149 
our specimens to this species on account of the strength of the supporting bundle, and the 
fact that the anthocodial points consist of a single pair and not of four pairs of spindles. 
9. Stereonephthya tilicoides n. sp. 
Station. Not recorded. 2 Ex. 
This new species has been established for two complete colonies and several fragments. 
From an encrusting base there arise a number of radiating branches each of which gives off finger 
shaped Araucaria- like lobes. All the lobes arise almost vertically, so that they lie approximately 
parallel. They are densely covered with polyps, so that little of the naked cortex is seen. 
The polyps have short stalks. The supporting bundle arises directly from the lobe and 
the anthocodise hang down at an acute angle like the flowers of a Campanula. The sup¬ 
porting bundle is strong and consists of a median dorsal ridge of three or four spindles flanked 
on each side by three or four laterals. The whole forms a very rigid buttress. Sometimes the 
three or four of the median ridge project, but most commonly only two. Five of the points of 
the anthocodise are represented, but the en chevron rows are very indefinite-, the three ventral 
points are rudimentary or absent. Between the tips of the points there are a number of inter¬ 
mediate spicules. These spicules are white and strongly refractive. They are arranged in small 
groups which stand out as white lobes. The tentacles are heavily armoured with a double row 
of short spicules which are linked up with the intermediates. 
The points are made up of from five to eight pairs, but six is the most common number. 
In some respects this species comes near S. ulex , but the branches cannot be described 
as “very rigid” or Madrepore-like. They are well armoured but not rigid. The polyps are also 
more closely arranged on the lobes than in A. ulex. 
Genus Siphonogorgia. 
The genus Siphonogorgia was established by Kolliicer in 1874 for Siphonogorgia godefroyi, 
n. g. et sp., which he regarded as intermediate between Alcyonidae and Gorgonidae (Briareinae 
in particular). But this was before the clear contrast was drawn between Pseudaxonia, where 
the axis is really coenenchymal, and Axifera, in which the axis is invested by the coenenchyma. 
The points of Kolliicer’s diagnosis were : 
(1) that the coenenchyma is hard and includes much connective tissue with many needles; 
(2) that the polyps occur only on the ends of the smallest branches, and are retractile into 
slightly projecting calyces; 
(3) that the coelentera are continued into canals, which penetrate the whole interior of the colony, 
inside both twigs and stem ; 
(4) that only four mesenteries are continued into the elongated coelentera, namely those that 
bear the gonads and the long narrow mesenteric filaments; 
(5) that the polyp-calyces have slightly developed opercula. 
Wright and Studer (1889) widened the conception of Kolliker’s sub-family, Siphono- 
gorgiacese, so as to include all the Nephthyidae that have abundant spicules in the partition 
