ECHINONEUS CYCLOSTOMUS. 43 in the sand and are found only by sifting the sand. A flour scoop proved the best implement for digging up the sand, and then passing it through a set of sieves, the lowest having a very fine mesh. By the use of such sieves some small specimens were obtained which it would have been practically impossible to have obtained in any other way. In rare instances, specimens were found clinging to the rock overturned, but as they were clinging by the abactinal tubefeet, it is evident they were only accidentally pulled up with the rock. Careful examination of the sand around the rock-fragments (not beneath them) and in areas not under rocks always failed to reveal Echinonéus or Brissus. It seems as though rock-fragment were the essential factor in determining the exact habitat. This may however be true in very shallow water only, for the Litchfield sand-flat is covered at high tide by only 1-3 feet’ of water and at low tide much of it is bare.
‘As in many regular Echini, the tube-feet of Echinonéus are used to hold sand-grains and particles of dirt close against the test, thus more or less concealing the animal.
“‘For this reason specimens only a few millimeters long are very hard to see, the large flat grains of sand (derived apparently from calcareous Algae) affording an admirable means of concealment. When placed in a glass dish of seawater, without any sand, Echinonéus stretches the abactinal tube-feet out far beyond the spines as though hunting for something, but after a short time, they are retracted below the spines. If sand is present the animal soon covers itself. In or on a sufficient amount of sand, the movement of the animal is very evident. Both spines and tube-feet appear to be used in locomotion. “The color of Echinonéus is whitish with a more or less marked reddish tint. One small specimen was found the whole abactinal surface of which was deep red. The tube-feet are clear, translucent red (more rose than scarlet, yet not a rose-red), in marked contrast to the test and spines, so that the ambulacra stand out sharply. The color seems to be very easily affected by reagents, becoming much darker than normal in alcohol, or glycerine. Formalin turns the red shades to yellowish green ultimately becoming greenish yellow. A strong salt solution (NaCl in fresh-water) kept the colors fairly well for a short time (2448 hours) but specimens so treated afterwards became dingy brownish. Mixtures of aleohol and formalin and of alcohol, formalin and glycerine were of no avail, whether made up with fresh- or sea-water. The specimens which are now nearest natural color are certain young ones treated only with pure alcohol or pure glycerine.
+t TROPICAL PACIFIC ECHINI.
“Tn size, Echinonéus rarely exceeds an inch in length. The specimens taken range from 4 to 27 mm. Most of them are between 8 and 18. The form of the test is very variable, the proportions of width to length, height to width, ete. showing much diversity. (In 1897, a bare test was found 42 mm. long.) “Tt is quite possible that large specimens occur in deeper water, outside the reef, but this is simply a conjecture.
“The absence of specimens less than 4 mm. long is greatly to be regretted, but it seems almost certain that such young ones must occupy a different habitat from the adults. Since it is probable that there is a free swimming larval stage, it is quite possible that the absence of specimens under 3 mm. in length from the sand where the adults occur may be explained as follows:— after the metamorphosis, or at least at the conclusion of the free-swimming stage, the young Echinonéus settles on the calcareous Algae, eel-grass, or other vegetation surrounding undisturbed fragments of coral rock on the sand-flat, and only slowly moves downward to and into the sand beneath the rock. By the time the sand is reached, the Echinonéus is 2 or 3 mm. in length.” Echinoneus cyclostomus Leske.
THE OUTLINE OF THE AMBITUS.
Plates 4, 5.
With material limited to a few specimens the specific characters seem easily distinguished, but with an abundance of specimens the supposed differences are at least doubtful. Pls. 4 and 5, give a far better idea of the characteristics of the test than any description.
The figure of a fossil E. michaleti given by Cotteau and of de Loriol’s E. abnormalis from Mauritius, are exactly identical in outline with the one shown (Pl. 5, fig. 4) from Mauritius, and also with that collected in 1868 by Count Pourtalés at Caryfort Reef, Florida (PI. 5, fig. 17) all of which have an almost circular ambitus; the anal system in the smallest Florida specimen being much longer than in the other.
In comparing two specimens of nearly equal length from the same locality, Lord Lowe Island, N.S. Wales (PI. 5, figs. 1, 2), it will be seen that the ambitus of the smallest, fig. 1, is between an ovoid and a pentagon in shape, and that the anal system is elongated, while in the other, fig. 2, the ambitus is unevenly pointed at the anterior end with a longer distance between the anterior ambulacrum on the left side and the anal system is short.
ECHINONEUS CYCLOSTOMUS. 45 The ambitus in a specimen from Mauritius, Pl. 5, fig. 4, is circular, while in two adult specimens, also collected at Mauritius in 1873, only one of which is figured, Pl. 5, fig. 3, it is more ovoid in shape with a broad blunt posterior end; and the well-defined tubercles, the raised and rounded ambulacra, show little similarity to any of the other figures on Pls. 4 and 5. Diversity in the shape of the ambitus is also shown on PI. 4, figs. 7, 20, and Pl. 5, figs. 7-10; all the specimens figured are from the Pacific and by the same collector. On PI. 4, fig. 7, the ambitus is an ellipsoid and the mouth and anus are small; the outline of the test is compressed to an ellipsoid with a posterior pointed end, and the anal system is elongated, nearly reaching the ambitus, (Pl. 4, fig. 20). The same irregularity of the ambitus is shown (PI. 5, figs. 7-10), and the actinostome and anal system varies in form and size out of proportion to the length of the tests. These specimens were water-worn, the tubercles being partly destroyed.
A specimen from Makemo Island, Paumotu Islands, Exped. ‘‘ Albatross” 1899-1900, presents an actinostome and anal system toward the right side of the test; in other specimens the odd anterior‘ambulacrum also frequently inclines toward either the right or the left side.
The larger of two specimens from Bourbon (Pl. 4, figs. 3, 11), recalls the specimen from Lord Howe Island (PI. 5, fig. 2), and the smaller one is identical with that from Makemo Island. Three tests from the Gilbert Islands (Pl. 4, figs. 12, 15, 18), of nearly the same size, are unlike in shape but are in every way similar to the young specimen from Mujeres Island, Yucatan (PI. 4, fig. 19). Comparison of two specimens of about the same size from Cuba, one (PI. 4, - fig. 23), ellipsoid, and one (Pl. 5, fig. 17), with long ellipsoid ambitus nearly one third longer than its breadth, shows little or no conformity to two individuals (Pl. 5, figs. 15, 16), from Port Antonio, Jamaica, the larger of which is ovoid in outline, the smaller ellipsoid with slightly projecting ambulacra, and large actinostome. Of three young specimens also from Port Antonio, the two small- _ est (Pl. 5, figs. 12, 13), are ovoid in outline, while the third (PI. 5, fig. 14) is ellipsoid with a slight deviation in character towards a young specimen of the same size from Mauritius (PI. 5, fig. 5).
Profile views of the tests strengthen the opinion that variation in shape is not wholly due to disparity of age or locality. In two adult specimens of the same length, one from Barbados (PI. 2, fig. 3), the other from the Pacific (PI. 3, fig. 5), the shape of the profiles are exactly alike, with equally curved actinal and abactinal outlines. A third specimen (PI. 4, fig. 9), from Makemo Island
46 TROPICAL PACIFIC ECHINI.
has the actinal surface on the anterior end rounded outwards, with the abactinal side in the apical region depressed. It is not unusual in large specimens to find the same features as are shown in a specimen from Gilbert Islands (PI. 4, fig. 13), with its actinal side perceptibly arched, or as in specimens from the Pacific (Pl. 4, fig. 21), and Cuba, (PI. 4, fig. 24), with straight tests and narrow posterior ends. The test seen from the anterior end facing the odd anterior ambulacrum (Pl. 4, figs. 5, 10, 14, 17, 22), varies in outline from ovoid to spherical and frequently, as in Pl. 4, fig. 10, one side is higher than the other. DEVELOPMENT OF THE CORONAL PLATES.
Plates6,7,13.
The arrangement of the primordial ambulacral plates in each ambulacrum is the same as in the spatangoids, and follows the formula given in Lovén’s Etudes; usually the sueceeding two pairs or so are simple plates after which they are in groups of three of which the upper is the largest. This plate is somewhat broader at the median ambulacral suture; the second one, a demiplate, is wedged in between the first and third, the latter being quite low. In the adult these plates, (Pl. 13, fig. 5), on the abactinal side are very low and about four times as broad as they are high. Each plate, except the largest one, has a double perforated pore, in the edge of the actinostome (PI. 13, fig. 3); the largest plate has three perforations, the one towards the oral plastron being open. In the young from Port Antonio, with the test 5.90 mm. in length, the shape of the actinostome is nearly circular; there is a primordial plate in each interambulacrum (PI. 6, fig. 1); and the same condition obtains in a specimen from Makemo Island (Pl. 13, fig. 1), the test of which measures 4.40 mm. in length.
In the adult the shape becomes a triangle with its horizontal side limiting the bivium, and the vertical side running between the right anterior interambulacrum and the right posterior ambulacrum; its anterior and left posterior corners usurp a part, or the whole, of the primordial plates in the right anterior and the left posterior interambulacra, cf. Pl. 7, a specimen from Mauritius, with test 37 mm. in length and also my earlier sketches in Lovén’s Etudes (Pl. 9, figs, 82, 83). An adult specimen, 27 mm. in length, from Makemo Island shows (PI. 13, fig. 2) the way this reabsorption varies; all the primordial plates are still in place and but little diminished in height.
ECHINONEUS CYCLOSTOMUS. 47 So far as I know the labium in the spatangoids undergoes little or no change in form, while in Echinonéus, even from a single locality, there is much variety; for instance in two young specimens from Port Antonio, 4.19 mm. (PI. 6, fig. 2), and 5.90 mm. in length (PI. 6, fig. 1), the labium is almost three cornered and shut out from the anal system by the first pair of interambulacral plates; while in one, 6 mm. long (PI. 6, fig. 3), the broad base of the labium borders on the edge of the anal system just as in a specimen, 4.40 mm. in length, from Makemo Island (Pl. 13, fig. 7). In two adult tests of practically the same size, 22, 25 mm. in length, from Port Antonio (Pl. 6, figs. 5, 6), the labium of the former (fig. 5) is T-shaped with its smaller end on the anal system, while in the latter (fig. 6), only asmall strip in the edge of the actinostome remains, the same as in a specimen of about the same size, 27 mm. in length, from Makemo Island (PI. 6, fig.8).
The actinal interambulacral plates are nearly as high as broad. At the ambitus where the test is ellipsoid they are higher than when it is ovoid, and continue thus on the abactinal side.
The peculiar way in which the plates overlap in the median interambulacral suture is not fully shown in Pl. 13, fig. 4. The apparent continuation of the ambulacra on to the interambulacral plates (Pl. 13, fig. 5), may be simply impressions due to the attachments of soft parts.
ORAL AND ANAL SYSTEMS.
Plate 15.
Among the specimens collected at Port Antonio, the smallest measures 3.70 mm. in length; the oral plastron is devoid of plates, its actinostome, measuring 1.18 mm. in diameter, and is nearly circular, becoming triangular by growth as described in the development of the coronal plates. When-further developed and adult the oral plastron is covered with solid, glossy plates without tubercles, imperfect polygons or rectangles in form. They are laid in irregular curves (Pl. 15, figs. 2, 3, 4, 6), and could serve as an orbicular shutter; their function is possibly the same as that of the teeth.
Upon comparing the shape of the anal system in a number of specimens from the West Indian and Pacific regions I find an indefinite irregularity in the outline; as for instance two tests from Lord Howe Island and of same size (Pl. 15, figs. 3, 4), in the first (fig. 3), the opening is large and irregularly ovoid, while in the other (fig. 4), it is noticeably shorter, with its angular base or aboral end
48 TROPICAL PACIFIC ECHINI.
towards the ambitus. In an adult specimen, with a test 27 mm. in length, from Makemo Island (fig. 2), the anal system is elongated, ellipsoid, with pointed adoral and aboral ends; and in two other specimens (figs. 5, 6), 27 mm. long, from Mauritius, the opening in fig. 5 is broader with the aboral end pointed, while in fig. 6 it is more symmetrically ellipsoid. Two young specimens from Port Antonio, tests 6 mm. and 10 mm. in length, show the same diversity; one (fig. 8), is smaller and an ellipsoid; the other (fig. 9), with its aboral end running into a sharp point, has little or no similarity to a specimen from Florida (fig. 11), which also differs from an adult individual from Barbados (fig. 12); in the latter the opening of the anal system recalls that of the specimen from Mauritius (fig. 5). In a specimen 29 mm. long from Bourbon (fig. 7), the outline with its broad rounded aboral end, has no similarity to any of the other figures given on Pl. 15. This change in outline and size of the anal system, which is probably caused by the amount of resorption in the first six or seven pairs of interambulacral plates, makes the anal system valueless as a specific character.
The order in which the plates are arranged on the buccal and anal plastrons in other families of the spatangoids has hitherto been regarded as irregular and of slight consequence. The figures given on Pl. 15 of the anal plastron show a remarkable regularity in the grouping of the plates; about seven to nine of. which are laid on the plastron adjoining the interambulacral plates with the largest ones in the adoral end slightly raised. The aboral end has three or more large plates with the addition of a number of small ones surrounding the anal outlet. In the specimens from the Pacific the largest plates on the plastron towards the adoral end, are separated at the median line either by a large plate, or a double row of rhomboidal ones, while in the West Indian specimens these plates are replaced by a number of long slender calcareous strips, which make the surface flexible. This is the only differential point I find between the Pacific and West Indian specimens, and it accords with Mr. Agassiz’s statement, Revision, p. 334-335.
ECHINONEUS CYCLOSTOMUS. 49 THe ABACTINAL SYSTEM.
Plate 14.
At first it was intended to limit this report to a few plates, and a more careful examination of the abactinal system was left for the future. In a young specimen from Makemo, 4.40 mm. in length, interior view (PI. 14, fig. 2), none of the genital plates have received their pore, excepting a few madreporic perforations on the madreporite. Of the two posterior genital plates the one towards the right posterior interambulacrum is slightly larger than its mate; both are about as high as broad and in contact with the ocular plates in the bivium; of the two anterior ones the madreporic is the larger. Seen from the exterior (fig. 7), the apex is slightly raised, carries a few miliaries and a small number of the glassy tubercles.
Another specimen 4.19 mm. in length, from Port Antonio (Pl. 14, fig. 3), differs from the previous one only in the left anterior genital being slender, and the right posterior genital having pushed itself between the ocular plates of the bivium, which in fig. 2 are in apposition.
Of two specimens, one 12.5 mm. long from Port Antonio (Pl. 14, fig. 4), the other 27 mm. in length (PI. 14, fig. 7) from Makemo, the genital plates correspond exactly to each other in form and size; the anterior plate being longer than the posterior which is the reverse of what is shown in figs. 2 and 3; here the two large plates are posterior, with the larger one in the right posterior interambulacrum, while in figs. 4 and 5 these are in the left posterior interambulacrum. The system of the four genital plates are arranged in a square and are in contact; their form and that of the pentagonal shaped ocular plates are very similar to the figure of Hemiaster expergitus in Lovén’s Etudes (PI. 11, fig. 93), and of Periaster tenuis in Agassiz’s Panamic deep sea Echini (p. 210). In two specimens (PI. 14, figs. 5, 6), from the Pacific, measuring 26, 27 mm. in length, one (fig. 5) has very small genital pores, about one third the size of the other (fig. 6); this may be due to sex; the glassy tubercles are larger, darker, and more numerous in fig. 5 than in fig. 6. This variation is found in alcoholic as well as in dry specimens.
50 TROPICAL PACIFIC ECHINI.
THe SPHAERIDIA.
Plates16,17.
Thirty-three figures of the sphaeridia from tests of intermediate sizes varying from 10 to 34 mm. in length, and from various localities in the West Indian and Pacific regions, are given on Pls. 16 and 17. No two are exactly alike, neither do they show any specific characteristics; but some of them correspond in form to those of Maretia or Brissopsis, ¢f. Lovén’s Etudes, Pl. 1. An examination of the sphaeridia in Echinonéus shows their height to be greater in proportion to the size of the test in the smaller specimens than in those full grown. The largest sphaeridium of a young specimen, from Makemo Island, 10 mm. long, is 0.10 mm. in height; in another from Fakarava Island, 12 mm. in length, it is 0.13 mm., and a third from Barbados, with the test 26 mm. in length, has the highest 0.20 mm., while an adult specimen from Lord Howe Island 34 mm. in length or three times as long as the youngest from Makemo Island, has its sphaeridia only 0.16 mm. in height.
Among the material at the Museum I found only three sphaeridia to each ambulacrum, although Lovén’s Etudes (Pl. 9, fig. 79), gives four. There are searcely two alike from the same ambulacrum; some have an uneven papillate surface, but commonly they are almost smooth. The row of sphaeridia (PI. 16, figs. 1-4), taken from odd anterior, right anterior, left anterior, and left posterior ambulacra respectively of a specimen from Fakarava Island, 12 mm. in length, vary in form with no indications of specific characters. Of six sphaeridia (Pl. 16, figs. 5-8), from two specimens from Lord Howe Island, 30 and 34 mm. in length, two (fig. 5) from the odd anterior ambulacrum are 0.16 mm. high, one of them being spherical with a wavy outline, the other smooth and pointed at the top; one from the left anterior ambulacrum (fig. 6) is depressed with flattened top; another from the left posterior ambulacrum (fig. 7), has a spherical outline, slightly truncated at the upper end. Two others (fig. 8), from the right anterior ambulacrum of a specimen measuring 30 mm. are 0.18 mm. high; one is unevenly cylindrical in outline, the other ovoid. This irregularity of form (Pl. 16, figs. 13, 14), can be traced even in the young test measuring 10 mm. in length. In ten figures from the different ambulacra (Pl. 17, figs. 1-6), of two specimens, from Barbados, measuring 26 and 30 mm. in length, the height of the sphaeridia varies from 0.17 mm. to 0.20 mm. This variation which exists between the sphaeridia is further accentuated as shown in fig. 1 where the sphaeridium from the right posterior ambulacrum is ovoid with two points at the top, its companion being ellipsoid with a sharp upper end. There is still another form,