A REVISION OF THE AMERICAN LEMNACEAE. 95 
mud along its margin, and still farther away from the water 
where the mud was quite hard. In all instances they pre- 
served a healthy appearance, and when placed later in a 
vessel of fresh water began active growth by sending out 
normal fronds. A diligent search through the dead 
plants left by the receding water failed to reveal a single 
fruit. 
Thus we have three species which produce specialized 
fronds whose ultimate object is to reproduce the species, 
independent of the normal sexual method, after a more or 
less prolonged resting period, in much the same manner as 
the tubers of Nelumbium and many other aquatics. Hence 
the terms “resting fronds’’ and ‘‘ resting stage ’’ are here 
used in preference to ‘* winter fronds.” 
These various forms and stages have wrought much con- 
fusion. We find that a stage of one species so closely 
resembles one of another species as to cause the very com- 
mon error of a wrong determination of specimens in her- 
baria. It is this confused condition, as well as the lack of 
an American standard to follow, that has induced me to 
undertake a revision that it is hoped will meet the require- 
ment of a clear understanding of the order. 
ANALYTICAL KEY. 
I. Fronds with roots and two reproductive pouches. 
A. Roots more than one, fascicled. SPIRODELA. 
Fronds rounded obovate, 5-15-nerved. S. polyrrhiza. 
B. Root solitary. LEMNA. 
a. Fronds long stipitate, mostly submerged, forming large 
masses. L. trisulca. 
b. Fronds short stipitate or sessile, floating on the surface. 
1. Symmetrical or nearly so. 
* Oblong-obovate; fruit more or less lenticular. 
+ Upper surface uniformly green; margin of the fruit 
unappendaged. L. minor. 
+ + Upper surface mottled with irregular brown streaks; 
margin of the fruit with rounded wing lobes. L. gibba. 
** Oblong to elliptical, small, green on both surfaces; fruit 
elongated, erect. L. minima. 
5 
by as s 
PAs. Soe ees 
