Igor] EREMOSPHAERA AND EXCENTROSPHAERA 311 
have been collected during almost every month of the year, 
always in considerable quantities. 
In August, 1899, while on Naushon, one of the Elizabeth 
islands, near the place on the east shore known as Tarpaulin 
cove, Eremosphaera was found in its typical habitat, namely a 
low sphagnum swamp. There was practically no difference 
between the forms from the two localities except that of size, 
and this was quite characteristic. Out of the two hundred and 
fifty measurements of the specimens from Ridge hill (fg. 7), 
the average diameter was 75.45 with a maximum of 105 and 
a minimum of 67.5; while of the same number of spheres from 
Naushon (fig. 3) the average diameter was but 35 with a min- 
imum of 31.54 and a maximum of 40.7. The figures above 
referred to will give a fair idea of the comparative sizes of the 
two lots of material. 
This marked difference in size has been noticed by Chodat 
(1), who in the spring of 1892 found Eremosphaera which was 
almost exclusively the large form, while in 1894 he discovered 
among Sphagnum and Carex a quantity of the small variety. 
The relative sizes of these two extremes is not given, except for 
the statement that by careful search one might find ‘‘giant indi- 
viduals” of 170". De Bary, in his original notice of the plant, 
gave the measurements as about 60, while Rabenhorst says 
43-49". Kischner gives 100-145, and De Toni covers both 
extremes by citing 100~150m, as the usual size, 30-804 spocuucas 
being occasionally found. It is evident, then, that the plant is 
one varying within wide limits. The maximum of the Naushon 
form not coming within 26m of the minimum of the Ridge hill 
form, if size be a sufficient criterion there would be no diff- 
culty in recognizing a variety ™m1nor, measuring from 30-414, and 
a variety major, measuring from 67-100 or more microns. There 
can be no question that the smaller forms, found by me at 
Tarpaulin cove, were mature plants. The arrangement of the 
chromatophores and general cell contents was in every way iden- 
tical with the larger forms, and division took place as readily in 
cells 334 in diameter as in those measuring 1004, Furthermore, 
