72 Rhodora [APRIL 
The chief difference, according to these descriptions, seems to be 
that the leaves of L. spongiosus average a little smaller than those of 
L. calycinus, and examination of specimens bears this out. The 
achenes of the two are identical in size and shape except for a little 
individual variation in both species. 
In an attempt to determine to what extent the foliage-characters 
can be relied upon many plants have been studied, and sketches, shown 
in plate 137, have been made to illustrate the variations observed. 
Fig. 1 shows a characteristic leaf of Lophotocarpus calycinus from 
Webb City, Missouri; fig. 2 is a type occasionally found, and is from 
the same herbarium sheet as fig. 1. The leaves are often much larger, 
sometimes a decimeter long, or as large as 12 X25 cm. in var. maxima 
(Engelm.) Robinson. Fig. 3 is of a similar leaf of L. calycinus from 
Delaware City, Delaware. Figs. 4 and 5 show two leaves from co-type 
material of L. spongiosus from the tidal mud-flats of the Delaware 
River at Wilmington, Delaware. The only difference between this 
and L. calycinus is that the leaves of the former are a little smaller— 
hardly enough ground for making a species. Figs. 6, 7 and 8 are from 
three other specimens from the same station, showing how variable 
a plant may be without having any character sufficiently constant to 
set off a species. Figs. 9 and 10 show leaves from Milford, and Figs. 
11 and 12 from Old Lyme, in Connecticut, and for both stations 
exhibit the most nearly hastate leaves found on the specimens in 
the Gray Herbarium. These also begin to show the tendency which 
I wish to point out, namely that the leaves of L. spongiosus, while 
everywhere variable, tend less and less to behastate in the northern 
stations forthe plant. 'Thisis more pronounced in specimens from the 
Mystic River, Massachusetts, which was tidal until the dam was built 
at Medford. On these plants a stump of a basal lobe is rarely present 
(figs. 13 and 14). At the mouth of the Merrimac River at Newbury- 
port, Massachusetts, the leaves are of the forms shown in figs. 15 and 
16, with all vestiges of the basal lobes lost. "The blades are reduced, 
short and strap-like, and the petioles are thick and spongy’. At 
Winnegance Creek, Phippsburg, Maine (figs. 17 and 18) the leaves 
become more compressed-subulate; on the Moulies River in Kent 
County, New Brunswick (figs. 19 and 20) and at Newcastle, New 
Brunswick, on the Mirimichi River (figs 21, 22, and 23) the blades are 
almost lost and the petioles are tapering, very thick and spongy. 
1. The plant of the Merrimac River mud-flats was described by Smith as L. spalt- 
ulatus, but it may readily be seen how the plant fits into this series of variations 
of L. spongiosus, 
