parts spirally arranged and indefinite in number to penta- 
cyclic trimerous flowers. * * * Engler has subdivided the monocotyledous into 
ten great alliances. The first six constitute the more primitive Spiral series, * * * 
the spiral arrangement and indefinite numbers occurring in one or more sets. — Coulter 
and Chamberlin. Morphology of Angosperms, p. 228. 
Among the first six orders of Engler occur the Arales, but surely nobody can 
pretend that this order has “naked flowers with parts spirally arranged.” The spi- 
rality lies entirely in the arrangement of the flowers on the spadix, and not at all 
in the arrangement of the parts of the flower. If we are to group families with 
reference to spirality, we should have two groups, one based on spirality of floral 
parts, the other on spirality of flowers on stalk. 
