75 
sists of a somewhat smaller body than that of the preceding 
species. In this the gramiles lying near the enveloping 
membrane are coloured red instead of green. The border of 
sarcode is narrower, and the processes, for the most part, are 
needle-shaped. The second form is represented by much 
paler specimens, with coarsely granular contents ; these are 
surrounded by a broad border of sarcode bearing very small 
processes, and often show a large, bright, and sharply defined 
circular gap (vacuole?). Whether this form really represents 
a distinct species remains as yet undecided. 
No. Ill (fig. 3a, 3b ). — This species is represented by green 
bodies, which are supplied with a narrow border of sarcode 
and a considerable number of needle-shaped processes. They 
differ from the first species in their size, the absence of 
tongue-shaped processes, the number and distribution of the 
needle-shaped processes, and in the increased proportion of 
green-coloured granules. The form of these creatures is 
circular, oval, or triangular, with very rounded angles. 
With a very highly magnifying power a cellular structure 
may be observed in the enveloping membrane, particularly at 
its outer surface, where each needle-shaped process is placed 
on the summit of an elevated cell. The lower part of each 
long needle-shaped process is often enclosed in a fine sheath 
(chitiue ?), which may be recognised after the death and dis- 
section of the creature. This sheath, which is about 7 1 77 / " in 
length and in breadth, is inserted into the summit of 
the swelling on the surface of the lining membrane. (This 
is illustrated by an ideal figure at 3b.) 
In conclusion. Dr. Focke states that he has reasons for 
believing that Eremosphaera viridis belongs properly to the 
Radiolaria. This organism, which presents a large green 
globular body, and multiplies by twofold division, was first 
considered to be a germ-corpuscle, and afterwards one of the 
Algse. 
