shows no enlargement, and the tip is hair-like. No regular series of short bristles
occurs in these tufts, though one or two are observed at the edges.
The first row of hooks is on the fifth segment, the first two elevations being devoid
of them., On the fifth the row is at the posterior part of the segment at some distance
ventrally from the setigerous process. They increase a little in length after the first,
and remain nearly equal for a considerable distance, gradually, however, becoming
elevated, so that about the twenty-fifth, prominent uncinigerous processes are formed,
and posteriorly they stand out like the “ legs” of caterpillars. Toward the tip of the
tail they diminish on the narrow segments and incline ventrally, so that those of opposite
sides approach, and they cease at the last segment. In an example from St. Andrews
the uncinigerous processes were irregular posteriorly, being crowded on one side, and
scantily distributed, even with blanks, on the other—probably from injury. The hooks
are arranged in a single row throughout and are smaller than those of Lanice conchilega,
but more numerous, no less than 137 occurring in a row anteriorly, but posteriorly
the number diminishes. Each hook (Plate CXXVI, figs. 6 a—d) in lateral view
presents two teeth above the main fang. The posterior outline is short and has a dimple,
whilst the base is elongated and convex inferiorly. The anterior outline (below the
main fang) is smooth, often slightly convex, and merges into the prow .(anterior
part of the base), which is prolonged as a stout process with a slightly dilated tip, so
as to resemble a stud, whilst below this the prow curves a little forward, and the inferior
outline is slightly convex. In some a smaller process of the prow (less than that shown
in fig. 6 b) occurs beneath the stud, as, for instance, from the south-west of Ireland.
The occasional occurrence of a second tooth above the main fang brings the Heterophenacia
Renouardi of Marion,1 from Marseilles, nearer this species.
The tube is generally attached inside bivalves from deep water, and contains more
of the tough membranous secretion and fewer shelly elements than that of Lanice
conchilega; moreover, its anterior end is not branched. The anterior glands and the many
glandular bodies which give a variolous appearance generally may also aid in the secretion,
and in some the body is surrounded by a milky secretion before death. "When in vigour
the animal rapidly constructs a tube. In many the tube shows a tendency to assume a
spiral form. The forms attached to the secretion are spines of Amphidoti, fragments
of shells, corallines, masses of mud and Foraminifera. In Greenland the tubes of
the arctic type are sometimes made of sand and Foraminifera (Ditlevsen). In the
Mediterranean they are formed of coralligenous algae, leaves of Posidonia, vegetable fibres
and other debris (Lo Bianco).
Reproduction.—In an example captured on the 16th October numerous ova were
found in the coelomic space. Lo Bianco (1909) observes that individuals carried ova
from August to October at Naples.
A distinct variety of Thelepus cincinnatus, if not a separate species, was dredged
by the “ Knight Errant ” at Station 8, lat. 68° 4', long. 7° 37', in 305 fathoms on mud.
Only the hook is at present available, but it differs from that of the ordinary form
in the depth and proportional shortness of the base, which is boldly convex inferiorly.
It agrees with the northern Thelepus in having only a single tooth above the main fang.
1 ‘ Revue des Sc. nat./ t. iv, March, 1876.
No member of the family has been the subject of more ambiguity or of a greater
expenditure of time and labour than this form, a result perhaps partly due to the
variability of its characters and the fineness of the distinctions between it and
neighbouring species. It agrees with the northern Thelepus cincinnatus, Fabr., in having
two pairs of simple branchiæ and in occasionally presenting the minute papillæ at the
dorsal end of the hook-rows 4—7, but it shows an additional tooth above the main fang of
the hook, and the base is somewhat shorter and broader, the stud above the prow projecting
nearly in a line with the trend of the anterior outline, whilst the stud in T. cincinnatus
makes a smaller angle with the anterior outline. The process beneath the stud is generally
larger. T. cincinnatus, var. andreanse, is distinguished from T. triserialis, G-rube, a southern
form characteristic of the Channel Islands and the Mediterranean, by the occurrence of
three sets of gills in the southern type, by its hooks having a longer base and often only
a single tooth above the main fang, though a second is occasionally seen. Streblo-
soma Bairdi differs in having much longer setigerous processes, three pairs of gills,
hooks which have the stud set at a different angle, and with a prolongation of the anterior
outline in a line with that above the stud. Two conspicuous teeth occur above the
main fang. From Thelepus. setosus, De Quatrefages, T. cincinnatus, var. andreanoe, is
separated by its two pairs of gills, and by the commencement of its hook-rows on the
fifth segment, whereas those of T. setosus begin on the third segment.
Yet after all these distinctions the feeling remains that the precise degree of
variability in each has never been fully studied, and that the environment of each
may be the cause of its special divergencies. All that can be done at present is to
define such variations around each centre or supposed species. The variations of the
hooks of the widely distributed Thelepus setosus, De Quatrefages, as recently shown by
Prof. Fauvel1 (1916), afford an excellent argument for caution in dealing with forms so
intricate. Future work may show that the northern Thelepus cincinnatus of Fabricius
may be the centre of the various forms hitherto described as species, for the presence or
absence of a pair of gills, or the curves and processes of a hook or even the modifications
of the nephridia are not necessarily in themselves features of fundamental value.
Von Marenzeller (1884) mentions that he has found, as Grube previously noted,
specimens in which the bristles did not reach the posterior end. He also observes
that Phenacia ambigrada■ and P. rétrograda of Claparède refer to young examples.
He appears to group all the varieties under this species. In all probability the author
refers to the same form as here described, both from the structure of the hooks and
other particulars.
Crawshay (1912) gives an account of the structure of the hooks and their variations
at Plymouth, and he probably refers to this form, which differs from the northern
T. ci/ndmiatus.
2. T helepus triserialis, Grube, 1885. Plate CXXV a, figs. 7—7 b—bristles and hook
Specific Characters.—General aspect of the cephalic region and body similar to that
of T. cincinnatus, but three pairs of branchiae occur, viz. on the second, third and fourth
1 ‘ Arch. Zool. Exper./ t. lv, p. 465, text-figs. 3—6.