97 
toppart of the colony. I he curved colony is somewhat laterally compressed; the greatest breadth 
is 3 cm., the smallest breadth, on the same level, is 1.5 cm.; the toppart is blunt. A worm-tube 
lies against the stem, almost from the base to the top ot the' colony. — The branches are not 
arranged in regular longitudinal rows., except sometimes for a short space. Their variable 
mutual distance is ± 3 mm. Many branches are simple, especially on the two flat sides of the 
colony. On the other sides many branches are branched. While the primary branches are almost 
at fight angles with the axis, with some exceptions which are distally inclined, the secondary 
branches are curved towards the outside of the colony: Tertiary branches occur also. The 
secondary branches diminish in length towards the top of the primary ones, and usually they 
are not arranged in longitudinal rows but in verticils of 3 to 4; their bases, in the same verticil, 
are not always on the same level; the members of one verticil are not always equal in length. 
I he tertiary branches are also arranged in verticils if they are present in a sufficient number. - _- 
Frequent fusions occur between neighbouring branches, not only of 
the coenenchyma, but also of the axis itself. 
On the top of the branches, which is separated from the rest of 
the branch by a sudden diminuation of the diameter, the smooth spines 
are arranged in longitudinal -rows, without further regularity. Their 
mutual distance is 300 a ; they are triangular with a concave distal side 
(fig. 90 a). On the slightly larger parts of the axis the longitudinal rows Fig. 90. Parantipathes coiumndris 
have disappeared, so that the distribution is wholly irregular, especially n>uch ' ) Brook ; a Spine on the 
since the mutual distance is very variable (fig. 90^); sometimes short on the lower part of an ultimate 
longitudinal rows put in appearance but never for a great distance. The 
spines are here more blunt and their base broadens gradually and imperceptibly into the axis. There 
is no difference in length between the spines of opposite sides of the axis, but there is a 
difference in shape (fig. 91). The round knobs are situated on 
the polyp-bearing side of the axis. The length of the spines is 
50 p.; their mutual distance is at least 225 p.. If there is any 
regularity in the distribution, there are 4—5 longitudinal rows, 
with a mutual distance of 375 u. between the spines. But the larger 
the branch, the more irregular is the distribution of the spines; 
Fig. 91. Parantipathes columnaris (Duch.) 
Brook. Spines on opposite sides of a 
branch; 52 >4 
their length diminishes and many spines disappear. — The axial lumen is very small (20 p.). 
The polyps (figs. 92, 93) may occur on every side of the primary branches, but not in a 
series, and on the outward side of the branches of a higher order, in a 
single series. The tentacles are short: 180 p,; the oral cone is very low, 
flat and broad, with a transversal diameter of 440 p„; the sagittal diameter 
is slightly less. The mouth is almost round. The tentacles are nearly 
radiate; on the larger branches the distance between the sagittal tentacles 
and the middle of the mouth is 375 p., and 450 p. between the lateral Fi s- 9 2 - Parantipathes eoium- 
»ar/j(Duch.) Brook, «, b Polyp 
tentacles and the middle of the mouth. So there is a tendency to a on a rat her thick branch; 
transversal elongation, which on the larger branches is not so conspicuous 6 14 x ' 
as on the thinner branches, where the polyps are very elongated (fig. 93), as is necessary in 
SIBOGA-EXPEDITIE XVII. ' 13 
vy 
