100 
differs from P. ziziiformis by the leaves, in Græbner, 1. c. 
p. 79 thus described: folia — subabrupte in petiolum 
brevem attenuata» Nothing is noted of the presence or ab- 
sence of teeth in P. brasiliensis, nothing of the stem-struc- 
ture; and the same must be said of P. Schweinfurthii, that 
in Flora of trop. Africa is placed at the side of P. Richardi, 
but in Græbner 1. c. among »lucentes» between P. lucens 
and P. Zizii. If authors had taken as a rule to note such 
essential characters, we should be released from uncertainty 
concerning plenty of species. For instance: Although they 
have contested for hundred years of P. fiuitans», none till 
now has observed that the margin of the submersed leaves 
3. Pot. ziziiformis Hagstr, Top of a stem-leaf. 4 /i. 
4 — 6. Pot. dentatus Hagstr. 4. Cross-section of the central 
cylinder, outlined. 5. Cross-sect, of part of the stipule at the base. 
6. Basis and top of a stem-leaf, nat. size. 
in young state always is provided with a fine serrulation 
that makes this species (we aim at P. nodosus not the hy- 
brid P. lucens + natans) to be referred to a group separated 
from P. polygonifolius and its relativs (see Raunkiær, Anat. 
Potamog.-studies, 266, and Fischer, Die bayer. Potamoge. 
tonen 1907. 31. 53!). P. decipiens Nolte, likewise, always 
with denticulation in the young leaves, nevertheless by 
Ascherson u. Græbner, Synopsis, 1897, and Græbner, Pota- 
mogetonaceæ, 1907, divided in two parts: lucens -f- perfolia- 
tus und lucens + prælongus, the latter Blätter — — ganz- 
randig; , foliis — — integris >, to which also P. berolinensis 
is referred, although its margins at least in the original 
