The whole structure will at first naturally be thought to be an external skeleton, like the true shell of a mussel,* or the hard investment of a lobster.† It is, however, nothing of the kind, but is a truly internal skeleton, inasmuch as the soft substance of the animal (the perisoma) coats it externally as well as internally.

The numerous, distinct, calcified plates which form the shell are arrayed in a very definite order. Of the two poles the inferior one is termed the "oral pole," as it is there that the mouth is to be found. The superior one is called the "apical pole," and this is the situation of the anus. Around the anus are placed some small and more or less irregular plates, termed "anal plates." External to and immediately below the anal plates are ten large ones, each perforated, and forming together a single circular row. These ten plates are the "ocular” and "genital" plates, and they alternate with each other so that each ocular plate is between two genital plates, and each genital plate is between two ocular ones. The perforation in each genital plate is the external opening of a genital duct. The perforation in each ocular plate is for an eye-spot. Below this circle of ten plates is the great bulk of the shell, which is called the "corona," and consists of five vertical tracts beset with perforations and termed the ambulacra. Alternating with these are five other tracts of greater width, destitute of pores and termed the interambulacra. Each ambulacrum and each interambulacrum is narrower at its apical than at its oral end, and consists of two vertical series of plates. Each of these plates is pentagonal, while two angles of each pentagon are right angles. A zigzag suture joins together the two vertical series of ambulacral plates in each ambulacrum, and a similar suture unites together the two vertical series of interambulacral plates in each interambulacrum. On the other hand, a straight suture unites the flat margins of each series of ambulacral plates with the flat margins of the adjacent series of interambulacral plates. The plates which form the ambulacrum are seen, when closely examined, to consist each again of three pieces (pore-plates), which form between them the six pores, or foramina, of each ambulacral plate.

The interambulacral plates, on the other hand, are single, and without perforations.

There are never more than two vertical series of plates in each ambulacrum and in each interambulacrum. Therefore there are five pairs of rows of ambulacral plates and five interspaces occupied by five pairs of rows of interambulacral plates.

See the article on "The Anatomy of the River-Mussel," by Mr. J. C. Galton, POPULAR SCIENCE REVIEW for July 1870.

† See the article on the Lobster in the same work for October 1838.

Thus there are altogether twenty vertical (or longitudinal) series of plates, ten of these being ambulacral and ten being interambulacral.

The relation of these alternating tracts to the circle of ocular and genital plates is such that an ocular plate is placed at the summit of each ambulacrum, while a genital plate stands at the top of each interambulacrum.

One of the five genital plates is larger than the rest, and has a porous space, which gives a worm-eaten look to its surface. This modified genital plate is termed the madreporic tubercle, and is of course, like the other genital plates, interambulacral in position, or, as it is sometimes called, inter-radial.

At the opposite, or oral, pole (beyond the corona formed of the ambulacra and interambulacra) are the buccal plates, which are small and irregular and are scattered in the buccal membrane which surrounds the oral aperture.

All these plates forming this singularly composite shell consist each of a network of calcareous spicula formed in the soft perisoma, and they thus have the same relation to the flesh of the echinus that the skeleton of a coral has to its investing body. The spicula meet and unite at right angles, and often form very definite patterns.

In the living state the shell is covered with spines, each spine being articulated to one of the tubercles before spoken of. The articulation is effected, at least in the larger spines, by means of a ligament which passes from a little pit on the surface of the projecting tubercle to another little pit situated on the convex lower end, or base, of the spine, like the ligamentum teres of our own hip joint. Thus great freedom of motion is allowed, while the motions themselves are effected by means of small muscles which extend from the shell around the tubercle to the projecting parts of the base of the spine. These muscles are, of course, invested by the external layer of the soft perisoma of the body. The spines are important organs of locomotion, and are the agents by which burrowing in the sand is effected. Locomotion, however, is also aided by tubular suckers, which abound in each ambulacrum. Each sucker springs from two pores, a minute tubular vessel passing out through cach pore, and then each pair of such minute vessels joining together to form one sucker. On the other hand, the two minute vessels unite again within the shell, to form a saccular dilatation, which is directly continuous with the ambulacral system of vessels to be presently described.

Each sucker is mostly cylindrical, but is flattened and enlarged at its free end into a disk. This disk even is furnished with five flattened calcareous pieces, of about equal size, together forming the rosette, and the stalk of the sucker is often beset

with spicula. These suckers, being very extensible, can be protruded and withdrawn ; and by their aid locomotion is indeed mainly effected.

There is a third and most singular group of organs scattered over the whole surface of the buccal membrane and corona. They are of minute size, so that to observe them requires a handglass, and they were formerly taken to be parasites. They are termed pedicellaria, and are complex organs, being even each furnished with a delicate internal skeleton. Each consists of a long slender stalk, ending in three short limbs (or jaws), which diverge from the distal end of the stalk. These jaws open and shut with a snapping action, while the stalk sways about.

The utility of these organs is not yet fully determined, but they have been observed to remove particles of excrement, passing them along, as it were, from hand to hand. They may, however, also pass along nutritious particles to the mouth, and they may serve to detach foreign bodies, ova, &c., which, if allowed to remain, would grow up parasitically and injure the echinus. They singularly remind us of the bird's-head processes of certain polyzoa or bryozoa (polyp-like animals, with ciliated tentacles).

On the inside of the corona, at the lower end of each ambulacrum, is a solid calcareous arch, termed an auricula, while within the space bounded by these five auriculæ is that singularly complex masticating organ termed "Aristotle's lantern," made up of a score of distinct parts. When, in the unmutilated animal, the mouth is looked at, the apices of five pointed teeth are seen to protrude more or less.

Each of these teeth is externally very like a cutting tooth. of a rat or squirrel, though much simpler in minute structure. It is constantly reproduced from a soft root at the upper end (where it is enclosed in a secreting bag) as it is worn away at the apex.

Each tooth is enclosed in a composite vertical plate, folded on itself and interambulacral in position. These are the alveoli. The adjacent upper edges of the alveoli are held together by strong horizontal pieces, which pass from one alveolus to the next, are oblong in shape, and are called the rotula. Each rotula is opposite an auricular arch.

The last set of pieces are the five radii, which are long slender parts fixed to the apices of the rotulæ like springs, extending upwards and outwards, and bifurcating at their ends.

All these complex parts act together, and separate and ap

"Composite" because it consists, primitively, of four parts.