290 The Ohio Naturalist. [Vol. XI, No. 5, 



have been adapted from older classifications. In this connection 

 it might l^e stated that the Bentham and Hooker scheme of 

 classification contains certain features which should not be thrown 

 overboard bodily without due consideration. 



As stated in a previous article of this series, all possible char- 

 acters and peculiarities should be considered in segregating 

 groups. Gross and microscopic, and external and internal mor- 

 phology, as well as embryology, physiology, and life cycle are 

 important and must be taken into account if contradictions in 

 supposed lines of descent are to be avoided. But when the groups 

 have thus been established, one or at most, a very few definite 

 characters in combination should give an exclusive definition. 



It will be evident to anyone, who has considered the subject 

 in some detail, that the groups of monocotyls and dicotyls cannot 

 be segregated on the basis of the flower alone, although the flower 

 is perhaps the most important stnictm'e in the Anthophyta to 

 indicate relationship. There may be apocarpous and syncar- 

 pous, apetalous and choripetalous, monosporangiate and bispo- 

 rangiate, and numerous other diverse developments in very 

 closely related groups. From an evolutionary point of view, the 

 starting-point of floral development inust be sought among the 

 homosporous and heterosporous Pteridophyta. The flower of 

 the higher plants then seems to have come from a definite, 

 bisporangiate strobilus or cone. This is especially apparent in 

 the angiospenns where the monosporangiate flower usually shows 

 vestiges of one or the other set of sporophylls. These vestiges in 

 the angiosperm flower are ver}^ conclusive, and in deciding whether 

 a given structure is primitive or specialized their recognition 

 becomes of primary importance. 



The general progression is then about as follows: 



1. Indefinite bands of sporophylls with further growth of the 

 axis. 



2. Definite bisporangiate strobili. 



3. Development of a perianth in the Anthoph^^ta. 



4. Reduction of the fioral organs to definite cycles and 

 numbers. 



5. Extreme modifications in the tyjjical floral organs and also 

 in the parts immediately surrounding. 



As often pointed out the evolutionary lines in the flower are 

 then : 



1. From spiral to cyclic and to reduced cycles, in the 

 monocotyls mostly trimerous, occasionally tetramerous or dimer- 

 ous, and in the dicotyls mostly pentamerous, but occasionally 

 tetramerous, trimerous or dimerous. 



2. From pentacyclic to tetracyclic or still fewer sets. 



3. From hypogynous to pcrigynous and epigynous conditions. 



4. From parts free to parts united, as from apocarpy to 

 syncarpy. 



