204 
DR, carpenter’s RESEARCHES ON THE FORAMINIFERA. 
which serves to bring the superficial and columnar segments of each zone into mutual 
communication. 
29. As the description now given of the superficial layer applies equally to both 
surfaces, we may now proceed to i\\e intermediate When this is laid open by a 
horizontal section (Plate V. fig. 6), it is seen to consist of a series of concentric zones, 
the cells of which alternate with each other, like those of the simpler type 17-). 
The cells are usually circular (or nearly so) in form ; but seem to differ consider- 
ably in size, even in different parts of the same zone. Their borders, however, very 
commonly present a funnel-like aspect; and thus we perceive that the diameter of 
the cavity is liable to vary, according to the part of it which the section happens to 
traverse, — a fact which becomes mure obvious when vertical sections are examined ; 
for it is then seen (Plate VI. figs. 7, 8, 9) that each columnar cell is narrowed by 
constrictions at intervals, so as to divide it imperfectly into a series of segments 
vertically superposed one upon another. The number of th&se segments varies 
according to the thickness of the disk ; so that it is anything but constant, either 
in different individuals, or in different parts of the same. Moreover, it may be often 
observed that the columns neither always pass from end to end in a straight line, nor 
maintain a complete isolation from each other (Plate VIII. figs. 1, 2) ; an inoscula- 
tion of two columns not being unfrequent, and more rarely a fusion of two columns 
into one. All these features of structure presented by the shell, are beautifully dis- 
played by the animal (Plate IV. fig. 4) ; the columns of sarcode (cc, c'c') exhibiting 
the imperfect transverse segmental division, the not unfrequent inosculation, and 
the occasional fusion, which we have seen to exist in the cavities which they occupy. 
At their upper and lower extremities, they unite with the horizontal bands {hh' , bb'), 
which pass continuously round, in each zone, between the intermediate and the super- 
ficial layers. 
30. Save in the case of such accidental inosculations as those just noticed (which are 
indicated in vertical sections like that represented in Plate VI. fig. 7, by the irregularly 
disposed apertures h, h), no other lateral communication seems to exist between the 
contiguous cells of the same zone, than that which is established by the annular sto- 
lons just mentioned. The cells of the successive zones communicate with each other, 
however, as in the simple type previously described (^ 17.) ; but with a curious modi- 
fication ; for whereas a horizontal section of the latter shows that each cell communi- 
cates with the two cells alternating with it in the interior zone (Plate V. fig. 1), a like 
section of the Orbitolite of complex type seems to show that such a connexion exists 
with only one cell of the interior zone, by a passage running obliquely from one to the 
other, and extending continuously through several successive zones (Plate V. fig. 6, i, k), 
I was long perplexed by the want of constancy in the direction of these passages; 
the very same section exhibiting opposite obliquities in contiguous parts (Plate VI. 
fig. 2). By the study of vertical sections, however, made tangentially instead of 
radially, so as to cross these connecting passages, I arrived at the explanation of this 
