a substantial stalk which represents the fused proximal parts of the individuals, as in a 
Xenia. The various stalks arise from a thickish membrane which encrusts pieces of sponge, 
coral, etc. In other cases or in other parts of the same colony, the polyps arise individually 
from the basal membrane, as in many Clavularians. 
(2) Each polyp is heavily armoured and often glistening with longitudinally disposed spindles, 
more or less definitely in chevron, each double row ending in a point at the base of the 
tentacles. The tentacles show a dorsal transverse row of minute curved spindles, but all of 
them are inturned. There is no question of the heavily armoured polyps being retracted 
into the basal membrane or into the stalk. 
(3) The characteristic spicules of the polyps and of the superficial coenenchyma are relatively 
substantial tuberculate spindles, sometimes straight, but usually curved (arcuate and lyriform). 
They bear in varying degrees of closeness short conical rather blunt tubercles, standing 
abruptly at right angles. These spindles differ in size, proportion of breadth to length, 
degree of roughness, and size of tubercles, which are sometimes compound and interlocking. 
The following measurements were taken : 0.95 X °-°5 mm -; 0.44 X 0.03 mm. 
The spicules on the aboral surfaces of the tentacles are minute spindles with very small 
roughnesses, and with somewhat blunt ends. They are arranged in two sloping rows, irregularly 
in chevron. Average dimensions are 0.12 X 0 - 01 mm. 
The specimens are much broken, the largest is 3.3 cm. by 1.3 cm., with a height of 1.7 cm. 
The polyps of the oldest specimens arise from a composite stalk, as in most Alcyonacea, 
and this may have a height of 11 mm., with a diameter of 5 mm. 
When a stalk-portion is developed above the encrusting base, it shows in cross section 
a continuation of the several cavities of the polyps which it bears, and the canal walls separating 
these cavities are supported by numerous typical spindles. The whole texture of the colony is 
stiff, though the heavily armoured polyps are flexible. It may be added that when there is a 
stalk portion the polyps arise from it at various levels and irregularly. 
Our specimens agree in appearance and detail with those described by Thomson and 
Henderson as Coelogorgia repens n. sp., but they include older stages which were not represented 
in the Zanzibar collection, and these show that the reference to Coelogorgia was an error. 
Thus they show that there is no possibility of distinguishing a primary axial polyp in the various 
clusters. Moreover, the typical species, C. palmosa , is arborescent; the walls of the axial polyp, 
surrounding the central coelenteric canal, are thick and include secondary canals. The spicules 
of our Siboga specimens do not show the compound warts characteristic of Coelogorgia palmosa , 
nor any tendency to form pseudo-clubs. 
This interesting type cannot be referred to the Stolonifera if we adhere to the definition 
of this order, as including forms whose basal portion does not rise above the level of a membrane 
or stolon , from which the polyps arise individually. In the Siboga type a distinctive feature in 
the older colonies is that several polyps arise from a common stalk growing up from the basal 
membrane. It may be added that the Siboga specimens could not be referred to any known 
genus of the Stolonifera. 
It seems to us that the occurrence of the polyps in bunches formed from a coherent 
continuation of their several canals, points to the inclusion of our type in the order Alcyonacea. 
