they never, at any time, participated in the rhythmic swing or feeding movements of these last.

Gentle pressure on the anus or the sexual organs excites or inhibits the swimmerets, according as they are already at rest or in motion, and leads, where possible, to a folding of the abdomen. The feeding and preening movements are also, as a rule, brought to a complete standstill by slight irritation of the anus, the after movements being in all cases more violent. So long as the nervous connexion with the tail-fin remained intact, the swimmerets can be excited to considerable activity by touching this region, but when this connexion is destroyed, it is with difficulty they are made to move at all.

The experiments, of which the above is a brief and preliminary account, were carried on at the Physiological Laboratory, Cambridge.

II. "Preliminary Report upon the Comatula of the 'Challenger' Expedition." By P. HERBERT CARPENTER, M.A., Assistant Master at Eton College. Communicated by Sir WYVILLE THOMSON, F.R.S. Received February 18, 1879. Published by permission of the Lords Commissioners of the Treasury. The collection of Comatulæ made by the staff of the " Challenger" includes specimens from 45 different localities, but few of which are deep-water stations. Comatule were only obtained seven times from depths exceeding 1,000 fathoms, namely at:

At lesser depths, 200-1,000 fathoms, Comatule were met with at 13 stations; but by far the greatest number both of species and of individuals were dredged at depths much less than 200 fathoms, and often less than 20 fathoms, at 26 widely distant stations.

At the present time I regard the collection as containing 111 species, mostly new; but as the work of examination and description progresses, it is not unlikely that forms which I now consider different may turn out to be merely local varieties of one and the same species, so that the number given above may be subject to alteration.

Of these 111 species, 59 belong to the genus Antedon, 48 to Actinometra, 1 to Ophiocrinus, and 3, which are peculiar in having ten rays to the calyx instead of only five, to a new genus for which I propose the name Promachocrinus (popaxos, "Challenger.") Challenger.") It may be thought that this peculiarity is hardly a sufficient reason for the erection of a new genus to receive these three species. It is, however, a much more striking one than that on which the genus Ophiocrinus is based, viz., the presence of five arms only, as the rays, unlike those of most Comatulæ, do not divide but bear the arms directly. In Promachocrinus on the other hand, there are ten distinct rays, the radial pentagon which is in contact with the centrodorsal consisting of ten separate pieces, and not of five only, as in Ophiocrinus and in the other Comatulæ.

In two of the species the rays are undivided as in Ophiocrinus; but in the third they divide, as in our common Antedon rosacea, so that there are twenty arms.

This character, the presence of ten rays, is evidently not an accidental one, like the existence of more or less than five rays in other Comatule and in Rhizocrinus. In the latter genus individuals with four to six rays are common, and cases of seven, though rare, may occur. Among the Comatulæ, however, it is very different. I have carefully examined three large Comatula collections besides that of the "Challenger," viz., those of the British and Paris Museums, and Professor Semper's collection from the Philippine Islands. Out of the nearly 200 species contained in these collections I have found but two specimens in which there are not five rays in the calyx. In one of these there are only four, and in the other six rays, though in other individuals of each species there are five, the normal number. The distribution of Promachocrinus is as follows:

P. Kerguelensis (20 arms). Balfour Bay, Kerguelen, 20—60 fathoms.

Ophiocrinus was obtained at four localities at depths varying from 565 to 1,070 fathoms, two in the South Pacific, off South Australia and New Zealand respectively, and two in the North Pacific, one off Japan, and one just north of the Philippine Islands. All the specimens belong to one species, which is by no means so slender and graceful as Semper's Philippine species from shallower water, but has a much more massive arm skeleton.

Among the numerous species of Antedon (59) and Actinometra (48)

the only species which I have been able to identify with any certainty

Müller's specific diagnoses are, as is well known, very incomplete ; and it is possible that a personal examination of his original specimens will enable me to identify more of his species than I can at present. I am inclined to think that besides the above-mentioned species, the "Challenger" collection also includes the following:

Act. purpurea.

Act. rotalaria.

Act. Wahlberghii.
Act. stellata (Lütk.).

The comparative distribution of these two genera is very striking. Relatively speaking, Actinometra is extremely limited in its range, both geographical and bathymetrical. It is almost exclusively a tropical genus, its northern limit being about 30° N. lat. and its southern 40° S. lat. Isolated species are known from the Cape of Good Hope, Natal, South Australia, and Port Jackson, but its chief home is Oceania, especially the Philippines and Moluccas, from which latter locality the "Challenger" brought home 11 species of Actinometra, but not a single Antedon. 14 species were found at Zamboanga, in the Philippines, but no Antedon; while at the Zebu Reefs, in another part of this group, two Antedons were obtained, but no Actinometra; and at Station 192, 11 Antedons, but no Actinometra, just the reverse of what was found at Banda, in the Moluccas. A few Actinometra species are also known from the west coasts of the Atlantic, as South Carolina, the West Indies, Bahia, and St. Paul's Rocks.

The bathymetrical range of Actinometra is likewise very narrow. Nearly all the "Challenger" species are from depths less than 20 fathoms, while only three come from a greater depth than 100 fathoms. These were all obtained at Station 174, where the depths of different hauls were 210, 255, and 610 fathoms. I have no information as to which of these hauls yielded the three species in question. The individual species of Actinometra, like the genus itself, are very local in their distribution. Act. solaris seems to have a fairly wide range in the Malay Archipelago and in Oceania, though oddly enough it does not occur in the "Challenger" collection. Each of the fortyeight species of this collection has its own locality. In no case have I been able to refer specimens from different localities to the same species, except that duplicates of the same species were found at two stations in Torres Straits (186, 187), very close to each other.

With Antedon, however, the case is different. Not only do nearly all the deep-sea Comatula belong to this genus, but some species of it

have a fairly wide range. Ant. rosacea ranges from the north of

Scotland to the Mediterranean, while Ant. Eschrichtii is found over a much wider area. It is the common Arctic species, having been obtained by our own expedition under Nares, as far north as lat. 81° N., while the expeditions of Sweden, Norway, and other countries have found it abundant in the seas of Spitzbergen and Nova Zembla. It is well known on the American coast, and was dredged by the "Challenger" off Halifax, while the "Porcupine" met with it in the "cold area of the North Atlantic.

The Challenger" dredgings round Heard Island yielded several specimens which agree so very closely with Ant. Eschrichtii that I am very strongly inclined to believe in the identity of the southern and northern forms. There are, however, some minor points of difference between them, and the southern form may really turn out to be the representative species of Ant. Eschrichtii, but not identical with it. I cannot venture to give a definite opinion upon this point until I have had an opportunity of examining a greater variety of specimens than are accessible to me just at present.

There are other Antedon species, which occur in duplicate from different localities. Two specimens from near the Kermadec Islands (S. 170), also occur in the neighbourhood of the Fijis (S. 174, 175). A third species was dredged at Stations 147 and 160, two localities in the Southern Sea, in nearly the same latitude, but separated by almost 90° of longitude. A fourth species came up from 1,070 and 775 fathoms, off the Admiralty Islands and Japan respectively.

The above facts would seem to show that, with few exceptions, the geographical range of the individual members of the family Comatulidæ, is exceedingly limited, nearly every species having its own locality, and that not a very extensive one.

This is not surprising when it is remembered how rarely Comatulæ have been found at great depths. The stalked Crinoids, on the other hand, are especially characteristic of the abyssal fauna, Pentacrinus, Bathycrinus, and Rhizocrinus, all having a very wide distribution. This is true, also, even with the individual species of the latter genus. This accords well with our paleontological knowledge. Chalk Comatulæ are exceedingly rare. Hagenow found one in Germany, which he named Hertha mystica. From the figure which he gives of its calyx, I should judge it to be an Antedon, which agrees well with the facts stated above. Lundgren* has found a calyx in the chalk of Sweden, which " comes very near to Antedon Fischeri, Geinitz." There

are also a few chalk Comatulæ in the Woodwardian and British Museums, viz., Glenotremites and similar forms, but they are as nothing compared to the remains of Pentacrinus and Bourguetticrinus, and even

* "Neues Jahrbuch für Mineralogie." Heft ii, 1876, pp. 180-182.