490 
Transactions of the Society. 
The subsequent history of the endocysts has not since been traced. 
It remains, like so much of the life -history of diatoms, to be yet 
observed by those who have ready access to living specimens, or who 
possess facilities for cultivating them, that is, if such purely pelagic 
genera are capable of cultivation, by the methods adopted by Dr. 
Miquel, of Paris, and by our late member, Mr. Haughton Gill. 
The object of the present paper is to place on record the occurrence 
of endocysts in a genus which, though it resembles the two above- 
named in being filamentous and pelagic, is otherwise quite distinct, 
viz. Thalassiosira. 
This genus was constituted in 1873 by Dr. Cleve, for a single 
species, T. Nordenshioldii, which occurs in the Arctic regions in 
immense masses, forming a large part of what is known to mariners 
as “ whale’s food.” Its main distinction from the genera which most 
closely approach it, is that the frustules, of which the filament is 
formed, are at some distance from each other, and are connected 
together by a thread of mucus, which extends from the centre of the 
valve of one frustule to the centre of the valve of the next adjacent 
one. 
A short time ago I received from a friend, a small tube of a 
surface gathering made near the South Shetland Islands in the 
Antarctic Ocean. It was composed principally of a Thalassiosira , 
which I propose to name T. aniarctica, differing from the Arctic 
species in the characters detailed below. Fig. 3 represents the normal 
condition of the frustules forming the filament, except that as they 
have been burnt on the cover-glass to better exhibit their silicious 
covering, the thread of mucus connecting them together is not shown. 
The larger specimens evince a tendency to separate less widely from 
each other, and to form such a filament as is represented in fig. 4. 
The s.v. of the frustule is shown at fig. 1, X 1000, in order to 
exhibit the minute structure. Within the frustules, and much more 
strongly silicious than they are, occur the endocysts, sometimes in 
every frustule of the filament, sometimes only in a few. The s.v., 
fig. 2, very closely resembles, if it be not identical with, a form 
from the Arctic regions described and figured by Prof. Grunow as 
Coscinodiscus poly acanthus. In f.v. (figs. 5, 6, and 7) the outlines of 
the two valves of the endocyst are different, one valve being always 
more convex than the other. 
In this respect it agrees with the “ endocysts ” observed by 
Mr. Lauder. That gentleman remarked that “ these bodies always lie 
with their similar ends towards each other throughout the filament ” ; 
and this peculiarity is observable also in the endocysts of T. antarctica. 
The convex valve of the endocyst is always turned to the convex 
valve of the next adjacent endocyst ; and the flatter valve in like 
manner. 
Although the endocysts are frequent in the material, I have 
observed no instance of their undergoing subdivision. 
