1897] PHYLOGENY AND TAXONOMY OF THE ANGIOSPERMS 165 
again, none at all (naked). Since these often occur on plants 
which are clearly related to those bearing a fully developed 
perianth, we are led to the conclusion that apetalous and naked 
flowers are modifications of the common flower structure. Thus, 
there can be no question as to the relationship of Clematis, Anem- 
one, Thalictrum, Caltha, Hydrastis, etc., to Ranunculus, Myo- 
surus, Coptis, Delphinium, and other genera of Ranunculacee. 
So, too, who questions the relationship of our apetalous maples 
(Acer saccharinum L. and A. negundo L.) to the remaining species 
of the genus, or of our ashes (Fraxinus sp.) to the old world 
petalous species? In these and many other cases we see clearly 
that the apetalous condition of the flower is one derived from 
the normal structure in which the complete perianth is present. 
There are, however, many apetalous dicotyledons whose rela- 
tionship botanists have not been able to agree upon. Thus 
Bentham and Hooker in their Genera Plantarum enumerate thirty- 
six families, including 84g genera, and 12,100 species, in the 
artificial group Monochlamydez, which they separate from their 
Polypetalea solely by the simple (or absent) perianth; Engler 
and Prantl in their Pfanzenfamilien bring together into a hetero- 
geneous group twenty-four families of mostly apetalous plants, 
including nearly 6000 species. All of these, excepting the Ola- 
cacez, are included in Bentham and Hooker’s Monochlamydee, 
So that we have in Engler and Prantl’s arrangement a reduction 
of Monochlamydez amounting to fully one-half. This has been 
accomplished by a distribution of apetalous plants among those 
whose flower structure differs only in regard to the perianth. 
That this reduction could have been carried further without 
doing violence to our knowledge of relationship will be admitted 
by most systematic botanists. Thus we may readily remove the 
Olacacez, which havea perianth consisting of calyx and corolla, 
and with them may go the sandalworts (Santalacee), proteads 
(Proteacez), loranths (Loranthacee), and perhaps the balan- 
ophorads (Balanophoracee), all of which are more or less clearly 
related to the typical Celastrales. So too the willows and pop- 
lars (Salicacez) differ from the tamarisks (Tamaricacee) only 
