is

•

,

/5 3 ?///Remoirs of tbe flDuseum of Comparative

_^<s ti^c/ / » m rr j D-P » TJ n rnr.T. ROW.

AT HARVARD COLLEGE.
VOL. XXXIV. No. 4.

A*

HAWAIIAN AND OTHER PACIFIC ECHINI.

THE PEDINID/E, PHYMOSOMATIDJE, STOMOPNEUSTID^E, ECHINIDJE,

TEMNOPLEURID5!, STRONGYLOCENTROTID^!,

AND ECHINOMETRIDJE.

BY

1!
!!

, 3-

: a-

g

Cl

m

HUBERT LYMAN CLARK.

•C

WITH THIRTY-TWO PLATES.

PLATES 90-121.

JfckOv

"^C^^fLi

IPublished by Permission of GEORGE M. BOWERS, U. S. Commissioner of Fish and Fisheries.]

CAMBRIDGE, U. S. A.:

prtnteo for tbe flDuseum.

JUNE, 1912.

J1ATNABD H. MEICJ

flDemotrs of tbe flDupeum ot Comparative

AT HAEVAED COLLEGE.

VOL. XXXIV. No. 4.

HAWAIIAN AND OTHER PACIFIC ECHINI.

THE PEDINID.E, PHYMOSOMATID.E, STOMOPNEUSTIDJE, ECHINID-ffl,

TEMNOPLEURID5!, STRONGYLOCENTROTID^E,

AND ECHINOMETRIDJE.

BY
HUBERT LYMAN CLARK.

WITH THIRTY-TWO PLATES.
PLATES 90-121.

(Published by Permission of GEOHQE M. BOWERS, U. S. Commissioner of Fish and Fisheries.)

CAMBRIDGE, U. S. A.:

IPrlnteo for tbc flDuseum.

JUNE, 1912.

CONTENTS.

No. 4. HAWAIIAN AND OTHER PACIFIC ECHINI. Based upon Collections made
by the U. S. Fish Commission Steamer "Albatross" in 1902, Commander CHAUNCEY
THOMAS, U. S. N., Commanding, and in 1906, Lieut. Commander L. M. GARRETT,
U. S. N., Commanding. THE PEDINID^E, PHYMOSOMATID.E, STOMOPNEUSTID.E,
ECHINID.E, TEMNOPLEURID.E, STRONGYLOCENTROTID.E AND ECHINOMETRID.E. By
HUBERT LYMAN CLARK. 180 pp. 32 plates. June, 1912.

CONTENTS.

PAGE

Pedinidse Gregory 213

General Characteristics 213

The Spines, Pedicellarire, Sphjeridia,

and Spicules 215

Caenopedina A. Ag 216

*Ca?nopedina hawaiiensis, nom.

nov., Plates 90, figs. 1-4; 91,

figs. 1-13; 105, figs. 1-5 . . 217

Ctenopedina mirabilis Mrtsn.,

Plates 91, figs. 14-17; 105,

fig. 8 221

*Ca?nopedina pulchella, comb,
nov., Plates 91, figs. 18-22;
103, figs. 1-3; 105, figs. 6, 7 . 223
Phymosomatidae Meissner .... 225
General Characteristics .... 225
The Spines, Pedicellarioe, Sphseridia,

and Spicules 226

Glyptocidaris A. Ag 228

Glyptocidaris crenularis A. Ag.,
Plates 90, figs. 5-10; 92, figs.
;-//; 106, figs. 1,2 .... 228
Stomopneustidfe Mortensen, Plate 90,

figs. 11,12 229

Echinidfe Agassiz 231

General Considerations 231

The Spines, Pedicellaripe, Sphseridia,

and Spicules 236

The Genera and Species of Recent

Echinidse 238

Psammechinus Agass. and Des. . 242

Lytechinus A. Ag 244

Lytechinus euerces, sp. nov.,
Plates 93, figs. 4, 5; 98, figs.
3,4; 107, figs. 4-6 . ... 247
Lytechinus dyscritus, sp. nov.,
Plates 93, fig. 10; 96, figs. 1-3 249

PAGE
Lytechinus callipeplus, sp. nov.,

Plate 96, figs. 4-6 251

*Lytechinus verruculatus, comb.

nov 253

Lytechinus anamesus, sp. nov.,

Plates 99, figs. 4, 5; 107, figs.

7-11 254

Lytechinus semituberculatus

Verr 258

Lytechinus pictus, comb, nov.,

Plates 99, figs. 6, 7; 107, figs.

12-14 258

Echinus Linne 260

Echinus Wallisii A. Ag., Plates

93, fig. 32; 108, figs. 1, 2 . . 265
Echinus euryporus, sp. nov.,

Plates 93, figs. 2, 3; 109, figs.

4-6 266

Echinus lucidus Dod., Plate 107,

figs. 1-3 267

Echinus atlanticus Mrtsn., Plate

110, figs. 1-3 269

Echinus tylodes, sp. nov., Plates

93, figs. 11-15; 109, figs. 1-3 269
Echinus anchistus, sp. nov.,

Plates 93, figs. 8, 9; 103, figs.

4,5; 111, figs. 1-3 . . . . 271

Parechinus Mrtsn 272

Parechinus magellanicus, comb.

nov 275

Nudechinus, gen. nov., Plate

111, figs. 7, 8 276

Nudechinus scotiopremnus, sp.

nov., Plate. 97, figs. 4-6 . . 277
Nudechinus stictus, sp. nov.,

Plates 93, fig. 1; 97, figs. 1-3 . 279
Evechinus Verr. 281

* Hawaiian species,

210

HAWAIIAN AND OTHER PACIFIC ECHINI.

PAGE

Toxopneustes Agassiz 282

Toxopneustes chloracanthus, sp.

nov., Plate 93, figs. 6', 7 . . 283

Tripneustes Agassiz 285

*Tripneustes gratilla Loven . . 286

Gymnechinus Mrtsn 286

Gymnechinus megaloplax, sp.

nov., Plate 102, figs. 2,3 . . 287
Gymnechinus epistichus, sp.
nov., Plates 93, figs. 22, 23;

102, figs. 4,5 289

Temnopleuridffi Desor 290

General Considerations 290

The Spines, Pedicellarise, Sphseridia,

and Spicules 291

The Genera and Species of Recent

Temnopleuridce 291

Trigonocidaris A. Ag 295

*Trigonoeidaris albida A. Ag. . 295

Hypsiechinus Mrtsn 296

Genocidaris A. Ag 296

Genocidaris apoda A. Ag. and
Cl., Plates 93, figs. 16, 17;

100, figs. 1-3 297

Orechinus Dod 300

*Orechinus monolini Dod., Plate

93, figs. 27-31 300

Lamprechinus Dod 302

Prionechinus A. Ag 302

*Prionechinus Chuni Dod. . . 304
Prionechinus ruber A. Ag. and

Cl./ Plate 100, figs. 4-6 . . 304
*Prionechinus sculptus A. Ag.
and Cl., Plates 93, figs. 24-26;

101, figs. 4-6 306

Prionechinus Agassizii Wood-
Mas, and Ale 308

*Prionechinus depressus A. Ag.

and CL, Plate 101, figs. 1-3 . 309

Opechinus Desor 310

Opechinus variabilis Mrtsn. . . 311

Temnopleurus Agassiz .... 311

Temnopleurus Hardwickii A. Ag. 312
Temnopleurus toreumaticus

Agassiz 312

Temnopleurus Reevesii Mrtsn. 313
Salmacis Agassiz, Plates 93, figs.

33-36; 111, figs. 4-6 ... 314

Temnotrema A. Ag 317

* Hawaiian

PAGE
*Temnotrema hawaiiensis, comb.

nov., Plate 99, figs. 1-3 . . 319
Temnotrema sculpta A. Ag.,

Plate 112, figs. 1,2 .... 321
Mespilia Agass. and Des. . . . 322
Microcyphus Agass. and Des. . . 322

Salmacopsis Dod 324

Salmacopsis olivacea Dod.,
Plates 103, figs. 6, 7; 112,

figs. 3,4 324

Amblypneustes Agassiz .... 325
Amblypneustes pachistus, sp.
nov., Plates 104, fig. 6; 112,
figs. 10, 11; 121, figs. 1-3 . . 327
Amblypneustes grandis, sp. nov.,

Plate 121, figs. 4-6 .... 329
Amblypneustes triseriatus, sp.
nov., Plates 104, fig. 5; 112,

fig. 5 331

Holopneustes Agass. and Des. . . 332
Holopneustes pycnotylus, sp.
nov., Plates 104, fig. 4; 112,

figs. 6-9 334

Goniopneustes Duncan .... 336
Goniopneustes pentagonus Dun-
can, Plate 93, figs. 18-21 . . . 336
Strongylocentrotidae Gregory .... 338

General Considerations 338

The Spines, Pedicellarite, Sphaeridia,

and Spicules 338

The Genera and Species of Strongy-

locentrotid;e 339

Echinostrephus A. Ag., Plate 105,

figs. 0-11 342

*Echinostrephus aciculatus A.
Ag., Plates 95, figs. 23, 24;

105, fig. 9 342

Pseudoboletia Trosch 344

*Pseudoboletia Indiana A. Ag.,
Plate 92, figs. 12-18 ... 345

Paracentrotus Mrtsn 347

Loxechinus Desor 347

Ceenocentrotus, gen. nov., Plate

95, figs. 16,17 348

Pachycentrotus, gen. nov., Plates

94, figs. 1-6; 98, figs. 5-8 . . 349
Heliocidaris Agass. and Des., Plate

94, figs. 7, 8 350

Heliocidaris stenopora, nom.
species.

CONTENTS.

211

PAGE

nov., Plates 95, figs. 18-22;
104, figs. 1-3; 110, figs. 4, 5 . 351
Strongylocentrotus Br., Plate 94,

figs. 9-33 352

Strongylocentrotus fragilis Jack-
son, Plates 94, figs. 28, 29;

113, figs. 3-6 354

Strongylocentrotus pulchellus A.
Ag. and CI., Plates 94, figs.
24-27; 98, figs. 1,2; 111, figs.

9, 10 356

Strongylocentrotus clrobachien-

sis A. Ag 358

Strongylocentrotus echinoides A.
Ag. and Cl., Plates 94, figs.
13-16; 113, fig. 1 .... 360
Strongylocentrotus polyacan-
thus A. Ag. and Cl., Plates 94,
figs. 30-33; 113, fig. 2 ... 362
Strongylocentrotus nudus A.
Ag., Plate 94, figs. 17-23 . . 363

Echinometridffi Gray 365

General Considerations 365

The Spines, Pedicellarite, Sphreridia,

and Spicules 366

The Genera and Species of Echino-

metridse 367

Parasalenia A. Ag 368

Parasalenia gratiosa A. Ag. . . 369

PAGE
Parasalenia Pohlii Pfeff., Plate

95, figs, 1-5 369

Echinometra 370

Echinometra Mathsei Bl. . . 372
*Echinometra picta A. Ag. and
Cl., Plates 95, figs. 6-12; 114,

figs. 5, 6 .373

*Echinometra oblonga BI., Plate

114, figs. 1,2 373

Echinometra insularis, sp. nov.,
Plates 95, figs. 13-15; 114,

figs. 3, 4 374

Echinometra Van Brunti A. Ag. 375

Podophora Agassiz 376

*Podophora atrata Agassiz . . 376
Podophora pedifera Agass. and

Des. . . . • 377

Colobocentrotus Br 377

Heterocentrotus Br 377

*Heterocentrotus mammillatus

Br. Plates 115-117 .... 378
Heterocentrotus trigonarius Br.,

Plates 1 18-120 379

Addenda 381

*Podocidaris ornata, sp. nov.,

Plate 102, fig. 1 381

Yoshiwara's Echini .... 282
Explanation of Plates 385

* Hawaiian species.

HAWAIIAN AND OTHER PACIFIC ECHINI.

COLLECTED BY THE U. S. FISH COMMISSION STEAMER "ALBATROSS," COMMANDER
CHAUNCEY THOMAS, U. S. N., COMMANDING IN 1902, AND LIEUT. COM-
MANDER L. M. GARRETT, U. S. N., COMMANDING IN 1906.

PEDINID^E Gregory.
GENERAL CHARACTERISTICS.

IT is quite remarkable that the recent species of this family represent but a
single genus and, with one exception, are very much alike in all essentials. The
solid spines, the firm test, and the bright colors make their resemblance to the
Echinidse rather striking but the grooved teeth and the perforated tubercles
indicate a nearer relationship to the Centrechinidse.1 On the other hand we find
such peculiarities in the alimentary canal and abactinal system that it is an open
question whether the relationship to the Centrechinidse is really as close as it
seems. Yet it must be admitted that the resemblance of the recent species of
Csenopedina to Centrostephanus, particularly to C. longispinus, is very close
and it is extremely difficult to draw the line so sharply as to warrant their being
placed in separate families. The structure of the ambulacrum, the perforation
of the primary tubercles, the plating of the buccal membrane and the appearance
of the lantern are identical in the two genera, while the abactinal system and the
pedicellarise of Csenopedina are so similar to those of Centrostephanus asteriscus
that no possible generic difference can be found in those characters. Indeed the
only distinct and constant differences are found in the primary spines (which are
ordinarily smooth and solid in Csenopedina, rough and hollow in Centrostephanus)
and in the primary tubercles (which are distinctly crenulated in Centrostephanus
but perfectly smooth in Csenopedina). The smaller spines of Csenopedina are
sometimes hollow however, and the larger primaries are, in C. hawaiiensis at
least, decidedly rough. Were it not for the large number of fossil forms known,
which seem to represent a different group from our modern Centrechinidse, the
family Pedinidss would not be either necessary or desirable, for Csenopedina
could well be referred to the Centrechinidse. But until a more complete revision

1 For the use of this name, see Jackson, 1912, Phylogeny of the Echini. Mem. Boston Soc. Nat.
Hist., VII, p. 27—28. The conclusion that Diadema cannot be used for Echini seems to be unavoidable.

214 HAWAIIAN AND OTHER PACIFIC ECHINI.

of the extinct species reveals the needlessness of the Pedinidae, it may be better
to leave the genus Csenopedina in that family. In any case, it certainly belongs
in the suborder Aulodonta Jackson.

The species of Csenopedina agree in having a large abactinal system, nearly
or quite equal to half the diameter of the test; the ambulacral primary tubercles
are continued well above the ambitus; and secondary tubercles are more or less
numerous. In other particulars they are much like Pseudopedina Cotteau, to
which genus they are perhaps more nearly allied than to the true Hemipedinas.
They cannot properly be referred to Pseudopedina however and therefore
Mortensen's revival of the name Csenopedina, under which the type-species
(cubensis) was originally described, may be accepted, without however agreeing
with him that recent species cannot well be referred to genera known originally
only as fossils.

Having already showed the close resemblance of Caenopedina to Centro-
stephanus, it will be unnecessary to repeat the various structural features of the
genus, but some of the details regarding which little has been published, may be
briefly referred to here.

The most striking feature of the internal anatomy is seen in the brevity of
the upper coil of the intestine, which, unlike that of Echinothrix and the other
Diadematidse, almost wholly lacks the interradial loops and follows a very short
course near the abactinal surface of the test (PI. 90, fig. 2). The lantern and
teeth (PI. 90, figs. 3, 4) are like those of the Centrechinidse in every detail, but the
perignathic girdle is less developed than is the rule in that family; the auricles
are slender and do not quite meet over the ambulacra. Rudimentary Stewart's
organs can be easily distinguished. The characters of the test, and other ex-
ternal features, are so well known we need only mention them here. The
ambulacra are typically centrechinoid, while the primary tubercles of both areas
are equally so. The tubercles are high and without a trace of crenulation and
the areolse are very little or not at all sunken. The abactinal system is very
large, with a big periproct and small ocular plates. There is no trace of an anal
tube. The actinostome is well plated, especially in the ambulacra, as in the
Centrechinidse, and the buccal plates carry spines as in Centrostephanus.

215

THE SPINES, PEDICELLARLE, SPH^RIDIA, AND SPICULES.

Plate 91.

The spines of the Pedinidse are typically smooth and solid. In the recent
species, they form the most obvious distinguishing mark to separate them from
the Centrechinidse. The larger primaries are ordinarily very solid, though they
may be slender; their length usually exceeds the diameter of the test. They are
finely striated, and when pointed (as in C. pulchella), they strongly resemble
those of Echinometra. As a rule however, they are not pointed but very blunt
and they even may be flaring at the tip. Under the magnifying glass, they reveal
the same longitudinal series of minute serrations which is found in many Centre-
chinidse, but ordinarily much smaller than in that family. The smaller primaries
and larger secondaries are often quite hollow, suggesting how little weight ought
to be placed on the characters afforded by the spines.

The pedicellarise, so far as they are known, show a marked resemblance to
those of Centrostephanus. While at least four different kinds occur on most
individuals, either globiferous or tridentate may be quite wanting. It seems to
be true, as in the case of many other Echini, that if globiferous pedicellarise are
abundant, the tridentate are wanting or very rare, while if the latter are abun-
dant, the former seem to be missing or infrequent. The stalks of the pedicel-
larise are much as in the Centrechinidse, those of the tridentate and ophicephalous
rather stout and a little enlarged at the upper end, while those of the globiferous
are very slender, scarcely more than a single rod at the middle, and are abruptly
enlarged where they join the head. The valves of the globiferous pedicellarise
are more or less concealed by the glands, which are often heavily pigmented.
They are quite small and terminate in 2-4 teeth. These teeth may be very long
(PI. 91, fig. 18) or quite short (fig. 17). The tridentate pedicellarise vary greatly
in size and form (PI. 91, figs. 1, 4, 5, 6) but the valves are nearly always narrow
and compressed. The heads range from .20 to 4 mm. in length, while the stalks
may hardly equal them or may be several times as long. The valves are either
straight or curved and may be in contact for most of their length or only just
meet at the tip. The ophicephalous pedicellarise are more or less common and
occur with valves of two quite distinct forms. That one which is really char-
acteristic (PI. 91, figs. 14, ^0) is very greatly constricted at the base of the blade,
while the other is broader (fig. 15) or longer (fig. 7) and much less constricted.
The latter form intergrades with the tridentate pedicellarise and have been

216 HAWAIIAN AND OTHER PACIFIC ECHINI.

called tridentate by Doderlein, but as they usually have a well-marked articular
loop, it seems better to call them ophicephalous. The triphyllous pedicellarise
are small and hard to detect but are usually common. The valves are .10-.20
mm. in length and quite variable in form, though generally rather wide and
truncate at tip (PI. 91, fig. 5). The stalks are, as usual in triphyllous pedicel-
larise, much longer than the neck, which is several times as long as the head.

The sphceridia (PI. 1 , figs. 13, 22) as in the Centrechinidse, are present on the
lower secondary element of the ambulacral plates, at the inner side of the foot,
from the peristome nearly to the ocular plate. They do not exhibit great variety
of form or possess any distinctive characters.

The calcareous spicules (PI. 91, fig. 21) are large, fenestrated plates, of very
diverse forms and sizes. The surface may be smooth or provided with a few,
or with numerous spinelets. Such deposits are obviously very different from
the normally triradiate spicules of the Centrechinidse.

QENOPEDINA.

A. Agassiz, 1869. Bull. M. C. Z., I, p. 256.
Type-species, Ccmopedina cubensis A. Agassiz, 1. c.

Careful comparative study of the material available, and of the descrip-
tions and figures of de Meijere, Doderlein, and Mortensen shows that there
are now known five species of this genus. They are distinguished from each
other most satisfactorily, by the characters found in the abactinal system, the
primary spines, and the coloration. The species of the West Indian region
(cubensis A. Ag.) has been longest known and is the type of the genus; the
specimens before us are all small, but Koehler has been so fortunate as to have
some very fine material from depths of 605-660 fathoms in the eastern North
Atlantic and his beautiful colored figure (1909, Ech. "Princesse Alice," PI. 1, fig.
1) well shows the appearance of this interesting species. The Japanese form
(mirabilis Dod.) was the next to be discovered and the "Albatross" collection
fortunately supplies sufficient material for a satisfactory understanding of that
species. The "Siboga" collected specimens of a Cgenopedina near the Kei
Islands, which was regarded by de Meijere as undescribed and to which he gave
the name indica. Doderlein (1906, "Valdivia" Ech., p. 176) has expressed the
opinion that this is only a variety of mirabilis but I am not able to agree with
him. The two forms seem to be quite distinct. In 1907 (Bull. M. C. Z., L,
p. 245) de Meijere's species was recorded from the Hawaiian Islands but a more

CLENOPEDINA HAWAIIENSIS. 217

careful comparison of these specimens with de Meijere's description and figures
shows that they are really quite a distinct species, which may be called
hawaiiensis. It is interesting and important to note in this connection that, so
far as can be judged from de Meijere's description and figures, indica appears
nearer cubcnsis, than to any other species. Indeed the differences between the
two, which de Meijere mentions are for the most part trivial, but it is possible
that a direct comparison of specimens would show more important distinctions
than those given below. The fifth species of the genus, from the Hawaiian
Islands, is strikingly different from any of the others and because of its fine
coloring well deserves its name, pulchella.

The following table shows the characters by which the five species may be
distinguished.

Anal system covered by numerous minute plates; primary spines long and slender, more

or less cylindrical, thickness less than .05 of length.
Primary spines not banded; genital plates with very few spines or at least, bare at

the centre.
Abactinal region not at all violet; genital plates, each with several secondary and

rather numerous miliary tubercles.

Actinostome distinctly smaller than periproct; few, scattered secondary
tubercles in ambulacra; tridentate pedicellarias common; usually more

or less greenish abactinally cubensis.

Actinostome equals periproct; a double row of secondary tubercles in each
ambulacrum; tridentate pedicellariie seem to be wanting; no green in

coloration indica.

Abactinal region distinctly violet; genital plates quite bare, each with about
8 tubercles chiefly on proximal margin; tridentate pedicellariae abundant and
varied; ophicephalous pedicellariao with markedly constricted valves, wanting hawaiiensis.
Primary spines greenish or very light brown with several broad ill-defined bands of red

or reddish brown; genital plates with more or less numerous secondaries . . . mirabilis.
Anal system covered by few (15-30) plates; primary spines short and stout, tapering to a

blunt point, thickness nearly .10 of length pulchella.

Caenopedina hawaiiensis, nom. nov.

Hemipedina indica A. Agassiz and Clark, 1907, Bull. M. C. Z., L, p. 245. NON Hemipedina
indica de Meijere 1903, Tijds. Ned. Dierk. Vereen., (2), VIII, p. 3.

Plates 90, figs. 1-4; 91, figs. 1-13; 105, figs. 1-5.

The largest of the specimens measures 37 mm. in diameter and the height of
the test is 17 mm. There are 10 interambulacral plates in each column and 15
ambulacral plates in each half-area. The longest primary is 41 mm. in length.
In another specimen, 32 mm. in diameter the longest primary is 50 mm. long

218 HAWAIIAN AND OTHER PACIFIC ECHINI.

and only 1.8 mm. in thickness. In a specimen 30 mm. in diameter, there are
10 interambulacral and 15 ambulacral plates to a column. The abactinal
system is 15 mm. in diameter and the actinostome 11.5 mm. In the smallest
specimen the test is 15 mm. in diameter and nearly 8 high; the abactinal system
measures almost 8 mm. across and the actinostome 6 mm.; there are 8 inter-
ambulacral and 10 or 11 ambulacral plates to a column.

The periproct is large, its diameter about half that of the abactinal system;
it is covered by upwards of five hundred small, rounded plates of which ten or a
dozen near the approximately central anus, are noticeably larger than the others
and carry a few miliary spines and pedicellari&3; there is no indication whatever
of an anal tube. The genital plates are large, broadly in contact with each other
and completely exclude all the oculars from the periproct. The genital pores
show such diversity as to indicate a sexual difference; in some specimens they
are moderately large, circular and near the centre of the plate while in the others
they are very small and occupy a notch in the distal angle of the plate, from
which a narrow, shallow but quite distinct groove runs down nearly or quite
to the ambitus. The madreporic genital is no larger than the others and the
pores occupy only a small area at the centre of the plate. Each plate carries
4 or 5 small spine-bearing tubercles near the periproct and 2 or 3 others may be
scattered on the plate; there are also a number of minute pedicellaria-bearing
granules which are hardly visible to the unaided eye. The ocular plates are
about one third as large as the genitals; the pore is near the distal margin and
proximal to it are 2 or 3 small tubercles.

The interambulacral plates are high but the surface of each is largely covered
by the primary tubercle; at both the outer and inner ends however, there are
secondary and miliary tubercles, which are largest and most numerous on the
actinal plates. The ambulacral plates are nearly as high as they are wide, with
the rather large pore-pairs occupying most of the outer end, while the inner half
is covered by the primary tubercle; there are, however, some small secondaries
beside the median suture especially actinally, and miliaries occur wherever there
is room especially abactinally. The pore-pairs form a nearly vertical series,
though below the ambitus the "arcs of three" are evident. The ambulacra are
half as wide as the interambulacra at the ambitus.

The buccal membrane carries a considerable number of small plates, but these
are chiefly in the ambulacra distal to the primordial ambulacrals. The latter
carry small spines as well as pedicellaris while the little plates carry pedicel-
lariae alone. The gills are relatively small and the cuts are broad and shallow.

(LENOPEDINA HAWAIIENSIS. 219

The primary tubercles are very large, especially at the ambitus. In the
interambulacra, their areolae coalesce vertically for nearly their full width, but
the uppermost and lowest are naturally not so extensively developed. In the
ambulacra, the areolse do not coalesce even at the ambitus but are separated by
a few minute miliaries. In both areas, the boss is moderately high without a
trace of crenulation, while the mamelon is very well formed and conspicuously
perforated. The larger secondary tubercles are also perforated and are other-
wise similar. The primary spines at and above the ambitus are long and slender,
those of the interambulacra greatly exceeding those of the ambulacra. They
are very finely striated longitudinally and very minutely rough. They are solid
but the central core is obviously less solid than the surrounding wall and towards
the tip of the spine becomes very poorly developed. Some of the smallest
primaries are perfectly hollow and the larger secondaries resemble them in that
respect as well as in general appearance.

The pedicellarise are abundant but are mostly tridentate. The ylobiferous
pedicellarise (PI. 91, fig. 10) are nearly or quite confined to the abactinal system
and may be wanting even there. They are quite small but are easily detected
because of the brown or purple glands which enclose the valves. The latter are
about half a millimeter long and terminate in two very short, sharp, diverging
teeth (PI. 91, figs. 11, 12). The tridentate pedicellarise occur in a great variety
of sizes and forms and are found on all parts of the corona, as well as on the
abactinal system and occasionally even on the actinostome. They may have
a short neck or none and the stalk may equal, or exceed more or less markedly,
the head. The valves may be straight and in contact for nearly their entire
length (PI. 91, fig. 4) or in contact only, to a varying amount, near the tip (PI. 91,
figs. 5, 6), or they may be strongly curved and in contact only at tip (PI. 91,
figs. 1, 2). The length of the valves ranges from .35 to nearly 4 nun. The
thickness of the head at the base varies from one quarter to one half the length
(compare PI. 91, figs. 1 and 4 or 5 and 6). The ophicephalous pedicellarise are
common but resemble some forms of the tridentate and are chiefly distinguished
by the loop on the valves. The latter (PI. 91, fig. 7) are somewhat elongated
with rather flattened blades and measure about .40-.50 mm. No pedicellarise
of the form characteristic of the genus (PI. 91, figs. 14, 20) were found. The
Iriphyllous pedicellarise are not uncommon but are so small as to be easily over-
looked. The valves, which are about .20 mm. in length, show considerable
diversity in form, as the blade may be nearly circular, or narrow and truncate
or, as is usually the case, broad and truncate; the one figured (PI. 91, fig. 8) is

220 HAWAIIAN AND OTHER PACIFIC ECHINI.

intermediate between the two latter. The sphceridia (PL 91, fig. 13) are some-
what elongated and a trifle angular; they occur on the lower secondary element
of most of the ambulacral plates. The calcareous spicules of the tube-feet are
large, irregular, fenestrated plates with an essentially smooth surface; they
occur in considerable numbers.

The coloration of this species shows almost no variation and is very charac-
teristic. The test is violet, at least abactinally, deepest and most marked on the
genital plates which are, at least proximally, very close to violet no. 507 of
Klincksieck and Valette's Code; actinally the violet fades into a dingy white.
The primary spines are dull reddish (about no. 87 K. & V.) at base, but fade
into a clear greenish yellow (about no. 266 K. & V.) at tip. The small spines
are very light and the buccal membrane and periproct are pale bluish. These
colors are-taken from an alcoholic specimen, but they are not essentially modified
by drying.

On first examination of the Hawaiian specimens, Mr. Agassiz and I were led
by the unbanded spines and the general features of the test to consider them
identical with the "Siboga" species, described by de Meijere as indica. A more
careful comparison of this material with the specimens of mirabilis, subsequently
received, together with a more critical examination of de Meijere's description
and figures showed that indica is quite distinct from any of the other species.
The characteristic differences have been set forth in the table above and do not
need to be repeated here, save to emphasize the fact that in indica the test and
spines are reddish flesh-color, the spines becoming whitish at the tip. The test
of de Meijere's specimen is recorded as 19 mm. in diameter and 12 mm. high,
which would indicate a much higher test than in hawaiiensis but as the photo-
graph is about 22 mm. by 11, the apparent difference is unimportant. Accord-
ing to de Meijere the actinostome in indica is as large as the abactinal system
and this may be a good specific character for in our specimens of hawaiiensis,
small and large alike, the actinostome is only three quarters as large as the
abactinal system.

The "Albatross" took this species at the following stations: —

Station 3865. Off Mokuhooniki Islet, Pailolo Channel, Hawaiian Islands.
Bott. temp. 44.8°-45°. 256-283 fathoms. Fne. vol. s., r.

Station 3879. Off Molokini Islet, south of Lanai, H. I. Bott. temp. 37.1°.
923-1081 fathoms. Glob, oz., r.

Station 3914. Off Diamond Head, Oahu, H. I. Bott. temp. 46°? 289-292
fathoms. Gy. s., m.

CENOPEDINA MIRABILIS. 221

Station 4178. Off Kawahioa Point, Niihau, H. I. Bott. temp. 43°? 319-
378 fathoms. Co. s., r., p.

Station 4179. Off Kawahioa Point, Niihau, H. I. Bott. temp. 42°. 378-
426 fathoms. Co. s., r., p.

Bathymetrical range, 256-1081 fathoms. Extremes of temperature, 46°(?)-
37.1°.

Eleven specimens.

Csenopedina mirabilis Mrtsn.

Hemipedina mirabilis Doderlein, 1885. Arch. f. Naturg., Jahrg. LI, 1, p. 24.
Caenopedina mirabilis Mortensen, 1904. Dan. Exp. Siam: Ech., p. 34.

Plates 91, figs. 14-17; 105, fig. 8.

This species has been very fully described by Doderlein (1906, "Valdivia"
Ech., p. 174) and the pedicellarue have been discussed and figured by Mortensen
(1. c.), so that there is little to add, but as many of the "Albatross" specimens
are considerably larger than Doderlein's, there are a few points worth mention-
ing. The largest specimen is 22 mm. in diameter, and 11 mm. high; the
abactinal system is 11 mm. across and the actinostome 10 mm. There are 9
interambulacral plates and 14 ambulacral, in each column. The longest pri-
maries measure 27 mm. It will be noticed that the test is a little higher, the
abactinal system a little smaller, the number of coronal plates, especially
ambulacrals, somewhat larger and the primary spines a little shorter than in
the type but the differences are too trivial to be of importance.

True tridentate pedicellarise seem to be very rare as only one was found, in
the examination of five specimens, but the globiferous ones are abundant, and
conspicuous because of the very dark-colored glands on their valves. The latter
are about .30 mm. in length and have, as figured by Mortensen, a truncate
or rounded tip with a prominent tooth on each side. But they are quite
variable, for between the two conspicuous teeth there are several others, which
are typically very minute, but one or more of them may be nearly or quite as
large as the lateral teeth; there are thus often three, four (PI. 91, figs. 16, 17),
or even five large terminal teeth. The length of these teeth is variable but I
have not happened to see any as long, relatively, as in Mortensen's figure. The
ophicephalous pedicellarise occur in two forms, that characteristic of the genus
(PI. 91, fig. 14) and one which intergrades with the tridentate (PI. 91, fig. 15).

222 HAWAIIAN AND OTHER PACIFIC ECHINI.

The latter are called tridentate by Doderlein, but ophicephalous is preferable
because of the articular loops.

The specimens studied show greater diversity of color than would be expected
from Doderlein's description, to which none of them answer exactly. The test,
although very light actinally, is light brown, dull violet or reddish brown above;
the color is deepest on the genital plates, while the periproct is often much lighter,
in decided contrast. The small spines are all light-colored and those of the
actinal side are nearly or quite white, but abactinally they are more or less
deeply tinged with yellowish green. The primaries are very pale brown or
yellowish green, with 3-6 narrow and indistinct bands of brownish red; as a rule
the bands are narrower than the light spaces which separate them, but near the
tip of the spine they may be as wide. The primaries are very slender and the
general appearance of this species is so different from hawaiiensis, as to leave no
doubt of its distinctness from that form. As hawaiiensis and indica are obviously
very nearly allied, I cannot follow Doderlein in considering indica merely a
variety or form of mirabilis.

The latter was taken by the "Albatross" at the following stations, the
specimens ranging from 7 to 22 mm. in diameter.

Station 3708. Off Ose Zaki, Honshu Island, Japan. 60-70 fathoms. Gr.
m., vol. s., a.

Station 4807. Off Cape Tsiuka, Japan, 41°36' 12" N., 140° 36' E. 44-47
fathoms. Sh., crs. g.

Station 4808. Off Cape Tsiuka, Japan, 41° 35' 50" N., 140° 36' 45" E. 47
fathoms. S., sh., crs. g.

Station 4900. Off Ose Saki Light, Eastern Sea, 32° 28' 50" N., 128° 34' 40" E.
Bott. temp. 52.9°. 139 fathoms. Gy. s., brk. sh.

Station 4933. Off Kagoshima Gulf, Japan, 30° 59' N., 130° 29' 50" E. Bott.
temp. 56°. 152 fathoms. Rky.

Station 4934. Off Kagoshima Gulf, Japan, 30° 58' 30" N., 130° 32' E. Bott.
temp. 60.6°-56°. 103-152 fathoms. Rky.

Station 4965. Off Hiro Misalu Light, Japan, 33° 35' 20" N., 135° 10' 50" E.
Bott. temp. 49.4°. 191 fathoms. Dk. gn.-gy. s., sh.

Station 5047. Off Kinka San Light, Japan, 38° 12' 50" N., 141° 49' 15" E.
Bott, temp. 49.6°. 107 fathoms. Dk. gy. s., brk. sh., p.

Bathymetrical range, 44-191 fathoms. Extremes of temperature, 60.6°-
49.4°.

Forty specimens.

C^NOPEDINA PULCHELLA. 223

Caenopedina pulchella, comb. nov.

Hemipedina pulchella A. Agassiz and Clark, 1907. Bull. M. C. Z., L, p. 245.

Plates 91, figs. 18-2%; 103, figs. 1-3; 105, figs. 6, 7.

In superficial appearance this species is so unlike any other member of the
genus that its real relationship was not suspected until the abactinal system was
examined, and it was only by the inspection of the tubercles that its generic
position was determined. The larger specimen is 14 mm. in diameter, while the
height of the test, the diameter of the abactinal system and that of the actino-
stome, are each about half as much. In the smaller specimen, which is only
about half as large the proportions do not seem to be essentially different, though,
as might be expected, the abactinal system and actinostome are perhaps a little
larger relatively. But while the smaller specimen has 6 interambulacral and 7
ambulacral plates in each column, the larger has 8 and 11 respectively. The
genital plates (PL 103, fig. 1) are large, heptagonal and broadly in contact with
each other. The central part of each plate is rough, one might almost say
sculptured, and on the proximal portion are two or three secondary tubercles.
The pore is near the centre. The ocular plates are small, scarcely one fifth as
large as the genitals; like the latter, their surface is rough but they carry no
tubercles. The periproct is small, not so large as a genital, and is covered by
about 20 rather large plates, none of which however carry tubercles.

The interambulacral plates are high, but those above the ambitus are each
almost completely covered by the primary tubercle; there is just room at the
corner of each plate for a small secondary tubercle. Actinally the primary
tubercles are much smaller and there is room, on both the outer and inner sides,
for some small secondary tubercles; miliary tubercles seem to be wanting.
Excepting one or two of the uppermost in each column, the ambulacral plates
each carry a primary tubercle ; those at the ambitus are largest but even they are
little larger than the actinal interambulacral tubercles. Aside from the primaries
only a very few tubercles and those, small actinal secondaries, are found in the
ambulacra. The pore-pairs are small, forming a narrow almost vertical area.
The ambulacra are about three fourths as wide as the interambulacra at the
ambitus.

Aside from the primordial ambulacral plates, the buccal membrane (PL 103,
fig. 2) is almost naked, only a very few, small, rounded, non-ambulacral plates
being scattered here and there. There are pedicellarise and a few small spines on

224 HAWAIIAN AND OTHER PACIFIC ECHINI.

the primordial plates. The gills are small and their slits insignificant. The
auricles are moderately high but widely separated, showing no tendency to meet.

The primary tubercles, especially those above the ambitus, in the inter-
ambulacra, are very large, with disproportionately large, perforated mamelons.
Of the secondary tubercles, only one or two are perforated. There is not the
least trace of crenulation on any of the tubercles. The primary spines are very
conspicuous, at least those on the uppermost three or four interambulacral plates
in each column. These measure from 12 to 22 mm. in length and from 1 to 2 mm.
in thickness; the thickness is from .06 to .15 of the length but is usually less
than .10. All the spines taper to a blunt point, the large ones rather abruptly,
the small ones gradually. They are very finely and uniformly striated longi-
tudinally but are quite smooth. All are perfectly solid throughout.

Pedicellarise of all kinds are fairly common. The globiferous are not con-
spicuous, for the glands are light brown, not dark brown or purple as in the other
species. Their valves (PI. 91, fig. 18) terminate in two slender teeth, remarkable
for their length. The tridentate pedicellarise are very variable, the valves rang-
ing from .20 to 2 mm. in length; they are sometimes broad and flat (PI. 91, fig.
19) but are usually narrow and compressed, and may be either straight and in
contact for some distance or curved and meeting only at tip. The ophicephalous
pedicellarise occur in the same two forms which were referred to under mirabilis;
the characteristic forms are remarkable for their very narrow valves (PI. 91, fig.
20), which are about .30 mm. long, including the loop; the other form intergrades
completely with the tridentates. The triphyllous show no peculiarities; the
valves are .10 -.13 mm. long. — Tlie sphceridia (PI. 91, fig. 22) are somewhat
elongated but show no characteristic features. The calcareous spicules of the
tube-feet (PI. 91, fig. 21) are remarkable only for being finely spiny on their
convex surface.

The coloration of this species, as shown in the alcoholic specimens, is un-
usually handsome. The test is white actinally, becoming rosy above; the genital
and ocular plates are deep brownish rose. The periproct is white. The large
primaries are light green (near no. 286 K. & V.) at base, dull rose-red (near no.
17 K. & V.) on the distal half and light, almost or quite white, at tip. The
ambulacral primaries and all the secondaries are white. The actinal inter-
ambulacral primaries are nearly or quite white with one or two bands of reddish.

Were it not for the fact that the outline of the test is circular, this beautiful
sea-urchin would be easily mistaken for an Echinometra, the short stout pri-
maries are so suggestive of that genus. It is obvious however that it has no near

PHYMOSOMATID.E. 225

relationship with the Echinometridae but is undoubtedly a Csenopedina, though
it is so strikingly different from the other members of that group.
The "Albatross" took this species only at the following place: —
Station 3991. Off Mokuaeae Islet, Kauai, Hawaiian Islands. Bott. temp.
43.7°. 272-296 fathoms. Fne. s., r.
Two specimens.

PHYMOSOMATID.E1 Meissner.
GENERAL CHARACTERISTICS.

Although the superficial appearance of the single living representative of
this family, Glyptocidaris crenularis, is quite like that of an Echinus, the structure
of the "lantern" shows that it belongs in the suborder Stirodonta Jackson and
the ambulacra are also very different from those of the Echinida;. The ambul-
acra are very similar to what is found in a number of fossil Phymosomatidse, but
as Duncan (1885, Quart. Journ. Geol. Soc., XLI, p. 449) pointed out, the fact that
the pores are not " diplopodous " abactinally prevents our placing the recent
species in Phymosoma and necessitates retaining the name Glyptocidaris, under
which it was originally described. The structure of the ambulacral plates is
very characteristic and is well shown in the Revision of the Echini (PI. VI, fig. 2)
where it may be seen that the pores are in arcs of five and not in alternate arcs
of two and three as one would naturally suppose from the photograph (Rev.
Ech., PI. VII a, fig. 6). Each ambulacral plate is made up of three primary
and two secondary elements. The demi-plates lie between the primaries so
that the adoral primary element is followed by a demi-plate, then by the middle
primary element, then by the second demi-plate and lastly by the aboral primary
element. It is easy to see in such a plate a modification of the more simple
centrechinid tripartite plate by the introduction of secondary elements in a very
different order of succession from what is found in the Echinidse and Echino-
metridse. In those families only two primary elements are normally retained
in each ambulacral plate, the elements between them being, in most cases, demi-
plates. When three primary elements are present, one is adoral and two lie
together aborally, while one or more demi-plates follow the adoral element.

1 In reviving the name Cyphosoma and establishing a family Cyphosomatidse, for this group,
Duncan (1885 and 1889) appears to have overlooked the fact, although the necessary data are given
in the Revision of the Echini, pt. 1, p. 151, that Cyphosoma, as a genus of Coleoptera antedates Cypho-
soma, a genus of Echini, by three years.

226 HAWAIIAN AND OTHER PACIFIC ECHINI.

The "lantern" of Glyptocidaris (PI. 90, figs. 7-9) is remarkably like that of
Arbacia. The teeth are strongly keeled, the jaws are very erect, the foramen
magnum is deep and the epiphyses do not meet above it. Unlike Arbacia, but
like Stomopneustes, the top of each half-pyramid gives rise to a process extend-
ing down on each side of the tooth giving support to the latter. These proc-
esses are not from the epiphyses. One is well shown in fig. 7, PL 90, but in
figures 8 and 9 this important character is not clearly indicated. The pits
beneath the epiphyses, characteristic of the Centrechinoida, as recently shown
by Jackson (1912, Mem. Boston Soc. Nat. Hist., VII, p. 178, 183) are
remarkably large and deep. The perignathic girdle is similar to that of Arbacia,
the auricles being more or less in contact but the apophyses between them
being inconspicuous. Unlike Arbacia the primordial interambulacral plate is
resorbed.

The alimentary canal (PI. 90, figs. 5, 6) is surprisingly like that of Centre-
chinus. There is a very long oesophagus followed by a capacious and much
folded intestine, the loops of the upper coil alternating with those of the lower.
The anus lies to the right and a little behind the centre of the periproct, and a
small but distinct suranal plate is evident on the opposite side. The right
posterior ocular plate is in contact with the periproct but all of the others are
excluded, in all of the five specimens examined.

The buccal membrane carries, in addition to the five pairs of buccal plates,
numerous smaller plates, many of which carry spines as well as pedicellarise.
The gill-slits are neither deep nor conspicuous, though the gills are well devel-
oped. Rudiments of Stewart's organs are obvious on the sides of the lantern-
membrane, below the ends of the compasses.

THE SPINES, PEDICELLARISE, SPH^RIDIA, AND SPICULES.
Plate 92, figs. 1-11.

The spines are long, slender, and pointed, and the primaries are especially
conspicuous. The surface of the spines is very smooth, while the inner structure
is very compact and, to a certain extent, resembles that of some Arbacia spines.
(See Mclntosh, 1883, Trans. Roy. Irish Acad., XXVIII, p. 255, PL 8, fig. 32.)
The tubercles are all imperforate and the larger secondaries as well as all the
primaries are very markedly crenulated. The boss is very high and the scro-
bicule is nearly or quite flush with the surface of the test.

The pedicellarise of Glyptocidaris are very characteristic. They have been

PHYMOSOMATIDjE. 227

briefly described by Doderlein (1906, Zool. Anz., XXX, p. 520), who proposes to
place the genus in Mortensen's family Toxopneustidaa, because the valves of the
globiferous pedicellariae terminate in a single prominent tooth. It is interesting
to note that the first protest against this classification is made by Mortensen
himself (1910, Vid. Med., p. 31), who considers the pedicellarise an unsafe guide
in this case.

The globiferous pedicellarise (PL 92, fig. 1) are remarkable for their stalks,
which have 3-5 branches on each side. These pedicellarise are well distributed
and fairly common. The valves (figs. 6, 7) are of very variable size, ranging
from .25 to .75 mm. in length, while the single terminal tooth is from one fourth
to one third as long. These pedicellariae are quite unlike anything found in the
Arbaciadse and resemble only in a general way certain forms occurring in the
Echinidse.

The tridentate pedicellarise (PL 92, fig. 2) are rather uncommon and in one
of the two specimens examined appear to be wanting. They usually have a
neck, which may be quite long, but in some cases appears to be wanting. The
valves (PL 92, fig. 5) are broad and in contact only along the distal half. They
are 1-1.60 mm. in length and the margin is either entire or finely sinuate.

The ophicephalous pedicellarise (PL 92, fig. 3) are very common, particularly
actinally. The valves (PL 92, fig. 9) are short and wide, somewhat rounded
triangular in outline, with a low wide "loop" which is essentially the same on
each valve. The apophysis is prominent and continuous with the coarse mesh-
work which occupies the centre of the blade. The margin is slightly sinuous at
least near the tip. The valves measure about .75 mm. in length and are nearly
two thirds as wide.

The triphijllous pedicellarise (PL 92, fig. 4) are small and not very abundant.
The valves (PL 92, fig. 8) are rather elongated and rounded at the end. They
measure about .25 mm. in length and are a little more than half as broad.

The sphceridia (PL 92, fig. 10) of Glyptocidaris are numerous, as many as
15-20 occurring on each side of each ambulacrum. They are not sunken in pits
but are borne on the surface of the upper demi-plate, of each ambulacral plate,
from the actinostome to some distance above the ambitus. They are nearly
spherical and are situated at the inner end of the demi-plate, beside the tube-foot.

The spicules (PL 92, fig. 11) in the tube-feet are in the form of rods, expanded
and perforated at the middle but more or less drawn out at each end.

228 HAWAIIAN AND OTHER PACIFIC ECHINI.

GLYPTOCIDARIS.

A. Agassiz, 1863. Proc. Acad. Nat. Sci. Philadelphia, p. 356.
Type-species, Glyptocidaris crenularis A. Agassiz, 1. c.

This, the only living genus of the family, is distinguished from all of its
extinct allies, by the structure of the ambulacra. Each plate is made up of five
elements, an adoral primary being followed by a demi-plate, a middle primary,
an upper demi-plate, and an aboral primary. Near the ambitus the pore-pairs
of the aboral primary, the upper demi-plate and the middle primary form an
oblique arc of three, which is followed by an arc of two made up of the pore-pairs
of the lower demi-plate and the adoral primary; this alternation is marked by
the fact that the outermost pore-pair of each plate is that of the middle element
while the pore-pair of the aboral primary is further in than that of the upper
demi-plate; that of the lower demi-plate is directly below that of the aboral
primary, while that of the adoral primary is further out. As we follow the
ambulacrum dorsally, we find the pore-pairs come to lie more and more in a
single vertical series and are never diplopodous as in Phymosoma and allied
genera.

Glyptocidaris crenularis A. Ag.

Glyptocidaris crenularis A. Agassiz, 1863. Proc. Acad. Nat. Sci. Philadelphia, p. 356.

Plates 90, figs. 5-10; 92, figs. 1-11; 106, figs. 1, 2.

This species appears to be confined to the coasts of northern Japan. It was
first discovered near Hakodate by Stimpson and has been recorded from the
same region by Doderlein (1906, Zool. Anz., XXX, p. 520). It appears to be not
only local but rare, for the "Albatross" took but few specimens. The largest
(PI. 106) measures 75 mm. in diameter and the primary spines are about 50 mm.
long.

These specimens were taken at the following stations : -

Station 4807. Off Cape Tsiuka, Japan; 41° 36' 12" N., 140° 36' E. 44-47
fathoms. Sh., crs. g.

Station 5046. Off Kinka San Light.. Japan; 38° 15' 7" N., 141° 44' 20" E.
Bott. temp. 50.8°. 82 fathoms. Dk. gy. s., p.

Four specimens.

STOMOPNEUSTID^;. 229

STOMOPNEUSTID.E Mortensen.
Plate 90, figs. 11, 12.

To Mortensen (1903, "Ingolf" Ech., pt, 1, p. 133) belongs the credit of
separating Stomopneustes from the other regular Echini and placing it in a
family by itself. While his reasons for so doing seem quite trivial, the exami-
nation of the internal anatomy reveals the soundness of his conclusion and
strongly suggests that where the pedicellarise and .spicules of a sea-urchin show
really important characters, there will be found morphological characters of
real significance. And the corollary would naturally follow that if the careful
study of an echinoid fails to reveal characters in the test or the internal anatomy,
of real morphological value, whatever differences the spines, pedicellariae, and
spicules may show, however interesting they may be, are not significant and have
little bearing on a natural classification.

The alimentary canal of Stomopneustes is very long and much looped, as in
the Centrechinidae. Its most noticeable peculiarities are the short oesophagus,
and the very large intestinal appendage. The lantern shows at once how dis-
tinct from either the Echinidse or Echinometridse, Stomopneustes is, for the
epiphyses are relatively small and do not arch over the foramen, as they do in
those families. The teeth are however strongly keeled. The general appearance
of the lantern and teeth is very similar to that of Glyptoddaris crenularis (PI. 90,
figs. 3, 4) and shows that the family undoubtedly belongs in the suborder
Stirodonta Jackson. The auricles in Stomopneustes are only moderately
developed, meeting but hardly fusing, in an arch.

The test is essentially similar to what is found in the Echinidse. The outline
of the ambitus is typically circular, but there seems to be a slight tendency to
elongation of one axis. The M. C. Z. collection contains a specimen 69 mm.
measured through interradius 2 and ambulacrum V, but only 67 mm. at right
angles to that line. Another specimen is 59 mm. through ambulacrum III and
interambulacrum 5, but only 57 mm. at right angles to that line. In both of
these specimens, the test is somewhat asymmetrical and the elongation may be
pathological and not normal. The primary interambulacral plates are resorbed.
The primary ambulacral (buccal) plates carry numerous pedicellariae and some
small spines. Scattered in the buccal membrane are a number of small plates,
some of which, at least, carry pedicellarise. The ambulacra are remarkable
for the excessive development of certain primary tubercles. Seen from the inside,

230 HAWAIIAN AND OTHER PACIFIC ECHINI.

the ambulacral plates, are obviously made up of three elements, as in the Echini-
dae, an adoral primary element and two secondary elements above it. But every
fourth plate is typically greatly enlarged at its inner end and shuts out the three
succeeding plates from the median line. Examination of the exterior shows
that it is the primary tubercle of this enlarged plate which has grown over the
four plates concealing their sutures, and giving the impression of high ambulacral
plates with a dozen pairs of pores. So far as known this type of ambulacral
plate does not occur elsewhere among Echini. The abactinal system of Stomop-
neustes is much like that of. the Strongylocentrotidse; that is, oculars I and V
are typically in contact with the periproct, while ocular IV, or less commonly
ocular II, may also reach it. The periproct is covered by numerous small plates,
many of which carry spinelets. No suranal plate is distinguishable in mature
specimens.

The spines of Stomopneustes are very stout and moderately long. In cross-
section they are obviously polycyclic. There are several (4-8) sphseridia on
the actinal part of each ambulacrum; each is attached to a miliary tubercle,
but is not associated with any depression or other shelter. The pedicellarise and
remarkable spicules have been fully described by other writers (see Mortensen,
1903, "Ingolf," Ech., pt. 1, p. 126). In a careful examination of four specimens
from the Persian Gulf, Mozambique, and New Guinea, only one kind of globif-
erous pedicellarise was found, the characteristic form, and I can therefore fully
confirm Mortensen's statement as to the scarcity of this to him very important
organ; for on the two largest and best preserved specimens, there appear to be
none, and on each of the small specimens but a single example. As Mortensen
gives no measurements it may be well to add some based on my observations.
The valves of the globiferous pedicellarite are about .80 mm. long, while the base
of each is about .35, and the tip of the blade .07 mm., in width. The tridentate
valves range from .30 to 1.10 mm. in length, the ophicephalous from .35 to .80,
besides the loop, and the triphyllous, which are the only really common pedicel-
larise, from .15 to .20 mm. The extraordinary spicules of the tube-feet are about
three quarters of a millimeter in length and about one tenth as much in diameter.

Lack of material makes it impossible to decide positively whether there is
more than a single species in this genus. Specimens from New Guinea do
not seem to be distinguishable from those from Mauritius, but they are not
"almost black" and so may not be identical with the supposed form "atropur-
pureus" from Queensland. The examination of numerous specimens from a
considerable number of localities makes it very doubtful whether the characters

ECHINID^. 231

which were supposed to distinguish atropurpureus have any real systematic
significance.

Stomopneustes has not been recorded from the northern Pacific Ocean, nor
did the "Albatross" find it in either Hawaiian or Japanese waters.

Agassiz.
GENERAL CONSIDERATIONS.

There can be little reason to doubt that the EchinidsB are the present-day
representatives of the stock from which in the past the Temnopleuridse, Strongy-
locentrotidse, and Echinometridse have been derived. Indeed the relationship
with each family is so close that it is impossible to fix a natural boundary, passing
which no exceptions will be found. It is necessary therefore to choose between
fixing arbitrary lines and the alternative proposition of uniting all four groups in
a single family. The real interrelationships of the regular Echini are much
better shown by following the former course. The question then arises as to
the characters for distinguishing the groups. As in the case of the families
hitherto treated, the test furnishes the characters of morphological significance.1
Mortensen (1903, "Ingolf" Ech., pt. 1 and 1904, Dan. Exp. Siam: Ech.) has
chosen the pedicellarise as the chief source for light on the relationships of these
Echini, and his lead has been followed by some other students of the group.
It is not altogether strange that my results differ greatly from his, and it is to be
regretted that they cannot be brought into something more nearly like harmony.
But in addition to the objections previously urged against the use of the pedicel-
larise as a factor of primary importance in classification, study of the Echinidse
and their allies has suggested two others which are of special weight in connection
with these forms.

The first important objection to Mortensen's system is the separation which
results, of species obviously and by all the characters of the test very closely
allied, and the union of species, which it seems impossible to associate together.
To illustrate this objection the placing of Echinus magellanicus Phil, and E. albo-
cinctus Hutton in separate families, may be mentioned, although the two are so

1 This was written before the publication of Jackson's magnificent monograph on the Phylogeny
of the Echini (Mem. Boston Soc. Nat. Hist., VII, 1912) in which the same course is followed. The
classification used herein is essentially identical with that to which he has been led by his prolonged
morphological studies.

232 HAWAIIAN AND OTHER PACIFIC ECHINI.

much alike that they can be separated only with the greatest difficulty and I
am unable to regard them as really distinct species. Mortensen himself ("In-
golf" Ech., pt. 1, p. 140) recognizes the close relationship of these two forms
but considers it desirable to separate them in order to make the family definitions
more precise. The difference in their globiferous pedicellariae, upon which
their position in different families is based, is so slight, that the comparison of
Mortensen's figures ("Ingolf " Exp., pt. 1, PI. 19, figs. 19 and 23) and the perusal
of Doderlein's paragraphs ("Valdivia" Ech., p. 232) in which Mortensen's
course is defended, is suggested to any one interested in the matter. It is un-
fortunately necessary in systematic work to use lines of division which do not
exist in nature, and it must be frankly admitted that some of those employed
here are open to serious criticism, but it is hoped that the separation of any two
closely allied forms upon such an utterly trivial basis as this which is supposed
to separate magellanicus and albocinctus has been avoided. Similar cases of
wide separation of forms which seem very closely allied are the cases of Strongylo-
centrotus Hindus and drobachiensis, S. tuberculatus and franciscanus, and S. albus
and S. gibbosus. On the other hand, we find more or less close relationship
proposed between S. albus and Echinus microtubcrculatus, between S. lividus and
the species of Echinus, between Splicer echinus granularis and Tripneustes and
between Heliocidaris and the Echinometridao. All of these seem most improbable
if not impossible associations. In view of these striking cases it is not easy to
understand how Mortensen can express himself as he does (1. c., p. 140) concerning
the "natural relation" of his proposed groups.

The second objection to Dr. Mortensen's system as applied to the Echinidse
is the inconsistency of relying on it for the greater part of the regular Echini
but failing to use it for the Temnopleuridse. I can hardly do better than
quote Mortensen's own words on this point (Dan. Exp. Siam: Ech., p. 56):
"It is a very surprising fact that the pedicellarise of the Temnopleurids prove
to be only of subordinate value for classification. * * * they mostly give only
specific characters. In the larger genera * * * the globiferous pedicellarise
assume the forms occurring both in the Echinida?, Toxopneustidse and Echino-
metridse; in some species * * * they even occur in the same specimen in both
the two forms which distinguish the families Toxopneustidae and Echinometridae.
This very curious fact, of course, does not alter the classificatory value of the
pedicellariae in the other regular echinids; but we are forced to seek the generic
characters of the Temnopleurids in the structure of the test." The conclusion
which Dr. Mortensen reached for the Temnopleurids, from his study of the

ECHINID.E. 233

pedicellarise of that family coincides with that which I have reached from the
study of all the regular Echini: i. e. that while the pedicellarise often afford
good specific characters (doubtless in correlation with other features) they are
not, taken by themselves, reliable as a guide in seeking for the true interrela-
tionships of the species. It does not seem that a character of such uncertain
value in the Temnopleuridae can possibly become of prime importance in the
closely related Echinidse. Taken in connection with the other characters how-
ever, the globiferous pedicellarise undoubtedly assist in tracing the differentiation
of the species, and in grouping them in genera, and in a few cases they are the
most obvious, if not the most important, generic character.

On turning to the test for the characters upon which to base a natural classi-
fication, it is apparent at once that in all the less specialized Echinidse as well as
in the Temnopleuridse the outline of the test is circular 1 and the ambulacral
plates are made up of an adoral primary element and two secondary elements,
the pairs of pores being placed in nearly vertical arcs of three. From this simple
•ancestral form, development has proceeded along at least four different lines.

(1) The Temnopleurids have developed more or less sculptured tests, with
the coronal plates often united by dowelling, while undergoing little if any modi-
fication of the ambulacral structure; until we reach the highly specialized
condition of Holopneustes where the lateral spreading of the poriferous areas,
associated probably with some sort of vertical pressure, has greatly increased
the number of ambulacral plates, with accompanying displacement, but without
increasing the number of their elements.

(2) A similar development of the ambulacra, by great increase in the number
of ambulacral plates, without altering their tripartite structure has occurred in
Tripneustes and its allies in the Echinidse proper. In the Strongylocentrotidse
and Echinometridse, development of the ambulacra has been along a different
line, for there is rarely any great crowding of the ambulacral plates or displace-
ment of their elements but instead there has been a more or less marked increase
in the number of component elements in each plate. The connection between
these families and the Echinidse is obvious and it is very interesting to note that
in the case of each one, there is a genus which might properly be assigned either
to the parent or the derived family.

(3) The Echinometridse differ from the Echinidae, besides the difference in
the ambulacral plates, in the elongated form of the test. Yet in Parasalenia

1 The case of Microcyphus annulatti.i Mortensen appears to be no exception, as a series of eight
specimens shows that the outline of the ambitus is circular or slightly pentagonal.

234 HAWAIIAN AND OTHER PACIFIC ECHINI.

an elongated test is associated with only three pairs of pores in each ambulacral
plate. As a further indication of the intermediate position of Parasalenia,
attention may be called to the fact that the elongation of the test in that genus
is through IIIb-5b, while in Echinometra it is ordinarily through 3-1 and in
Heterocentrotus, the most extreme of the family, it is through IVb-lb or 4a-IIa.
There is thus a progressive movement of the axis to the animal's left, which is
least marked in Parasalenia. It seems that it would be perfectly proper to
class that genus with the Echinidse if one preferred, although for convenience
it is here placed in the Echinometridse.

(4) The genus Echinostrephus, in a somewhat similar manner, serves to
connect the Echinidse, characterized by only three pairs of pores in each
ambulacral plate, with the Strongylocentrotidse, characterized by more than
three pairs. For in one species of Echinostrephus there are three pairs of pores
and in the other there are four. The line of division between the two families
thus cuts the genus Echinostrephus exactly in two and it might therefore be
properly placed in either family. On account of the specialized form of the test
however, I place it in the Strongylocentrotidse, and consider this view strength-
ened by the fact that specimens of Strongylocentrotus, of two different
species, occur in which the ambitus is above the mid-zone, as is so character-
istic of Echinostrephus. This peculiarity has not been noted in any species of
the Echinidse.

Since so much stress is laid on the importance of the number of pore-pairs in
an arc, or rather on the number of elements in each ambulacral plate, it is proper
to utter a word of caution in regard to the use of this character. It must con-
tinually be borne in mind that the oldest plates in each ambulacrum are those
at the peristome, while the youngest are next to the ocular plate. The former
often show therefore features characteristic of recent ancestors, while the latter
show youthful characters, i. e., are not fully developed. Consequently it is in
the area between these two regions, aptly termed by Jackson the "mid-zone,"
that the specific characters must be sought. For practical purposes, the region
just above the ambitus (in Echinostrephus, just below the ambitus) will show the
species character in adult specimens. In young specimens, obviously the species
character will be further above the ambitus and in very young specimens it will
be found, if at all, near to the ocular plate. On the other hand in very old
specimens, old age characters will begin to appear near the ocular, senescence
having set in. In an old specimen therefore of a Strongylocentrotus with six
elements in the ambulacral plates, as its specific character, there will be found

ECHINID^E. 235

near the peristome five (sometimes four or even three) pore-pairs in each arc;
at the ambitus and above there will be six and near the ocular plate, there will
often be five or only four. In a really young specimen of the same species there
will be four (or only three) pairs near the peristome, five at and above the ambitus
and six near and adjoining the ocular plate. In the Echinidse and Temnopleur-
idse, there is no change in the number of elements in each plate as one passes
from the peristome upward but the relation between them becomes more and
more simple and obvious as one passes from the mid-zone to the ocular. This
increasing simplicity is most marked in forms like Holopneustes and Tripneustes.
Finally, it should be remembered, that any individual plate is liable to mal-
formation and to variation from the typical condition, so that plates with only
four or five elements are occasionally found intercalated between those having
six or even more.

Immature specimens are often a source of difficulty and error. The absence
of genital pores is one of the most obvious evidences of immaturity but unfortu-
nately these pores appear long before maturity so that their presence is no
criterion of age. Disproportionately large abactinal and actinal systems are
youthful characters, while the appearance of the periproct, the pore-pairs, and
the spines often show whether a specimen is mature or not. The identification of
young Echini is often very difficult and it is frequently impossible to distin-
guish the young of allied species or even of allied genera, unless a series of speci-
mens showing growth changes is available for comparison.

No reliable system of classification permits the certain identification of
isolated, immature specimens. One of the very few conveniences of using the
pedicellarise as the basis of classification in Echini, is that they are essentially
the same in the youngest specimens in which they are found, as in the adult.
In reality however this is an important argument against their validity for
systematic purposes, for it can hardly be questioned that a character which
appears full-fledged in early youth and undergoes no change in ontogeny, has
no phylogenetic significance. The occasional convenience therefore of using
the pedicellariae in identifying young Echini should not be construed as evidence
of their systematic value.

236 HAWAIIAN AND OTHER PACIFIC ECHINI.

THE SPINES, PEDICELLARLE, SPH^RIDIA, AND SPICULES.

Q3 fia-s / 1 % 99 9^ Q9

yo, ngb. 1—10, zz, zo, oz.

The spines of the Echinidse are smooth and solid and afford few characters
of value for systematic purposes. Generally there is no very marked difference
between the primaries, secondaries, and miliaries, but in a few cases the primaries
are conspicuously longer. As a rule the primaries are relatively short and stout
but in a few cases they are long, slender, and pointed. The secondaries also are
usually rather short and stout, but are occasionally very slender and sometimes
are noticeably rough.

The pedicellarise of the Echinidse show considerable diversity, all of the four
kinds being more or less common. The following summary of their characters
is necessarily largely a repetition of the very complete publications of Mortensen.
The globiferous pedicellarise are remarkable for the very great development of
poison-glands in connection with the valves. These glands may be on the stalk
or on the valves of the pedicellarise, but in either case are commonly quite con-
spicuous. They may be present in both positions. The stalks themselves are
made up of slender calcareous threads, which are only ver,y slightly connected
with each other; or the threads may be stouter and very fully united together
making either a somewhat tubular or a solid stalk. In some cases the head is
connected with the stalk by a neck of greater or less length but often the neck
is wanting. The valves show a greater or less specialization in the different
species, the most specialized valves being found in those species which have some
specialized condition in the test. In their simplest condition, the valves have
the blade open, but more or less deeply concave and in addition to the terminal
tooth which is of moderate size there are additional teeth on each lateral margin.
In a slightly advanced condition the terminal tooth is larger, there are only one
or two teeth on each side and there are cross-bars of lime connecting the margins
of the somewhat compressed blade. Further specialization occurs by the closer
and more complete union of the margins and the development of the terminal
tooth with the accompanying reduction or loss of the lateral teeth. In the most
specialized condition the blade is cylindrical and terminates in a single very
prominent tooth, but in other cases, there is a very large lateral tooth on the
left side, slightly below the tip. Sometimes, in species which typically have this
unpaired tooth, valves with a similar but somewhat smaller tooth on the right
(PI. 95, fig. 24) are found. The steps by which the transformation from the

ECHINID.E. 237

simple to the most specialized form takes place are easily followed, although as a
rule the form found in any one species is very constant. In some species how-
ever, like P. magellanicus, considerable diversity is found even in one specimen.
Aside from the great differences shown in the actual size of the valves, there is
much variation in the relative length and breadth of the blade, the base and the
terminal tooth. In many species the tissue surrounding the valves is more or
less filled with calcareous spicules, which may be either dumb-bell shaped, or
bluntly, or sharply bihamate.

The tridentate pedicellarise reveal as great a diversity of structure as that
shown by the globiferous, but the diversity is much less correlated with specific
limits and it is difficult to detect anything like progressive specialization. The
simplest condition of the valves is probably that shown by many of the small
ones, where the blade is about twice as long as the base, perhaps a little more
than twice its own width, only slightly curved and roundly pointed at the tips.
From this simple condition, specialization has diverged, on the one hand towards
excessively elongated, compressed valves, and on the other towards broad,
stout valves little compressed. Several forms are often found on one specimen,
but the two extremes do not occur together. The amount of calcareous mesh-
work in the cavity of the blade is variable; sometimes it is almost wholly wanting
while in other cases it occupies nearly the entire inner surface of the blade.
These pedicellariae have a slender stalk, and usually more or less of a neck. The
valves themselves may be more than three millimeters in length but are usually
about a millimeter, though they are often much less. It is an interesting fact
that the globiferous and tridentate pedicellarise seem to be more or less supple-
mentary to each other, for when one kind is unusually abundant, the other is
often quite wanting. Thus in some specimens, even to some extent in certain
species, tridentate pedicellarise are common enough, but we search in vain for
the globiferous, while in others tridentates are not found but the globiferous
occur in sufficient quantity.

The ophicephalous pedicellaria? are always present in greater or less numbers
and although ordinarily easily recognized, they may intergrade more or less
with the tridentate. They have no neck but the valves have well-developed
"articular loops," which differ strikingly in size on the three valves. The
stalk is thick and solid. In the form of the valves two quite distinct types occur,
the unconstricted and the constricted. In the former there is no sharp distinc-
tion between base and blade, and the valves are more or less triangular; they
are sometimes narrow and elongated and then intergrade very naturally with

238 HAWAIIAN AND OTHER PACIFIC ECHINI.

the tridentate through the suppression of the loop, which is in any case rather
small. In the constricted pedicellarise, the loop is usually very well developed,
and a marked constriction separates the blade from the base of the valve; the
blade itself may be nearly circular or more or less elongated, but it usually
contains a very considerable calcareous mesh-work.

The Iriphyllous pedicellarise seem to be constantly present, though on account
of their very small size, they are often difficult to find, especially when the other
sorts of pedicellarise are very abundant. Their heads are usually from .10 to .20
mm. long and are borne on stalks four or five times as long, with which they are
connected by a very extensile neck. The valves themselves show little diversity
in form and scarcely any structural peculiarities. The distal half is commonly
wider than the base and is distinctly truncate, though in a few species it is
rounded. The breadth of the valve is usually about equal to the length and may
exceed it. In most cases the ending of the apophysis in the blade is quite indis-
tinct, but it may appear as a well-marked fork and rarely it gives rise to ridges
which run out onto the blade.

The sphceridia show very little indication of specialization. They are present
in some numbers on the actinal part of each ambulacrum, but are not sunken
in any depressions, nor have they any unusual relation to the plates. They
occasionally occur on the buccal plates, and as Mortensen has shown, their
intergradation with miliary spines is sometimes quite evident. They are more
or less elongated and are usually smooth but may be quite rough, at least at the
tip.

The calcareous spicules of the tube-feet, when fully developed, are of the
typical bihamate form, but they are often found with the ends blunt instead of
sharp. Mortensen ("Ingolf" Ech., pt. 1, PI. 21, fig. 31) has given a good series
of figures showing the transition from simple granules to dumb-bell shaped
spicules and from the latter to complete bihamate rods. Sometimes the bi-
hamate spicules are more or less branched or provided with teeth at and near
the tip. Spicules are sometimes abundant but are often uncommon and very
hard to find, and are not of the least value for systematic purposes in this family.

THE GENERA AND SPECIES OF RECENT ECHINHXE.

There are rather more than fifty recent species, belonging in this family as
here limited; the fossil forms unfortunately must be left out of account.
The recent species form a homogeneous group and it is difficult to

ECHINID.E. 239

arrange them in genera, which shall be at the same time, natural and sharply
denned. For few groups show more clearly than do the Echinidse, the pro-
gressive differentiation of species and while it is possible to trace out the probable
lines of development, it is exceedingly hard to arrange the species in genera in
such a way as to indicate those lines. Taking any single character as a standard
we can indeed make lines of division sufficiently sharp to be easily seen, but such
a course works havoc with the natural relationships. Moreover some of the
most important characters show such a perfect series of steps from the simple
to the specialized condition, even within the limits of a single species, that they
are of little value in defining genera, or at least, must be used with great caution.
As an illustration of this point reference may be made to the tuberculation of
the ambulacra. Mortensen ("Ingolf" Ech., pt. 1, p. 93) first called attention
to the marked difference between having a primary tubercle on each ambulacral
plate, and having one, only on each second, third, or fourth plate. Associated
as this is with a change in the form of the plates and the arrangement of the
pores, it is indeed a most important character. The use of it however as a
factor of primary importance in the classification of the family is nullified by the
fact that in Echinus acutus, specimens may be found in which every ambulacral
plate has a normal primary tubercle, while in others not only are many ambula-
cral plates smaller than their fellows and lacking a primary tubercle, but some
of the abactinal ones are actually made up of only two elements, a most unusual
condition in this family. Between these two extremes, all possible intermediate
stages are found. An attempt therefore to define genera with reference to the
tuberculation of the ambulacra, necessitates putting some specimens of acutus
in one genus and some in another, while some would be exceedingly hard to
place. A similar difficulty arises when stress is laid on the condition of the
gill-cuts. Thus in Lytechinus l variegatus, the adults have deep and well-marked
gill-cuts while in L. semituberculatus they are smaller and less distinct and in
L. verruculatus they are scarcely specialized at all. Moreover as Mortensen
has pointed out ("Ingolf" Ech., pt. 1, p. 115) in small specimens of variegatus,
the gill-cuts are no more noticeable than in many other species. There can be
no question however, that the absence of primary tubercles on many ambulacral
plates and the presence of sharply defined gill-cuts are evidences of specialization
and while their occurrence in Echinus and Lytechinus respectively is not of any
value for the definition of those genera, in Toxopneustes and Tripneustes, where
they have become fully differentiated and fixed, they are important generic
characters.

1 For the use of this and other generic names, see below under the respective genera.

240 HAWAIIAN AND OTHER PACIFIC ECHINI.

The arrangement of the ocular plates with reference to the periproct, a
morphological character so beautifully worked out in Jackson's recent mono-
graph (1912, Mem. Boston Soc. Nat. Hist., VII., p. 86-164) is of great impor-
tance in tracing the lines of development in the Echinidse, but it is impossible to
rely on it alone. The same is true of the amount of calcification in the buccal
membrane, the relative sizes of the spines and tubercles and the characters of
the globiferous pedicellarige, though all of these are of the greatest help. Finally
it may be emphasized that, as Mortensen has pointed out, color is often a sug-
gestive character, and although in certain species (as Echinus acutus and Lytechi-
nus variegatus) it is very variable, in most cases it affords a good deal of assistance
in determining specific limits.

Having thus indicated the characters in which the specific differentiation
of the Echinidoe is best shown, it is desirable to point out what seem to have been
the lines of development and the resultant, most highly specialized genera.
There can be little question that the two species of Psammechinus (Echinus
miliaris Gmel. and E. microtuberculalus Bl.) are the least specialized members
of the family, at present known. This is shown by the uniform series of am-
bulacral plates, the small exsert oculars, the absence of distinct gill-cuts, the
heavily plated buccal membrane, the slight differentiation of primary spines
and tubercles, and the character of the globiferous pedicellarire. From some
such stock, the species of Echinus have undoubtedly come and it is not difficult
to trace possible lines of differentiation. It seems probable that Alexandri
is not far from the ancestral stock, as shown- by the ambulacra and abactinal
system, though the buccal membrane, the primary spines, and the globiferous
pedicellarise, all show considerable specialization. Although Wallisii, atlanticus,
and gracilis all show a very high degree of specific differentiation, it seems likely
they originated from the Echinus-stock near Alexandri. Echinus acutus is the
most highly variable species of the family, and while some of its forms are very
near Alexandri, it shows tendencies towards specialization in several different
directions. The resulting extremes are so utterly unlike, one would never
consider them conspecific were it not for the completeness of the intermediate
series. From acutus as a centre, such species as esculentus and tenuispinus have
arisen on the one hand by marked changes in the ambulacra without accompany-
ing changes in the abactinal system, while on the other hand margaritaceus has
developed with less specialization of the ambulacra but with greater changes
in the abactinal system. From similar stock the development, through E.
armatus, of Evechinus, the most specialized genus of this branch of the family,

ECHINID.E. 241

well characterized by its remarkable ambulacra, abactinal system, and pedicel-
larise, can be traced. From Psammechinus, another line of development runs
out through Lytechinus to Toxopneustes and Tripneustes, characterized by
increasing specialization of the ambulacra, abactinal system, gill-cuts, and
pedicellariae. It is not easy to draw a line between Psammechinus and Lytechi-
nus, as the less specialized members of the latter genus have no noticeable gill-
cuts and only one ocular, or none, insert. For convenience, we have drawn an
arbitrary line based chiefly on the specialized globiferous pedicellaria? of Lyt-
echinus. It is probable that Gymnechinus, with its remarkable abactinal
system and globiferous pedicellarise, has been differentiated from Lytechinus
through such a group as Nudechinus, which is sharply distinguished from
Lytechinus by the thin, naked buccal membrane. From Psammechinus again,
still a third group of species has developed in the far south, for which Mortensen's
name Parechinus is employed. Of this group, angulosus is nearest the original
stock in its ambulacra, spines, and pedicellarise, but shows decided specialization
in the abactinal system and buccal membrane; annulatus is a very close ally.
In magellanicus, the buccal membrane shows further loss of calcareous matter
and the globiferous pedicellarise, though variable, are more specialized than in
angulosus. The most extreme member of this group is huttoni, as shown by the
globiferous pedicellarise and the tuberculation of the test.

While it is impossible to show in a linear arrangement, the relationships of
the genera as here pointed out, the following table will indicate the limits which
are assigned to each one. It will be observed that it is not possible to
use the generic names Sterechinus Koehler, Pseudechinus Mortensen, Proto-
centrotus Doderlein, Notechinus Doderlein, or Selenechinus de Meijere. It is
to be regretted that no one of them is available for the little group of species for
which the name Nudechinus is herein proposed.

Periproct approximately central; if two oculars reach it, they are typically I and V.
Buccal membrane more or less heavily plated (except L. rufus) ; a primary tubercle

on every ambulacra! plate.

Oculars plates all exsert; abactinal system well covered with tubercles; gill-
cuts insignificant; valves of globiferous pcdicellarise with lateral teeth . . Psammechinus.
Ocular I often, and in some species ocular V also, insert or nearly so; abactinal
system usually with few tubercles; gill-cuts, in specimens over 30 mm. h. d.,
usually deep and sharply defined; valves of globiferous pedicellariae with-
out lateral teeth Lytechinus.

Buccal membrane not plated but with more or less numerous, small, scattered

plates or none in addition to primordial ambulacrals.
Gill-cuts shallow and not sharply defined.

Poriferous areas not very broad; pore-pairs in arcs of three.

242 HAWAIIAN AND OTHER PACIFIC ECHINI.

Ocular plates small, completely excluded from periproct (except in
margaritaceus) ; primary spines much longer, and primary tuber-
cles much larger, than secondaries; buccal membrane with more
or less numerous plates; valves of globiferous pedicellariae with
small lateral teeth on each side near tip (except in armatus)',
spines unicolor or becoming lighter at tip; size usually large,

often exceeding 75 mm. h. d Echinus.

Ocular I usually, ocular V not rarely, nearly or quite insert; a
primary tubercle on each ambulacral plate; primary spines
and tubercles not especially conspicuous; buccal membrane
thin and usually bare (except in angulosus and annulatus);
size small rarely exceeding 40 mm. h. d. and usually under 30.
Valves of globiferous pedicellaria; with one or more lateral teeth Parcchinus.
Valves of globiferous pedicellaria; with no lateral teeth . . . Nudechinus.
Poriferous areas very broad; pore-pairs at ambitus in three distinct
vertical series; a primary tubercle only on every second, third, or

fourth ambulacral plate Evechinus.

Gill-cuts deep and sharply defined; a primary tubercle usually only on every
second, third, or fourth ambulacral plate; valves of globiferous pedicel-
lariae with no lateral teeth.
Poriferous area not one half as broad as interporiferous; pairs of pores in

arcs of three Toxapneusles.

Poriferous area more than half as broad as interporiferous; pairs of pores

in three, more or less well separated vertical series Tripneustes.

Periproct excentric at right, with oculars I and II insert Gymnechinus.

PSAMMECHINUS.

Agassiz and Desor, 1846. Ann. Sci. Nat., (3), VI., p. 368.
Type-species, Echinus miliaris Gmelin, 17SS. Linn<5 Syst. Nat., ed. 13, p. 3169.

Mortensen's extraordinary course in selecting "Echinus variegatus Lam'k." :
as the type of this genus and proposing a new generic name for miliaris and its
allies ("Ingolf" Ech., pt. 1, p. 108, 114) is so contrary to all the accepted rules
of nomenclature, that it is necessary to discuss the matter fully here. The name
Psammechinus was proposed by Agassiz and Desor in 1846 for a subgenus of

1 The argument of Lambert (1906, Mem. Geol. Soc. France, XIV, p. 66, footnote 3) that the use of
the name variegatus for this common and well-known species is not justifiable, would be unassailable
were it not for an error in the first premise, that "Cidaris nariegata Leske est un re'alitd un Tripneustes."
There can be little question that Klein's figure, to which Leske refers, is a Tripneustes, but it is equally
certain that the figures of Seba and Gualtieri to which Leske also refers are the West Indian species,
long known as Toxopneustes variegatus, and a careful reading of Leske's description shows that it is
the latter he is describing and not a Tripneustes; his description of the ambulacra clearly proves this.
Moreover Leske and all the earlier writers (except Klein) refer to the green and white coloration, and so
far as known, no one has ever found a green and white Tripneustes. It seems clear therefore that varie-
galus must be retained as a specific name but should be credited to Leske 1778, instead of to Lamarck,
1816. The nomenclature of the genus Tripneustes will not suffer, for the species to which Leske's name
variegata. was long applied, was described by Linn6, and thanks to Loven's work, has now for more than
twenty years borne its Linnsean name, gratilla.

PSAMMECHINUS. 243

Echinus, into which they put a rather heterogeneous group of species. In 1855,
Desor raised the group to generic rank and in 1862, Dujardin and Hupe rear-
ranged the species, but in neither instance was there any revision of the group
attempted. In 1863 however both A. Agassiz and Liitken, quite independently
recognized the unnatural association of species in the group, and removed from
it the species with deep gill-slits. Agassiz, whose paper has a few months
priority, is clearly the "first reviser" and he definitely restricts Psammechinus
to the forms with shallow gill-slits (though he gives no diagnosis) and names
miliaris as the first species. To the species with deep gill-slits he gives the name
Lytechinus, and under it names three species, all of which Mortensen, and all
other recent writers, regard as synonyms of variegatus. While the action of
Agassiz does not settle the type of Psammechinus, it does forever preclude the
use of variegatus as the type of that genus; unless indeed the name is used in the
same sense and with the same contents as when originally proposed! Liitken's
paper, while entirely in agreement with Agassiz's, very naturally gives a different
name (Psilechinus) to the variegatus group, which of course is a synonym of
Lytechinus, but he also fails to designate a type for the restricted Psammechinus.
He suggests verruculatus as a typical example of the genus and it might have been
accepted as the type, were that not impossible since verruculatus is not among
the species known to Agassiz and Desor, and therefore is not in their genus.
In 1867, Verrill (Trans. Conn. Acad., I, p. 302) definitely designates "Echinus
variegatus" as the type of Lytechinus. In 1869, Pomel (Rev. des Ech., p. 42)
says that Psammechinus Agassiz and Desor is not a homogeneous group and
ought to be restricted to the type of miliaris and microtuberculatus. Schiz-
echinus is very unnecessarily proposed for variegatus and its allies. It seems
impossible to doubt that at the time of the publication of the "Revision of the
Echini" (1872) all students of the Echini were agreed that the name Psammechi-
nus belonged to miliaris and its allies, while variegatus typified a very different
group, for which three different names had been suggested, the earliest being
Lytechinus A. Ag. In the "Revision of the Echini" Mr. Agassiz thought best
to unite Psammechinus with Echinus, and Lytechinus with Toxopneustes and
these unions have been almost universally accepted. Mortensen (1903, " Ingolf "
Ech., pt. 1, p. 106, 114) has however shown excellent reasons, in the structure
of the ambulacra and in other characters for separating the groups thus united,
but it is clearly impossible to follow his nomenclature. While it seems possible
to argue as to whether any type has ever hitherto been definitely and correctly
assigned to Psammechinus, the type of Lytechinus is beyond doubt. Lambert's

244 HAWAIIAN AND OTHER PACIFIC ECHINI.

(1906, Mem. Geol. Soc. France, XIV., p. 67) and Mortensen's (1907, "Ingolf"
Ech., pt. 2, p. 174) statements that Anapesus Holmes must take precedence
over Lytechinus for variegatus and its allies, shows that neither of them has
consulted Holmes's paper (1860, Post-Pliocene Fossils S. Carolina, p. 5, PI. II,
fig. 2) but each has been misled by Pomel. The mistake of the latter was doubt-
less due to the unfortunate error in the "Revision" (p. 167, 168, and 172) by
which Anapesus Holmes is made a synonym of Toxopneustes instead of being
assigned to Arbacia, to which it is correctly referred on p. 72 of the "Revision."
Holmes's excellent figure permits no doubt on this point. It is most fortunate
that miliaris may be accepted as the type of Psammechinus for of all the species
included by Agassiz and Desor in their subgenus, it is the only one to which the
diagnosis given, accurately applies. Mortensen's attempt to fit that diagnosis
to variegatus is scarcely convincing.

As herein limited, Psammechinus includes only the two well-known European
species, which may be distinguished from each other as follows : —

Buccal membrane well covered with whitish plates; tuberculation of test, coarse;

secondary tubercles very large miliaris.

Buccal membrane completely covered by green or greenish plates; tuberculation of

test, fine; primary tubercles much larger than secondaries microtuberculatus.

LYTECHINUS.

A. Agassiz, 1863. Bull. M. C. Z., I, p. 24.
Type-species Cidaris variegata Leske, 1778. Add. ad Klein, p. 85.

It is not necessary to repeat here what has just been said under Psammechinus,
regarding the type of this genus. Careful study of large series of specimens
from Bermuda, South Carolina, Florida, Yucatan, various West Indian islands,
and Brazil shows that the Bermudian form, originally described under the name
of atlanticus, which has recently been reinstated by Jackson (1912, Mem. Boston
Soc. Nat. Hist., VII, p. 121), can hardly be maintained as a valid species.
Typical specimens from Bermuda are strikingly different from Floridian and Caro-
linian specimens, while in both regions, the characteristic green and white West
Indian form seems to be unknown. We have specimens from Brazil however,
which are much like the Bermudian form, while many Bermudian specimens are
distinctly green. It seems clear that variegatus is a highly variable species, which
in Bermuda is developing into a very slender spined, deep purple form, while
along the continental coast it is becoming stout spined and deep pink. In the
Brazilian region, variation does not seem to have become fixed in any special line.

LYTECHINUS. 245

To indicate these facts it seems desirable to use subspecific names, and the
characteristic Bermudian Lytechinus is accordingly designated as L. variegatus
atlanticus A. Ag., while that from the continental coast is called L. variegatus
carolinus A. Ag.

In attempting to draw lines of division between this genus and Psammechinus
on the one hand and Toxopneustes on the other, there is very great difficulty
as the species of these genera appear to form an almost unbroken intergrading
series. The only sharp line appears to be in the globiferous pedicellarise, which
in miliaris and microtuberculatus are quite simple with lateral teeth, while in all
the other species, the lateral teeth are wanting and the blade is tubular with a
long terminal tooth which reaches its most extreme development in Toxopneustes.
By this character, a number of small tropical species are separated from Psam-
mechinus, with which genus they are otherwise quite closely allied. As however
they all show a more or less decided tendency to have one or two oculars insert
and the gill-cuts more sharply defined, it has seemed justifiable to associate
them with Lytechinus. The line of division between this genus and Toxop-
neustes is not suggested by the pedicellarise but is based on the reduction of
calcareous matter in the buccal membrane and the increased specialization of
the ambulacra in Toxopneustes. In the latter character however, the Japanese
species, T. elegans Dod., is very little advanced over Lytechinus, while the calcifi-
cation of the buccal membrane is more or less variable in both genera.

As used here, Lytechinus is a somewhat heterogeneous group of nine species,
of which four are here described for the first time. Of the other five, two (varie-
gatus Leske from the West Indian region and semituberculatus Agas. and Des. from
the Panamic region) have long been known as congeneric species, while the rela-
tionships of the third and fourth (verruculatus Ltk. and rufus Bell from the Indo-
Pacific region) and the fifth (pictus Verr. from Lower California) have only
recently been made clear. Mortensen (1903) first pointed out the relationship
of verruculatus to Lytechinus and (in 1904) of rufus to verruculatus, while pictus
has up to the present time been confused with other species. It is therefore
a pleasure to now give it its rightful position as a valid species of this genus.
The nine species are easily distinguished from each other when adult, but young
specimens of the larger species are not always easily recognized, and since the
specific characters shown by the abactinal system and the gill-cuts are not
assumed fully until the individuals are 8-10 mm. in diameter or even larger, it
has been found necessary to base the distinctions of the following table largely
on color. Not having seen specimens of Bell's " Salmacis rufa" follow

246 HAWAIIAN AND OTHER PACIFIC ECHINI.

Mortensen who has examined the type and other specimens, in regarding it as
nearly related to verruculatus. Bell says nothing about the buccal membrane;
de Meijere says it is "sehr nacht"; Mortensen that it contains a "great number
of small irregular plates." It seems probable that, as Mortensen suggests, this
species serves as the connecting link between Lytechinus and the group here
called Nudechinus. It is not possible to determine whether the species dyscritus
and callipeplus are based on full grown specimens or not but it does not seem
probable that they are.

Test white or whitish more or less shaded with yellowish green abactinally ; spines
white or whitish, often yellowish or yellow-green at base; oculars small,
all exsert or rarely I insert.
Test high; primary tubercles small, little larger than secondaries; primary

spines short, rather stout; size moderate, up to 30 mm. h. d euerces.

Test much flattened; primary tubercles very conspicuous; primary spines

long; size small, less than 10 mm. h. d dyscritus.

Test and spines not as above.

Oculars all exsert or sometimes I insert; test variegated with reddish or red;
primary tubercles, at least abactinally often reddish; primary spines
more or less red or with red bands.

Spines pale red, not banded callipeplus.

Many primary spines with 2 or 3 red rings rufus.

Oculars relatively large, not usually all exsert; test and spines not as above.

Spines light with 1-4 faint, narrow rings of brown, dull green or pinkish;

oculars I and V usually insert; size moderate, rarely exceeding

30 mm. h. d verruculatus.

Spines not as above.

Oculars variable, often exsert but often I, or I and V insert; test
depressed, low; primary tubercles, at least abactinally, often
very dark; spines dull greenish or light; size moderate, seldom

exceeding 25 mm. h. d anamesus.

Oculars I and V usually insert, and sometimes IV also; test and
spines not as above; gill-cuts in adults usually deep and
sharply defined.

Spines bright yellow-green (K. & V. 282) ; in very young indi-
viduals they are darker and are tipped (sometimes banded
also) with white; abactinal interambulacral areas very bare,
most of the plates above ambitus carrying no tubercles
between the two rows of primaries; size moderate, up to

45 mm. h. d semituberculatus.

Spines not bright yellow-green; abactinal interambulacral areas

with secondary tubercles.

Abactinal interambulacral plates with many secondary
and miliary tubercles, only the extreme inner end of
each plate, bare; coloration prevailingly rose-purple
when young, becoming paler and duller with age, with
no indications of either green or deep violet; primary

LYTECHINUS EUERCES. 247

spines stout and blunt; size moderate, up to 40 mm.

h. d piclus.

Abactinal interambulaeral plates largely bare, usually with
only 3 or 4 small tubercles; primary spines slender
and pointed; size large, up to 80 mm. h. d.
Color variable, but rarely deep violet ; primaries mod-
erately slender; ocular IV seldom insert (in
only 8 per cent of the specimens) .
Colors prevailingly green and white; primary

spines rather slender variegatus typicus.

Color prevailingly dull pink; primary spines

stouter variegatus carolinus

Color deep, rich violet; primary spines very slender;
ocular IV often insert (28 per cent of the speci-
mens) variegatus atlanticus.

Lytechinus euerces,1 sp. nov.
Plates 93, figs. 4, 5; 98, figs. 3, 4; 107, figs. 4-6.

The largest of the specimens (PL 107, figs. 4-6} measures 32 mm. in diameter
while the height of the test is 22.5 mm. There are 18 interambulaeral plates in
each column and 23 ambulacral plates in each half-area. The longest primary
from just below the ambitus, is only about 6 mm. in length. The abactinal
system is 8 mm. in diameter and the actinostome 12 mm. A smaller specimen,
18 mm. in diameter, has the test 12 mm. high, the abactinal system less than 5 mm.
across and the actinostome 8 mm. There are 14 interambulaeral and 16 ambula-
cral plates in each column. In a still smaller specimen, 10 mm. in diameter and
6 mm. high, the abactinal system is 4 mm. across, the actinostome 6 mm., and
there are 8 interambulacral and 10 ambulacral plates in each column.

The periproct (PL 98, fig. 4) is moderately large, decidedly exceeding a
genital plate in area. It is covered by a dozen or more plates, of which those
adjoining the anus are quite small; the outer series are much bigger and the one
adjoining the left anterior genital plate is so much the largest it may properly
be called a suranal; in the largest specimen, it carries a well-developed secondary
tubercle and spine. The genital plates are of approximately equal size, though
the madreporic plate may be a trifle the largest. They usually form a closed
ring, but 1 and 5 may be separated. They are noticeably bare, carrying only
1-3 secondary tubercles, situated near the proximal margin. The pores are
circular and quite large, in the distal part of the plate. The oculars are small

= well protected.

248 HAWAIIAN AND OTHER PACIFIC ECHINI.

and bare, carrying only 1-3 very small tubercles. They are usually all exsert
but in one specimen I is distinctly insert and in another it is nearly so.

The interambulacral plates at the ambitus are rather more than twice as
wide as high. Each plate carries a small imperforate, non-crenulate, primary
tubercle and 6-8 well-spaced secondaries, some of which are nearly as large as
the primary. Miliary tubercles are small and abactinally they are few and very
indistinct; actinally they become more numerous and better defined. The
ambulacral plates are high, those at the ambitus half as high as wide and those
near the ocular plate as high as wide. Each plate carries a small primary tubercle
and 2-4 secondaries one of which is often nearly equal to the primary. The
pore-pairs are small, well spaced and form a nearly vertical series, especially
abactinally. The poriferous area is thus very narrow, occupying only the outer
third of each plate, though near the actinostome it becomes a little wider. At
the ambitus, the interambulacra are not quite twice as wide as the ambulacra.

The buccal membrane (PI. 98, fig. 3) is heavily plated. The primordial
ambulacrals are approximated in pairs and are distinctly larger than the other
plates. All the plates are thick, white, and polygonal and carry more or less
numerous pedicellarise. In the small specimens, there are practically no gill-
cuts but in the large ones, the cuts are well defined though not very deep.

The primary spines are remarkably short, slender, and pointed. They
show under the lens about a dozen longitudinal striations. The milled ring is
only imperfectly developed. The secondary spines are similar but much more
blunt. Miliaries are few and scattered, and are remarkably long and slender.

Pedicellarise are abundant, the tridentate being the most common. The
globiferous pedicellarise are fairly common. The valves (PI. 93, fig. 4) are
slender, about .65 mm. long, with the base not quite half so wide. The stalks
are very slender, two or three times as long as the head, and there is no neck.
The heads contain numerous spicules which are distinctly bihamate and not
dumb-bell shaped; at least none of the latter were seen. The tridentate pedicel-
larise show great diversity in size, the valves (PI. 93, fig. 5) ranging in length
from .35 to 1.25 mm. The blade is moderately broad, rounded at the tip and
contains a small amount of calcareous mesh-work. In the small tridentates,
the valves are more compressed, there is no mesh-work, the margins are somewhat
more dentate and the tip is bent in more. The ophicephalous are not rare and
are rather conspicuous as the head is quite heavy and the stalks are long and
stout. The valves are somewhat constricted but not markedly so. They
measure about .40 mm. besides the loop which may be half as much again. The

LYTECHINUS DYSCRITUS. 249

triphyllous are very small and are common. The valves measure about .15 mm.
in length and are two thirds as broad. The sphceridia are spheroidal or less
commonly ellipsoidal ; they occur only on the actinal part of the ambulacra and
five or six may be grouped at the peristome. They seem to be perfectly smooth.

The coloration of this species is variable only in the amount of green. The
general appearance of both test and spines is creamy white, while the plated
buccal membrane is clear white. Abactinally, particularly along the sides of
the interambulacra, there is a more or less marked shading of light green; in
some specimens this is very marked, while in others it is almost wanting. Above
the ambitus, the primary spines are as a rule, light green at base fading out into
white at the tip. In some specimens, the spines are mostly green, but usually
the white greatly predominates.

This interesting species was collected by the "Blake" in the West Indian
region, but the specimens were identified either as Trigonocidaris albida or as
Toxopneustes variegatus, although in most cases the doubtfulness of the identifica-
tion was indicated on the label by one or more question marks. The resemblance
of small specimens to Trigonocidaris albida is very marked, the absence of the
sculpturing of the test being the only important difference, although the differ-
ence in color is more or less evident. From Lytechinus variegatus, this species
is easily distinguished by the exsert oculars and the very short primary spines,
as well as by the coloration. The geographical range of euerces as indicated by
the "Blake" collection is from the Gulf of Mexico, northwest of the Tortugas,
to Barbados, but three fourths of the thirty-nine specimens are from the vicinity
of St. Lucia and Barbados. The bathymetrical range is from 84 to 300 fathoms.

Lytechinus dyscritus,1 sp. nov.
Plates 93, fig. 10; 96, figs. 1-3.

The larger of the two available specimens, which are nearly of a size, is 7 mm.
in diameter and only 3 mm. high. There are 8 interambulacral plates in each
column and 9 ambulacral. The longest primary, from the ambitus, is about
5 mm. long. The abactinal system is 3 mm. in diameter and the actinostome is
3.5 mm. across.

The periproct (PI. 96, fig. 3) is rather large, distinctly bigger than a genital
plate and is covered by 4 or 5 large plates, of which the one adjoining genital 3

= hard to determine.

250 HAWAIIAN AND OTHER PACIFIC ECHINI.

is largest, covering half of the area or more. The genital plates are approxi-
mately equal and broadly in contact with each other. Each carries one second-
ary tubercle on the proximal margin, and occasionally there is a miliary tubercle
beside it. The madreporic genital is perforated by only a few pores and these
are situated in an elevation near the centre of the plate. The oculars are very
small, all broadly exsert; each one carries a small secondary tubercle or two.
There are no genital pores visible but with a lens, the ocular pores can be seen
close to the distal margin of the plates.

The interambulacral plates at the ambitus are rather high, but the greater
part of the surface is covered by the big primary tubercle, around which a few
minute secondary tubercles are scattered. The ambulacral plates are high but
their primary tubercles are very unequally developed. On some plates they are
very large occupying most of the surface, while on others they are so small as to
be more naturally called secondaries. There are usually more primaries de-
veloped on one side of the area, than on the other, giving the ambulacra a one-
sided appearance. The pore-pairs are very small, three for each plate, and the
poriferous area is a narrow, nearly straight, vertical line. At the ambitus, the
ambulacra are about three fourths as wide as the interambulacra.

The buccal membrane (PI. 96, fig. 2) is thin and although well covered with
plates, it is not heavily plated as in euerces. The primordial ambulacral plates
are much larger than the others and form a nearly closed ring. There seem to be
no spines or pedicellarise borne by any of the actinostomal plates. Gill-cuts are
scarcely to be detected at all.

The primary spines are noticeably long and correspondingly conspicuous, but
are rather slender and taper to a point. Those at the ambitus are the longest
and those of the ambulacra are smaller than the interambulacral. The secondary
spines are very few; they are long and slender.

Pedicellarise appear to be very scarce. The globiferous valves (Plate 93,
fig. 10) have the blade tubular and terminating in a single tooth, without lateral
teeth. They are relatively large, measuring about .45 mm. in length. The
ophicephalous do not show any special peculiarities but the valves are constricted
above the base. No tridentate or triphyllous pedicellarise were found, neither
were there any sphgeridia, or spicules in the tube-feet. No doubt the poverty
of the available material accounts in large part for this lack.

The ground color of this species is white but abactinally there is, under a lens,
a more or less evident yellow-green shade. This color is most marked on the
outer ends of the genital plates, on the periproctal plates, along the sides of the

LYTECHINUS CALLIPEPLUS. 251

interambulacra, especially around the tubercles or in the tubercles themselves,
and at the bases of the abactinal spines.

The two specimens upon which this species is based are labelled " Echino-
metra lucunter, Florida. L. Agassiz," but in the catalogue both name and
locality are followed by a question mark. The specimens are obviously imma-
ture, but the description and figures here given show clearly they do not belong
in Echinometra. The buccal membrane is not plated nearly so heavily as in
other species of Lytechinus, but the globiferous pedicellarise indicate a probable
relationship with that group. When compared with specimens of L. variegatus
of the same size, the shape of the test, the characters of both abactinal system
and buccal membrane, the long primary spines, and the coloration all serve to
show that these little Echini cannot be the young of that species. And I have
failed to find any other known species to which they show any closer relation-
ship. They may prove to be the young of euerces, but the long primary spines,
the conspicuous primary tubercles and the low, flat test separate them rather
sharply from that species, as at present known.

Lytechinus callipeplus,1 sp. nov.
Plate 96, figs. 4-6.

The largest specimen is a bare test, 11.5 mm. in diameter and 6 mm. high.
The abactinal system is 4 mm. in diameter and the actinostome is 6 mm. across.
The type is 8.5 mm. in diameter and 5 high, with the abactinal system not quite
3 mm. across and the actinostome, 4.5; the longest spine is 2.5 mm. long. In a
smaller specimen, 6.5 mm. in diameter, the longest spines are 3 mm. long. In
the largest specimen, there are 11 or 12 interambulacral, and 12 ambulacra!
plates in each column, while in the type, there are 10 or 11 and 10 respectively.
In a specimen, 7 mm. in diameter, the numbers are 9 and 9.

The periproct (PI. 96, fig. 6) is moderately large, decidedly larger than a
genital plate. It is covered by about four plates of which the one adjoining
genital 3 is largest. The genital plates are about equal and are about as high
as broad. Each one carries a small tubercle near the proximal margin and
there may also be one or two miliary tubercles on the plate. The genital pore
is large and distinct at the distal end of the plate. The madreporic genital is
fairly well covered with the minute pores of the water system. The oculars are

1 Ka\\ur£7rAo« = beautifully robed.

252 HAWAIIAN AND OTHER PACIFIC ECHINI.

large, and broadly exsert, except I, which is insert or nearly so in specimens more
than 7 mm. h. d. Each ocular carries 1-3 small tubercles. The ocular pore is
well developed near the distal margin of the plate.

The interambulacral plates are high, only those at the ambitus being nearly
twice as wide as high. Each plate carries a small, but well-developed primary
tubercle and 4-9 secondary tubercles, one or two of which approach the primary
in size. The ambulacral plates are remarkably high, as high as wide or higher.
Each plate carries a primary tubercle and 2-4 small secondaries of variable size.
The pores are relatively large, three pairs in each plate, and are placed close to
the interradial margin. The poriferous area is narrow and nearly straight.
The ambulacra are about three fourths as wide as the interambulacra at the
ambitus.

The buccal membrane (PL 96, fig. 5) is very fully covered with thick white
plates, among which the primordial ambulacrals, although distinctly the largest,
are not conspicuous. The larger plates all carry pedicellarise in small numbers,
often only one to a plate. The gill-cuts are sharply defined but are not very deep.

The primary spines are rather short, and though relatively thick at the base,
taper rapidly to a blunt point. They are finely, longitudinally striated. The
scattered secondaries are short, thick, and pointed.

Pedicellarise are not at all common, though the ophicephalous, because of
their white color and long white stalks may be rather conspicuous. The globif-
erous pedicellarite are small and show no special peculiarities. The valves have
the usual tubular blade without lateral teeth and measure .25-.30 mm. in length.
The ophicephalous show considerable variation in size; the valves, which are
somewhat constricted above the base, range from .14 to .25 mm. in length, while
the articular loop adds .03-.07 more. The stalks are stout and long, exceeding
many of the secondary spines. No triclentate or triphyllous pedicellarise and
no sphseridia were observed. The calcareous spicules are distinctly bihamate.

The color of the test is quite variable; actinally it is white like the plated
buccal membrane but abactinally it becomes pale brownish. The median
ambulacral and interambulacral areas abactinally are more or less brick-red or
less commonly reddish brown. There are traces of red also on the abactinal
system. In some specimens, there are patches of light green at the ambitus
and a similar color may sometimes be noted on the genital plates. The primary
tubercles, abactinally at least, are more or less reddish. The plates of the peri-
proct are white. The primary spines are pale red, lightest at the tip ; those on
the actinal side, especially in young specimens, show faint bands of red.

LYTECHINUS VERRUCULATUS. 253

This pretty little species was collected by the "Blake" in the West Indian
region, but many of the specimens were not distinguished from Genocidaris
maculata, a species with which they were often taken and from which they are
not easily distinguished, unless the specimens are dry. The coloration is very
distinctive even though variable, while the small number of ambulacral plates
as compared with the interambulacral is quite remarkable. The relationship
of this species to both euerces and variegatus is obvious, but it cannot be confused
with either. It was taken by the "Blake" only in the vicinity of Dominica,
Grenada, and Barbados, in 69 to 170 fathoms of water. Of the seventeen
specimens at hand twelve are from Barbados,

Lytechinus verruculatus, comb. nov.

Psammechinus verruculatus Llitken, 1864. Vid. Med., p. 166.

It seems probable that Mortensen is right in associating this species with
variegatus. Although none of the specimens before me are nearly so large as
those which de Loriol had from Mauritius, there seems to be no doubt of their
identity. The "Albatross" specimens reveal the same peculiarities of coloration
described by Mortensen (1904, Siam Ech., p. 123). In the smallest specimen
the rings on the spines are red and the same color appears on small spines of
larger specimens. Many primaries are red at base while in a few specimens,
the base of the primaries is violet. There is great diversity in the depth of the
colors marking the test. Although all the "Albatross" specimens are immature
ranging only from 5 to 12 mm. in diameter, the seven examined in regard to the
ocular plates showed the adult character nearly acquired. For five have
ocular I broadly insert, and in three specimens, V also is almost in. Of the
other two specimens one is a very rare variant, with only ocular IV insert and
the other is a much more interesting variant with I and II insert as in Gym-
nechinus; the periproct is not however excentric.

Station 3847. Off Lae-o Ka Laau Light, Molokai, Hawaiian Islands. Bott.
temp.? 23-24 fathoms. S., st.

Station 3871. Off Mokuhooniki Islet, Molokai, H. I. Bott. temp.? 13-43
fathoms. Fne. wh. s.

Station 3872. Off Mokuhooniki Islet, Molokai, H. I. Bott. temp. 74.6°.
32-43 fathoms. Yl. s., p., co.

Station 3955. Off Laysan Island, H. I. Bott. temp. 74°. 20-30 fathoms.
Co., r., alg.

254 HAWAIIAN AND OTHER PACIFIC ECHINI.

Station 3970. Off French Frigate Shoal, H. I. Bott. temp.? 17-17?
fathoms. Crs. s., sh., co.

Station 4031. Off Diamond Head, Oahu, H. I. Bott. temp.? 27-28
fathoms. Fne. co. s., for., co.

Station 4032. Off Diamond Head, Oahu, H. I. Bott. temp.? 27-29
fathoms. Fne. co. s., for., co.

Station 4149. Off Modu Manu, H. I. Bott. temp. 77.7°. 33-71 fathoms.
Co., corln.

Station 4162. Off Modu Manu, H. I. Bott. temp.? 21-24 fathoms. Co.

Station 4168. Off Modu Manu, H. I. Bott. temp. 78.3°. 20-21 fathoms.
Co. s., for.

Bathymetrical range, 13-71 fathoms. Extremes of temperature, 74°-78.3°.

Sixteen specimens.

Lytechinus anamesus,1 sp. nov.
Plates 99, figs. /,, 5; 107, figs. 7-11.

The largest specimen measures 25 mm. in diameter and 13 mm. high. It has

15 interambulacral and 19 ambulacral plates in each column. The abactinal
system is 8 mm. in diameter and the actinostome is 9 mm. across. The longest
primary is 15 mm. long. In a specimen 11 mm. in diameter and 5 mm. high,
the abactinal system is 4 mm. and the actinostome 5 mm. in diameter. There
are 10 interambulacral and 12 ambulacral plates in each half-area. A specimen
17 mm. in diameter is 9 mm. high, while another, 18 mm. horizontally is only 8
mm. vertically. A specimen 18 mm. in diameter has the primary spines about

16 mm. long, while in another, 17 mm. in diameter they only measure 9 mm.
The smallest specimen in the series is 7 mm. in diameter and 3.5 mm. high. The
abactinal system measures 2.5 mm. across and the actinostome 4 mm. There
are in each column, 10 interambulacral .and 12 ambulacral plates and the primary
spines are about 2 mm. long.

The periproct (PI. 99, fig. 5) is large, much larger, as a rule, than a genital
plate. It is covered by a number of plates (4-10 in the young, 12-20 in adults)
of which one, adjoining genital 3, is usually much the largest. In adults, many
of the plates carry small tubercles. The genitals are of approximately equal,
moderate size, about as high as wide, and often form a closed ring, but often
1 and 5 are separated and often 5 and 4 also. Each plate has a large genital

= intermediate.

LYTECHINUS ANAMESUS. 255

pore near the "distal tip, while on the proximal margin it carries a secondary
tubercle. There may be one or rarely two other tubercles on the plate but
otherwise its surface is quite smooth. The madreporic genital is conspicuous,
with a large group of pores. The ocular plates are large and each carries at
least one secondary and a number of miliary tubercles. The pore is evident
near the distal margin. Of fifty specimens examined, all exceeding 15 mm. h. d.,
19 (38%) have no oculars insert, 17 (34%) have ocular I insert, 11 (22%)
have oculars I and V, while 3 (6%) have ocular V only. Of the ten largest
specimens, three have no oculars insert, three have I, three have I and V, and one
has V alone. Where such diversity is shown, it is clear that the oculars do not
furnish a very helpful specific character.

The interambulacral plates at the ambitus are low, the width exceeding twice
the height. Each plate carries a conspicuous primary tubercle, which, at and
below the ambitus, is accompanied by two or three secondaries and several
miliaries; one of the secondaries is sometimes almost as large as the primary.
Above the ambitus, there are few secondaries and only scattered miliaries, and
the uppermost half dozen plates have their inner ends noticeably smooth and
bare. The ambulacral plates are relatively higher at the ambitus than the
interambulacral. The arrangement of their tubercles is essentially the same,
so that, while the test is well covered with tubercles actinally, there are ten
distinct, bare, areas abactinally, radiating out from the periproct. The am-
bulacra are relatively wide, about four fifths as wide as the interambulacra, at
the ambitus. The pore-pairs are in distinct arcs of three and as the pores are
large, the poriferous areas are relatively broad.

The buccal membrane (PI. 99, fig. 4) is heavily plated, especially in the young;
in some adults, the distal plates show indications of resorption and the membrane
is occasionally visible between them. The primordial ambulacral plates are
much larger than any of the others and form a well-spaced ring. They carry
pedicellarise, as do some of the larger plates distal to them. The gill-cuts are
fairly well defined but are not deep.

The primary spines are of variable length but are quite conspicuous. They
are longest at the ambitus, where their length may nearly equal the diameter
of the test. They are slender and taper very gradually from the low, incon-
spicuous milled-ring to the rather blunt point. Secondaries and miliaries are
similar, save for their much smaller size, but they are not abundant anywhere.

Pedicellarise of all kinds are abundant but are not especially distinctive.
The globiferous have valves about .60 mm. long, of which the base is about one

256 HAWAIIAN AND OTHER PACIFIC ECHINI.

half; the width of the base is .30 mm. The blade is tubular, ending in the
usual sharp tooth, which may be as much as .20 mm. long. The ophicephalous
are conspicuous because of their long, stout stalks and heavy heads; the valves
are .30-.40 mm. long, besides the loop, which adds .03-. 10 mm. more, and are
not constricted above the base; their general form is similar to those of Echinus
lylodes (PL 93, fig. 12}. The tridentate pedicellariae vary greatly in size but are
otherwise all alike; the valves, which range from .40 to 1.00 mm. in length, are
rather narrow and are strongly compressed at the base of the blade. The
triphyllous are not peculiar; the valves measure about .20 mm. in length by .15
in breadth of blade. The sphceridia are not peculiar. The calcareous spicules
of the tube-feet although arcuate, are scarcely bihamate.

The ground color of the test is dull white, at least actinally, becoming grayish
or greenish abactinally, where it is blotched or marbled with dull olive-green or
purplish brown. The spines are dull greenish, or yellowish becoming lighter
at the tip; they are often dirty cream-color. Although very variable in detail,
the general impression of this sea-urchin as seen from above is dull olive-green,
mottled with cream-color, or cream-color mottled with dull shades.

This interesting species seems to be abundant off southern and Lower Cali-
fornia in water of moderate depth. Apparently it is not littoral for none of the
collectors who have gathered the shore forms there seems to have found a speci-
men. It cannot be confused with any other echinoid of that region, nor does it
approach very closely to any known member of the Echinidse. But the variable
character of the abactinal system, the tendency of the buccal membrane to
become less heavily plated in adults, and the appearance of the pedicellarise
all point to the "intermediate" position it occupies between euerces and semi-
tuberculatus.

It was taken by the "Albatross" at the following places:—

Station 2838. Off Cedros Island, Lower California; 28° 12' N., 115° 9' W.
Bott. temp.? 44 fathoms. Gn. m.

Station 2899. Off Santa Barbara, California; 34° N., 120° 23' W. Bott.
temp.? 44 fathoms. Gy. s., brk. sh.

Station 2906. Off Santa Barbara, California; 34° 23' 30" N., 120° 19' 30" W.
Bott. temp. 55.5°. 9G fathoms. S., m.

Station 2907. Off Santa Barbara, California; 34° 24' 30" N., 120° 20' W.
Bott. temp.? 44 fathoms. Fne. gy. s.

Station 2913. Off San Diego, California; 32° 25' 30" N., 119° 3' 30" W.
Bott. temp. 59°. 26 fathoms. Brk. sh.

LYTECHINUS ANAMESUS. 257

Station 2922. Off San Diego, California; 32° 27' 15" N., 119° 5' 15" W.
Bott. temp. 57.1°. 47 fathoms. Fne. gy. s.

Station 2930. Off San Diego, California; 32° 25' N., 1 17° 18' 45" W. Bott.
temp. 52.9°. 60 fathoms. M.

Station 2931. Off San Diego, California; 32° 25' 30" N., 117° 16' 45" W.
Bott. temp. 55.9°. 34 fathoms. Gy. s., sh.

Station 2932. Off San Diego, California; 32° 26' 15" N., 117° 16' 15" W.
Bott. temp. 58°. 20 fathoms. Gy.'s., brk. sh.

Station 2934. Off San Diego, California; 32° 33' 30" N., 117° 16' W. Bott.
temp. 58.2°. 36 fathoms. Gy. s.

Station 2938. Off San Pedro, California; 33° 35' 15" N., 118° 8' 30" W.
Bott. temp. 58°. 47 fathoms. Fne. gy. s., st.

Station 2939. Off San Pedro, California; 33° 36' N., 118° 9' 30" W. Bott.
temp.? 27 fathoms. Fne. gy. s., st.

Station 2942. Off San Pedro, California; 33° 38' 45" N., 118° 13' 45" W.
Bott. temp.? 20 fathoms. Gy. s., brk. sh.

Station 2943. Off Santa Barbara, California; 34° 0' 30" N., 119° 28' 30" W.
Bott. temp. 56°. 31 fathoms. Rky.

Station 2944. Off Santa Barbara, California; 34° N., 119° 28' 30" W. Bott.
temp.? 30 fathoms. Rky.

Station 2945. Off Santa Barbara, California; 34° N., 119° 29' 30" W.
Bott. temp.? 30 fathoms. P.

Station 2951. Off Santa Barbara, California; 33° 55' 30" N., 119° 55' W.
Bott. temp.? 48 fathoms. Fne. gy. s.

Station 2965. Off Santa Barbara, California; 34° 21' 20" N., 119° 38' 30"
W. Bott. temp. 58°. 27 fathoms. Fne. gy. s., r.

Station 2966. Off Santa Barbara, California; 34° 20' 40" N., 119° 38' 50" W.
Bott. temp. 58.5°. 30 fathoms. Crs. m.

Station 2967. Off Santa Barbara, California; 34° 21' 15" N., 119° 39' 10" W.
Bott. temp. 58°. 30 fathoms. Crs. m.

Station 2969. Off Santa Barbara, California; 34° 20' 40" N., 119° 37' 45" W.
Bott. temp. 58°. 26 fathoms. Gy. s., p., st.

Station 2975. Off Santa Barbara, California; 34° 1' 30" N., 119° 29' W.
Bott. temp. 57°. 36 fathoms. G., brk. sh.

Station 2978. Off Santa Barbara, California; 33° 59' 45" N., 119° 22' 15" W.
Bott. temp. 56.5°. 46 fathoms. Gy. s.

Station 2983. Off Guadeloupe Island, Mexico; 28° 58' 30" N., 118° 15' 45"
W. Bott. temp. 55.8°. 58 fathoms. Gy. s., brk. sh.

258 HAWAIIAN AND OTHER PACIFIC ECHINI.

Station 2984. Off Guadeloupe Island, Mexico; 28° 57' 15" N., 118° 15' 45"
W. Bott. temp. 49.8°. 113 fathoms. Gy. s., brk. sh.

Bathymetrical range, 20-113 fathoms. Extremes of temperature, 59°-49.8 °.
Four hundred and fourteen specimens.

Lytechinus semituberculatus Verr.

Echinus (Psammechinus) semituberculatus Agassiz and Desor, 1846. Ann. Sci. Nat., (3), VI, p. 368.
Lytechinus semituberculatus Yen-ill, 1S67. Trans. Conn. Acad., I, p. 301.

Now that it has become clear that Psammechinus pictus Verr. is quite distinct
from this species, it would seem that semituberculatus is not found on the con-
tinental coast, but is confined to the Galapagos Islands. All of the specimens
taken by the "Hassler" and "Albatross" are from the Galapagos, nor did the
collectors on those vessels meet with it elsewhere. The extensive collections of
Echini from Lower California and Mexico, studied by Verrill and Liitken con-
tained no specimens, and there are none in the M. C. Z. collections from the
mainland coast. The bright coloration is very distinctive and seems to be quite
constant. It is much nearer that of some West Indian specimens of variegatus
than it is to the dull shades of pictus.

The "Albatross" took this species at the following places: —

Hood Island, Galapagos.

Indefatigable Island, Galapagos.

Station 2810. Off Hood Island, Galapagos; 1° 22' S., 89° 39' 30" W. 6.5
fathoms. Co. s.

Fourteen specimens.

Lytechinus pictus, comb. nov.

Psammechinus pictus Verrill, 1867. Trans. Conn. Acad., I, p. 301.
Plates 99, figs. 6, 7; 107, figs. 12-14.

The series of specimens collected by the "Albatross," taken in connection
with those which have gradually accumulated in the M. C. Z. collection, enables
us to give this species its rightful place. Verrill's original description is suffi-
ciently complete so that it is unnecessary to repeat its details, but since its main
points have been obscured by his considering a large specimen of Toxopneustes
roseus as an adult pictus (1. c., p. 581), it seems desirable to give figures of this
interesting species and to discuss some of its characteristic features.

LYTECHINUS PICTUS. 259

The series at hand ranges from 5 to 40 mm. in diameter. In the full-grown
specimens, the height is clearly more than half the diameter. The abactinal
system (PI. 99, fig. 7) is rather small, only about .25 h. d. The actinostome is
also small, about .35 h. d. Oculars I and V are well insert, but in most of the
small specimens, ocular V is more or less fully excluded. The adult character
in this respect seems to be acquired when the individuals are between 15 and
20 mm. in diameter. The buccal membrane (PI. 99, fig. 6) is heavily plated in
all the specimens, except the largest, where more or less resorption has occurred
and the membrane is visible between the plates. Coronal plates are numerous;
in a specimen 20 mm. h. d., there are 14 interambulacral and 17 ambulacral
plates in each column, and in the specimen, 40 mm. h. d., there are 21 and 30
respectively. The gill-cuts, though well defined are not deep, even in the largest
specimens. The primary spines are short, the longest ones only about one
fifth the diameter of the test. Pedicellarise are only fairly common. The
globiferous were found in both young and old ; the valves range from .50 to .60 mm.
in length and have the usual tubular blade and long end tooth; the blade may
be constricted, just above the broadly expanded base. The ophicephalous are
fairly common; the valves of those found in the large specimens, are about
.60 mm. long, nearly triangular, and not at all constricted, while in the young
specimens, the valves are only about .35 long, they are distinctly constricted
and the lime of which they are composed is strongly tinged with brown. The
tridentate were found only in the large specimens. They closely resemble those
of Toxopneustes (Plate 93, fig. 7); the valves measure .30-1.00 mm. in length.
The triphyllous are fairly common, but very small; the valves measure about
.15 mm. in length by .13 mm. across the blade. The sphceridia are very numer-
ous, for although they do not extend far up the ambulacra from the peristome,
as many as twenty may be found at the base of a single ambulacrum; of course,
in the young they are far less abundant. They are not peculiar in either size
or shape. The spicules in the tube-feet are arcuate but not bihamate, the ends
being blunt or almost knobbed.

This species has hitherto been regarded as the young of some larger species.
The specimens received at the Museum of Comparative Zoology have usually
been labelled Toxopneustes pileolus or Tripneustes depressa. The structure of
the ambulacra shows conclusively that these identifications are wrong. Although
nearly related to L. semituberculatus, there is not the least superficial resemblance
to that species. On the other hand, except for differences in color, the resem-
blance to L. verruculatus is noteworthy.

260 HAWAIIAN AND OTHER PACIFIC ECHINI.

The M. C. Z. collection contains specimens of pictus from San Diego, Gala.,
and La Paz, and Cape St. Lucas, Lower California.

The "Albatross" collected it at the following stations:—

Station 2824. Off Espiritu Santo Island, Lower California; 24° 22' 30" N.,
110° 19' 30" W. Bott. temp.? 8 fathoms. Brk. sh.

Station 2825. Off Espiritu Santo Island, Lower California; 24° 22' 15" N.,
110° 19' 15" W. Bott. temp.? 7 fathoms. Brk. co.

Station 2827. Off Espiritu Santo Island, Lower California; 24° 11' 45" N.,
109° 55' W. Bott, temp.? 10 fathoms. Sh.

Station 2828. Off Espiritu Santo Island, Lower California; 24° 11' 30" N.,
109° 55' W. Bott, temp.? 10 fathoms. Sh.

Station 2829. Off Cape St. Lucas, Lower California; 22° 52' N., 109° 55' W.
Bott. temp. 74.1°. 31 fathoms. Rky.

Station 3002. Off San Jose" Island, Lower California; 25° 2' 15" N., 110°
43' 30" W. Bott. temp.? 17 fathoms. S., sh.

Station 3005. Off San Jose" Island, Lower California; 25° 2' 45" N., 110°
43' 30" W. Bott. temp.? 21 fathoms. S., sh., corln.

Station 3006. Off San Jose" Island, Lower California; 25° 2' 30" N., 110°
43' 30" W. Bott. temp.? 8 fathoms. Sh., s.

Bathymetrical range, 7-31 fathoms.

Thirty-one specimens.

ECHINUS.

Linne", 1758. Syst. Nat., ed. 10, p. 663.
Type-species, Echinus esculentus Linn6, 1. c.

This genus as here limited contains seventeen species of which three are now
described for the first time. Having already (p. 240) discussed the general inter-
relationships of this group, it is not necessary to repeat here the arrangement there
outlined. Whether Alexandri is nearest the original stock or not, there can be
little doubt that the North Atlantic is the geographical centre of the genus.
We find in European waters, besides Alexandri, no less than five species (acutus,
elegans, esculentus, melo, tenuispinus) , while in the tropical Atlantic we find
four more (atlanticus, gracilis, tylodes, Wallisii). If it is correct to consider
affinis as only an extreme and as yet incompletely differentiated form of acutus
var. norvegicus, that species extends in deep water along the eastern coast of
the American continent at least as far as New Jersey. Of the other species,
one (lucidus) is from the coast of Japan, and is nearly related to Alexandri and

ECHINUS. 261

elegans, while another (armalus), the relationships of which are doubtful, was
taken by the "Siboga" in the East Indies. The remaining species are all from
the southern hemisphere. Two (Gilchristi from South Africa and anchistus, from
deep water off the coast of Chile) are very near to some forms of acutus, though
now so widely separated geographically. Other southern species are horridus
and euryporus from "Challenger" Station 308, off the coast of Chile, in 175
fathoms. Doderlein has recorded horridus from off South Africa also, and a
young specimen from off St. Paul Island still further to the east; the latter
individual is so small that its identification seems very doubtful. The most
characteristic of the southern species is margaritaceus , which appears to have
a wide distribution in the Antarctic Ocean.

The status of many of these species is open to question, due no doubt, in
part at least, to the lack of material. It is probable that knowledge of the color
in life would help in determining the true position of some forms. There seems
to be a general and justified feeling of doubt as to whether melo is specifically
different from acutus. It has seemed best however, to retain melo for the present
as few specimens are available and those are quite easily recognizable. The
similarities between A lexandri and elegans are striking and lucidus is very near
indeed to the latter. One has to search very carefully for any differences what-
ever between these three species and all found seem exceedingly trivial. More-
over the three are very near some forms of acutus var. norvegicus. Typical
acutus, when full grown, is so very different from norvegicus that it was hard to
accept Mortensen's claim that they are identical, but examination of a large
series of specimens shows that it is not practicable to draw the line between
the two forms. It is necessary however to go still further for affinis is not
distinguished from norvegicus by any constant characters; the one is simply the
most highly specialized form of the other. This form (affinis) is remarkable
for the frequency with which the uppermost ambulacral plates have only two
elements, as noted and figured by Jackson (1912, Mem. Boston Soc. Nat. Hist.,
VII, p. 118, text-fig. 115).

There is no question about the validity of esculentus and Mortensen's descrip-
tion of tenuispinus would indicate the desirability of accepting that species.
There is no doubt about the standing of gracilis and atlanticus, which are very
distinct and easily recognized species, and the same is true of tijlodes. I
consider Wallisii as somewhat doubtful as it may be simply the full-grown adult
of Alexandri; the smallest Wallisii is 60 mm. h. d. while the largest Alexandri is
only 50 mm. The test in Wallisii is so thick and rough, that it is hard to believe

262 HAWAIIAN AND OTHER PACIFIC ECHINI.

that Alexandri would reach such a condition, even with a great increase in size.
As de Meijere's description of armatus fails to make clear the character of either
the buccal membrane or the abactinal system, its true position is doubtful.
It appears to be a connecting link between Echinus and Heliocidaris, having
the ambulacra of the former, with the globiferous pedicellarise of the latter.
Although, as already stated, Gilchristi and anchistus are near some forms of acutus
(notably affinis), they seem to be valid species and the same may be true of
euryporus, but it is very difficult to separate the last from elegans in any satis-
factory way. More material is greatly needed for the elucidation of horridus;
the specimen from South Africa figured by Doderlein seems to be identical with
the "Challenger" specimens. One of the latter, the only specimen accessible,
is very remarkable for the form of the test; although when collected it was
broken into a number of pieces, some of which are missing, there is no question
that in life the vertical diameter greatly exceeded the horizontal, and so far as it
can be estimated, must have been nearly twice as much. Such a high test has
not hitherto been recorded among Echini either living or fossil.

As regards margaritaceus, I disagree with the eminent Continental zoologists
who make it the type of a genus Sterechinus, in which they recognize
four species (margaritaceus, Neumayeri, antarcticus, diademd). Although the
material studied is not extensive, it is very representative, consisting of speci-
mens from Patagonia, Kerguelen, Heard Island, Coulman Island, "Gauss"
winter station, and Antarctica (Mission Charcot), and including specimens
identified by Koshler as Neumayeri and by Mortensen as diadema, antarcticus,
and Neumayeri. Although these specimens reveal a certain amount of diversity,
it is not nearly so great as that shown by acutus or by our common northern
sea-urchin, Strong ylocentrotus drobachiensis. Moreover the differences are not
only slight but are very inconstant and it seems unwise to distinguish more than
a single species. Owing to the characteristic form and appearance, and particu-
larly the specialized abactinal system, it would have been advantageous to
recognize the genus Sterechinus (see below under Evechinus) but horridus is
such an obvious connecting link in certain particulars, though so different in
others, and the resemblance to esculentus in ambulacra and buccal plates is so
noteworthy that margaritaceus cannot be removed from Echinus.

The species of Echinus which seem valid may be distinguished as follows,
although in some cases the general appearance is of more value than the trivial
characters here used. Attention should be called to the fact that acutus occurs
in two places in the table owing to its variable ambulacra. Not having seen

ECHINUS. 263

armatus, the globiferous pedicellarise are used as the distinguishing mark of that
species.

Ambulacral plates, at least at and below the ambitus, each with a primary

tubercle.
Test rather low, often depressed, vertical diameter rarely exceeding .55

horizontal.

Ambulacral plates few (18-20 in specimens 45-50 mm. h. d., 24-28 in
specimens 65-100 mm.); valves of large tridentate pedicel-
lariae, broad and flattened.
Test not specially thick or rough; v. d. about equal to half h. d.;

abactinal system much smaller than actinostome Alexandri.

Test very stout and rough; v. d. .55-.65 h. d.; abactinal system

nearly equal or even exceeds actinostome Wallisii.

Ambulacral plates more numerous (25 or more in specimens over 40
mm. h. d.); valves of large tridentate pedieellariae, slender and
compressed.

Pore-pairs small; poriferous areas narrow (each about one sixth
of ambulacrum at ambitus) with no noticeable tendency to
widen at peristome; arcs of pores at ambitus especially two
outer pairs nearly vertical ; ambulacral tubercles usually dimin-
ishing unequally towards abactinal system, or else unequal in
two halves of same ambulacrum; abactinal system usually

large (.30 h. d. or more) with periproct about half as large . acutus var. norvegicus.
Pore-pairs large; poriferous areas moderate (each about one fifth
or more of ambulacrum at ambitus), tending to become
somewhat expanded at peristome; arcs of pores at ambitus,
distinctly oblique, becoming more or less nearly horizontal;
abactinal system usually about .25 h. d. with periproct not
usually half as large.

Actinal membrane with scattered plates distal to the buccal
circle, of which some (one at least) are rounded and
thickened, and carry pedicellariae; valves of globif-
erous pedicellariae with 2-3 lateral teeth on each side.
Primary tubercles of ambulacra, of irregular size, not
diminishing equally from ambitus upward and down-
ward; color of test dull purplish brown . . . euryporus.
Primary tubercles of ambulacra, regular, diminishing
equally from ambitus; color variable but usually

with some red clegans.

Actinal membrane with numerous small thin plates, none
of which carry pedicellariae; valves of globiferous pedicel-
lariae with only one lateral tooth on each side .... lucidus.
Test high, v. d. two thirds h. d. or more.

Test remarkably bare above ambitus, there being but few spines, be-
sides the long widely spaced primaries ailanticus.

Test well covered with spines above ambitus.

Test beautifully marked with deep green and white (or yellowish) ;

primary spines very short, white (or yellowish) gracilis.

264 HAWAIIAN AND OTHER PACIFIC ECHINI.

Test not marked with green and white; primary spines moder-
ately or quite long, red or reddish.
Test stout, rough, and almost sculptured, with tubercles

elevated and sutures and pcripodia depressed tylodes.

Test not very stout nor rough, very closely covered with

secondary and miliary tubercles horridus.

Ambulacral plates at and below ambitus often lack primary tubercles; usually

plates lacking tubercles alternate with those which have them.
Valves of globiferous pedicellarise with lateral teeth, at least one on each

side.
Buccal plates do not carry spines; ambulacral plates not twice as many

as interambulacrals.

Primary tubercles very small, wanting from many abactinal
interambulacral plates; diameter of areolsc at ambitus less than
half the height of interambulacral plates; pore-pairs close to
central primary tubercle of ambulacral plate, more or less
removed from interambulacrum and usually with tubercles
between them and marginal suture; primary spines, slender,

green melo.

Primary tubercles large, present on abactinal interambulacral
plates; areola; at ambitus more than half the height of
interambulacral plate; pore-pairs close to or near interam-
bulacrum, with no tubercles between them and marginal
suture; primary spines more or less stout, not uniformly
green.

Abactinal coronal plates with few secondary tubercles;
median abactinal interambulacral areas more or less
bare.

Test not exceedingly flattened, often high, v. d. = .45
h. d. or more, usually over .50; one or more peri-
proctal plates with at least one spine; test and pri-
mary spines usually with more or less red .... acutus.
Test very flat, v. d. = .40 h. d. or less; suranal plate more
or less circular and very conspicuous but neither it
nor any other periproctal plate carries a spine; no red

(in preserved specimens) anchistus.

Abactinal coronal plates with numerous secondary tubercles;

no bare spaces, in the abactinal interambulacra .... Gilchristi.
Buccal plates with at least a few spines among the numerous pedicel-
larise; secondary spines and tubercles numerous; ambulacral
plates nearly or quite twice as many as interambulacrals.
Test rather stout, high, v. d. usually more than .60 h. d. ; primary
spines short, with secondaries nearly as large; all oculars
broadly exsert.

Color orange reddish or purplish; abactinal spines more or
less numerous; coronal plates numerous (in specimens
50-60 mm. h. d., 22 or more interambulacral and 50 or

more ambulacral in each column) esculentus.

Color white or whitish; abactinal spines few and scattered;

ECHINUS WALLISII. 265

coronal plates less numerous (in specimens 50-60 mm.
h. d., about 18 interambulacral and 38-40 ambulacral in

each column) tenuispinus.

Test thin, rather fragile, low, v. d. less than .55 h. d.; primary
spines much longer than the slender secondaries; usually two
or more oculars reach the periproct; if all are excluded, one

is very slightly so margaritaceus.

Valves of globiferous pedicellaria; have a lateral tooth on only one side but

that one is conspicuous armatus.

Echinus Wallisii A. Ag.

Echinus Wallisii A. Agassiz, 1880. Bull. M. C. Z., VIII, p. 77.

Plates 93, fig. 32; 108, figs. 1, 2.

The type of this species (PI. 108, figs. 1-2) has never before been figured.
The statement in the original description that the pairs of pores are in sets of
two was based on the examination of the abactinal part of a single ambulacrum,
which possesses this peculiarity. Examination of other ambulacra and of
another specimen shows that this is not a characteristic feature at all, but is
only seen clearly in one ambulacrum. It is interesting to note (as stated above,
p. 261) that in the affinis form of Echinus acutus the abactinal ambulacral plates
very commonly consist of only two elements. As already indicated it is by no
means clear that Wallisii is not the adult form of Alexandri or possibly a local
variety. As bearing on the point, it may be noted that : — the two specimens
at hand are respectively 60 and 100 mm. in diameter, 35 and 55 mm. high. The
abactinal systems are 15 and 29 mm. across, while the actinostomes are only
15.5 mm. and 26 mm. Although the primary spines are all more or less broken,
it is evident that in the type they measured between forty and fifty millimeters.
They are slender and smooth and have about 24 longitudinal striations. The
secondaries are short, rarely exceeding 12 mm. and are very slender. They are
not abundant, but are scattered over the test, with a fair degree of uniformity.
The test itself is very stout and rough, especially in the smaller specimen. The
roughness is due to the elevation of the tubercles and the depression of the
sutures. At and below the ambitus, each ambulacral plate carries a normal
primary tubercle, but above the ambitus, the tubercles are of irregular size and
occurrence. They are not so large as those of the interambulacra but they
cover a large part of the plate. The pores are large and the poriferous areas
quite broad. In the smaller specimen there are 26 ambulacral, and 16 inter-
ambulacral plates while in the larger specimen the figures are 28 and 20 respec-

266 HAWAIIAN AND OTHER PACIFIC ECHINI.

tively. The ocular plates are all broadly exsert in both specimens and they,
as well as the genitals, are somewhat swollen. The periproct is large, covered
by numerous small plates, among which a suranal can easily be distinguished.
The actinostomal membrane is thick, and carries few plates outside the buccal
circle. The buccal plates are thickly covered with pedicellariae but carry no
spines. The gill-cuts are shallow and poorly defined. The pedicellarise, sphseri-
dia, and spicules are essentially the same as in Alexandri. The globiferous
pedicellarise (PI. 93, fig. 32) have heads about .70 mm. long and are borne on
long, flexible stalks, 10-12 times as long as the head. The valves of the large
tridentate pedicellarise measure up to 1.75 mm. in length, and the blade contains
more mesh-work than in Alexandri. The spicules are bihamate.

Echinus euryporus,1 sp. nov.
Plates 93, figs. 2, 3; 109, figs. 4-6.

The single specimen upon which this species is based is 46 mm. in diameter
and 26 mm. high. The test is well arched and the actinostome is not sunken.
There are 15 interambulacral plates in each column, and 26 ambulacrals. The
abactinal system is 12 mm. across, and the periproct is 15 mm. The ocular
plates are all broadly exsert. The periproct is covered by small but somewhat
swollen plates, among which the suranal can be easily distinguished. Each
interambulacral plate at ambitus carries, besides its primary tubercle, 8-12
secondaries and 50-60 miliaries. All the tubercles, even the miliaries, are
sufficiently elevated to be quite distinct and the test therefore appears rather
rough. In the ambulacra, the primary tubercles do not diminish uniformly
from the ambitus to the actinostome and abactinal system, but are of rather
unequal and irregular sizes. The poriferous area is moderately broad at the
ambitus, with the pores large and the arcs quite oblique. At the peristome,
the poriferous areas become decidedly widened, so that each one is wider than
the interporiferous area, and the arcs approach the horizontal. This feature,
upon which the proposed name is based, serves to distinguish euryporus from
any forms of acutus with which it might otherwise be confused, but it does not
assist much in separating it from elegans which resembles it in this particular.
The genital and ocular plates, except the distal tips of the genitals, are well
covered with small tubercles. The buccal membrane carries besides the small

1 e'vpvs = broad + irdpos = a way.

ECHINUS LUCIDUS. 267

primordial ambulacral plates, a few rounded scattered plates, distal to them.
All the plates carry more or less numerous pedicellarise, but no spines. The
primary spines are 20-25 mm. in length, rather slender and tapering. The
secondaries are much shorter and even more slender relatively. The pedicella-
rise are abundant, but not peculiar. The valves of the globiferous (PI. 93, figs.
2, 3) measure about .80 mm. long; the base is about .30 mm. long by .23 wide.
There are two or three teeth on each side of the blade near the tip. The small
tridentate pedicellarise have broad, curved, truncate valves, but the large ones
have narrow straight, pointed valves, which may be over a millimeter long.
The valves of the ophicephalous pedicellarise are very broad and are not con-
stricted; they are .3S-.45 mm. long besides the loop. The triphyllous valves
are about .16 by .14 mm. The spicules are bihamate and the sphaeridia are not
peculiar.

The color of the test is light purplish brown, lightest on the poriferous areas
and along the median interambulacral suture. The spines are light reddish
brown. Actinally both test and spines are much lighter, practically white.

The specimen upon which this species is based was taken by the "Challenger"
(St. 308) off the coast of southern Chile, in 175 fms. It is listed, without com-
ment, in the Report on the "Challenger" Echini, as Echinus norvegicus. Al-
though very close to that species, it is even nearer to elegans, so that its rightful
position is doubtful. The color is so unlike any elegans, that it looks very
different, and this fact, in connection with the geographical isolation of the
specimen favors its specific separation.

Echinus lucidus D6d.

Echinus lucidus Doderlein, 1885. Arch. f. Naturg., Jahrg. LI, 1, p. 97.

Plate 107, figs. 1-3.

The series of specimens, ranging in diameter from 12 to 40 mm., has made
very clear the close relationship which this species bears to elegans and it is
doubtful if the two characters by which they are separated in the table above
(p. 263) are likely to prove either important or constant. I am not ready
however to affirm the identity of the two and expect that in life there may be
important color differences. Among the preserved specimens however, there
are dingy yellowish and whitish individuals of both species. The specimens of
lucidus show great diversity in the length and stoutness of the primary spines,

268 HAWAIIAN AND OTHER PACIFIC ECHINI.

and in the elevation of the test. Thus the largest specimen (40 mm. h. d.)
(PI. 107, fig. 3} has primary spines less than 7 mm. long, and the vertical diameter
of the test is only 17 mm., while another specimen from the same station, 30 mm.
h. d., is 15.5 mm. high and has primary spines 17 mm. long. A specimen (PI. 107,
fig. 1) from an adjoining station is 37 mm. in diameter and almost 21 mm. high.

The "Albatross" took lucidus at the following stations: —

Station 4917. Off Kagoshima Gulf, Japan; 30° 24' N., 129° 6' E. Bott.
temp. 42.7°? 361 fathoms? Gy. s., glob., brk. sh.?

Station 4957. Between Kagoshima and Kobe, Japan; 32° 36' N., 132° 23' E.
Bott. temp. 39.8°. 437 fathoms. Gn.-br. m., fne. gy. s., for.

Station 4958. Between Kagoshima and Kobe, Japan; 32° 36' 20" N.,
132° 24' 30" E. Bott. temp. 40.1°. 405 fathoms. Gn.-br. m., fne. gy. s., for.

Station 4959. Between Kagoshima and Kobe, Japan; 32° 36' 30" N.,
132° 23' 20" E. Bott. temp.? 405-578 fathoms. Gn.-br. m., fne. gy. s., for.

Station 4965. Between Kobe and Yokohama, Japan; 33° 35' 20" N., 135°
10' 50" E. Bott. temp. 49.4°. 191 fathoms. Dk. gn.-gy. s., sh.

Station 4980. Between Kobe and Yokohama, Japan; 34° 9' N., 137° 55' E.
Bott. temp. 39°. 507 fathoms. Br. m., fne. s., for.

Station 5048. Between Hakodate and Yokohama, Japan; 38° 9' 24" N.,
141° 52' 30" E. Bott. temp. 40.7°. 129 fathoms. Dk. gy. s., brk. sh.

Station 5049. Between Hakodate and Yokohama, Japan; 38° 12' N.,
142° 2' E. Bott. temp. 37.8°. 182 fathoms. Dk. gy. s., brk. sh., for.

Station 5051. Between Hakodate and Yokohama, Japan; 38° 11' N.,
142° 12' E. Bott. temp. 38.1°. 399 fathoms. Dk. gy. s., brk. sh., for.

Station 5078. Off Omai Saki Light, Japan; 34° 12' 20" N., 138° 2' 30" E.
Bott. temp. 38.9° 475-514 fathoms. Fne. gy. s., glob.

Station 5079. Off Omai Saki Light, Japan; 34° 15' N., 138° E. Bott. temp.
39.1°. 475-505 fathoms. P.

Station 5082. Off Omai Saki Light, Japan; 34° 5' N., 137° 59' E. Bott.
temp. 37.7°. 662 fathoms. Gn. m., fne. s., glob.

Station 5083. Off Omai Saki Light, Japan; 34° 4' 20" N., 137° 57' 30" E.
Bott. temp. 38.1°. 624 fathoms. Fne. gy. s., glob.

Station 5084. Off Omai Saki Light, Japan; 34° N., 137° 49' 40" E. Bott.
temp. 36.8°. 918 fathoms. Gn. m., fne. s., glob.

Station 5088. In Sagami Bay, Japan; 35° 11' 25" N., 139° 28' 20" E. Bott.
temp. 41.8°. 369-405 fathoms. Gn. m.

Bathymetrical range, 129-918 fathoms. Extremes of temperature, 49.4°-36.8°.

Fifty-six specimens.

ECHINUS TYLODES. 269

Echinus atlanticus Mrtsn.

Echinus atlanticus Mortensen, 1903. "Ingolf " Eeh., pt. 1, p. 101.

Plate 110, figs. 1-3.

As no figures of this fine species have ever been published, it seems worth
while to give illustrations showing the general appearance. The specimen
figured is from the type locality, "Challenger" St. 343, off Ascension Island,
Atlantic Ocean, 425 fathoms.

Echinus tylodes,1 sp. nov.
Plates 93, figs. 11-15; 109, figs. 1-8.

Although there are but two specimens, and one of those quite young, I feel
little hesitation in naming this new species. The type (PI. 109, figs. 1-3) is
57 mm. in diameter and 39 mm. high, with an abactinal system 13 mm. across
and an actinostome 19 mm. in diameter. There are 17 interambulacral, and 23
ambulacral plates in each column. The longest primary spines are only 18 mm.
long. The test is well arched, thick, and rough. The roughness is due to all
the tubercles being more or less elevated, without sunken areolae, while the
sutures and peripodia are more or less depressed. Each coronal plate bears a
well-developed primary tubercle, and these diminish regularly from the ambitus
upward and downward. There is a faint indication of a vertical ridge connecting
these tubercles; this is much more evident in the ambulacra than in the inter-
ambulacra and is most noticeable abactinally. The secondary spines and
tubercles are not very numerous, but on cleaned plates the tubercles are very
conspicuous. The plates of the abactinal system are distinctly swollen but
carry very few tubercles. The oculars are all broadly exsert. The periproct is
nearly 6 mm. across and is covered by numerous small rounded plates, among
which the suranal can scarcely be distinguished. The ambulacra are 13 mm.
wide at the ambitus and the interambulacra, 22 mm. The pore-pairs are large,
and the poriferous areas rather wide, becoming somewhat expanded at the
peristome. The arcs of pores are quite oblique, and approach the horizontal
at the peristome. The gill-cuts are broad and shallow, almost wanting. The
buccal membrane is thick and carries only a few scattered plates distally.

The smaller specimen is only 19 mm. in diameter and 10 mm. high, and is

1 TV\OS = a swelling or lump + eioos = form.

270 HAWAIIAN AND OTHER PACIFIC ECHINI.

thus relatively more flattened than the adult. The abactinal system and
actinostome are 6 and 9 mm. in diameter respectively and are thus relatively
much larger than in the adult. The primary spines are also longer than in the
adult, equalling about one half the diameter of the test. The latter is stout and
rough as in the large specimen, and the abactinal system is essentially the same,
though, as might be expected, the suranal plate is more obvious. The buccal
membrane is thinner and contains numerous, minute plates.

Pedicellarise are abundant. The globiferous are found mostly on the abacti-
nal surface. The valves (Plate 93, fig. 13) are .55-.70 mm. in length; the blade
and base are about equally long, and the latter is .20-.30 mm. wide. Besides
the conspicuous terminal tooth (PI. 93, fig. 14) there are two or three lateral
teeth on each side near the tip. The ophicephalous pedicellariae occur every-
where. The valves (PI. 93, fig. 12) are short, wide, and stout, not at all con-
stricted; they measure .20-.32 mm. long besides the loop, which adds about a
quarter more; the width is .75-. 85 of the length. The tridentate are less common
and are chiefly actinal in position; the valves (Plate 93, fig. 15) are short, only
.30-.50 mm. long, and are .10-. 13 mm. wide near the tip of the blade. The
triphyllous are very abundant everywhere, and the valves (PI. 93, fig. 11) are
as wide as long .10-. 12 mm. The sphseridia show no particular features and
the spicules, though excessively scarce in these specimens, appear to be normally
bihamate.

The two specimens agree fairly well in color. The test is light reddish buff,
so light in the small individual as to be almost white. The spines are pale red,
brighter in the young one, where the shade is quite orange; they are lighter,
almost whitish, distally but in the adult each one has a dusky brownish tip;
this however is not conspicuous and may be artificial.

The larger of these two specimens was taken by the "Blake" (Station 319)
in 262 fathoms, off South Carolina, was labelled and catalogued, without careful
examination, " Toxopneustes variegatus," and is listed thus in the final report
on the "Blake" Echini. The smaller specimen was taken by Count Pourtales
in 1868, off Sombrero Key, Florida, in 195 fathoms and is labelled "Echinus
Fkmingii. Echinus norvegicust " There appears to be no reason to doubt the
specific identity of the two specimens, or the distinctness of the species. Small
specimens might be mistaken for Wallisii, save for the bright coloration, but in
the adult, the height of the test is very characteristic.

ECHINUS ANCHISTUS. 271

Echinus anchistus,1 sp. nov.
Plates 93, figs. 8, 9; 103, figs. 4, 5; 111, figs. 1-3.

The greatly flattened test is the most striking feature of this southern species
and serves to distinguish it at a glance from its neighbor, euryporus. The
largest specimen at hand (PI. Ill, figs. 1-8) is 23 mm. in diameter but only
9.5 mm. high; the height is thus .41 v. d. In smaller specimens the height is
relatively greater and in the smallest specimen (7 mm. in diameter) it is a trifle
over .50 h. d. The abactinal system is about .30 h. d. and the periproct is more
than half of that. The actinostome is about equal to the abactinal system.
In the largest specimen there are 12 interambulacral and 18 ambulacral plates
in each column, while in the smallest the numbers are 8 and 10 respectively.
Although the type has 18 ambulacral plates, there are only 9-11 primary tubercles
n each half-area, and in the little specimen only 6 or 7 can be counted. The
pore-pairs are small, the arcs nearly vertical and the poriferous areas very nar-
row, particularly at the peristome, in striking contrast to the condition in eury-
porus. The coronal plates at ambitus are fairly well covered with secondary
and miliary tubercles but abactinally they become barer. They are not at all
rough and the areolse are not depressed nor are the bases of the tubercles notice-
ably elevated. The ocular plates (PI. 103, fig. 4) are all excluded from the
periproct by the wide genitals, and both sets of plates carry many small tubercles.
The periproct is covered by relatively few, rounded, flat plates, among which
the suranal is very prominent. None of these plates carry tubercles. In the
type the plates are so few and scattered that the periproctal membrane is plainly
visible among them, but this is not usually the case. The actinostomal mem-
brane (PL 103, fig. 5) contains numerous small and thin plates distal to the buccal
circle, but none of them bear any pedicellarise. The gills are very small and the
gill-cuts are almost wanting.

The primary spines are slender and fragile, when unbroken about half as
long as the test diameter, smooth and even polished, but showing clearly numer-
ous longitudinal striations. The smallest primaries are not so smooth and the
very slender secondaries, which are not numerous, are distinctly rough (under a
lens) particularly at the tip.

Pedicellaria? are common but not abundant. The globiferous are found
chiefly on the ambulacra. The valves (PL 93, figs. 8, 9) are very similar to

= nearest of kin.

272 HAWAIIAN AND OTHER PACIFIC ECHINI.

those of horridus, having a very square-cut base and 2 or 3 lateral teeth on each
side of the tip of the blade. They measure about .60 mm. long, with the end-
tooth some .10 mm., and the base .15 mm. wide. The ophicephalous have the
valves about .35 mm. long, besides the loop which may be .08-. 10 more; they
are distinctly constricted near the middle. The tridenlate are found chiefly on
the actinal surface. The valves are .80-1.40 mm. in length, narrow, com-
pressed and in contact only at the slightly expanded tip. The valves of the
triphyllous pedicellarise are more elongated than usual, measuring .13 by .095
mm. Nothing peculiar was noted in regard to the sphaBridia and no spicules
were found.

The color is dirty yellowish or pale buff, with the primary spines more or less
nearly pure white, and the periproctal plates and secondary spines whitish.

Although this species is very near some forms of acutus var. norvegicus, the
test is much flatter than in any specimens of that variety and the periproct and
globiferqus pedicellarise are different.

It was collected by the "Albatross" in February, 1888, at the two following
stations : —

Station 2788. Off Chile; 45° 35' S., 75° 55' W. Bott. temp. 36.9°. 1050
fathoms. Gn. m.

Station 2789. Off Chile; 42° 36' S., 75° 28' W. Bott. temp. 35.9°. 1342
fathoms. Bu. m.

Eighteen specimens.

PARECHINUS.

Mortensen, 1903. "Ingolf" Ech., pt. I, p. 108; 134.
Type-species, Cidaris angulosa Leske, 1778. Add. ad Klein, p. XVII; 28.

It is fortunate that Mortensen designated no type for his genus Parechinus
for had he selected either miliaris or microtuberculatus, the name would be simply
a synonym of Psammechinus. As it is those two species must be left in the
latter genus (see p. 242) and thus angulosus becomes ipso facto type of
Parechinus. Doderlein's name Protocentrotus thus becomes a synonym of
Parechinus, having the same type.

I have placed in this genus with angulosus, the closely related annulatus
and the more distantly connected magellanicus and Huttoni. While these four
species show a similar tendency towards a specialized abactinal system, it is
interesting to note that whereas in angulosus, the buccal membrane of which is
more or less fully provided with plates, 54 per cent of the specimens have all the

PARECHINUS. 273

oculars exsert (as in Psammechinus) and 30 per cent have ocular I insert, in
magellanicus, the buccal membrane of which is thin and naked, only 3 per cent
have all the oculars exsert and 88 have I insert. Few specimens of annulatus
have been examined and they resemble young angulosus in having the oculars
all exsert as a rule, with ocular I not rarely insert. Of Huttoni too few specimens
are known to make deductions of any value. Benham, in his description of the
species, says that the oculars are all exsert but in the specimen, which he kindly
sent the Museum, ocular I is broadly insert. Unfortunately he says nothing
as to the character of the buccal membrane, but as he first placed the species
in Mortensen's genus Pseudechinus, it is fair to infer that Huttoni agrees with
magellanicus in that particular; in the M. C. Z. specimen the buccal membrane
is wanting.

Although the species of Parechinus are all southern, their geographical
distribution is unknown. New Zealand and Australian material is infrequent
in most museums, and such specimens as have been accessible are lacking in
some important particulars; either the locality labels are open to suspicion, or
the tests are bare and usually lack the buccal membrane as well as all spines
and pedicellarke; they often lack the abactinal system also. With the exception
of bare tests not a single specimen of this family has been seen (save Eve-
chinus and Tripneustes) from any definite locality in either Australia or New
Zealand, and in every case of a bare test where the label reads "Australia" there
is reason for suspecting its correctness. I agree with Mortensen that angulosus is
not certainly known from any place except South Africa, but we have unquestion-
able specimens labelled "Red Sea," "Australia," "Hakodate, Japan," and
"Nicobar Islands." The last is of the same lot apparently as the one in the
Copenhagen Museum, as it was received by the M. C. Z. from the Vienna
Museum. The specimens recorded by Benham (1909, Rec. Canterbury Mus., I,
no. 2, p. 25) from New Zealand as angulosus seem to be magellanicus. This
opinion, which was originally based on the published description, has since been
confirmed by Dr. Benham's kind sending of one of the Stewart Island specimens.
There is little reason to doubt that Filhol's margaritaceus (auct. Perrier), said to
be common on Stewart Island and in Cook Strait is magellanicus. This species
is quite variable and Doderlein (1906) has suggested names for two varieties,
Hassleri from Eastern Patagonia and novos-amsterdamice from New Amsterdam
Island. The latter is a well-marked variety, characterized by the very small
abactinal system and the stout greenish spines. Having received a specimen
from Dr. Doderlein, and compared it with a large series of specimens from southern

274 HAWAIIAN AND OTHER PACIFIC ECHINI.

South America, Falkland Islands, and Marion Islands, its specific rank seems
probable. On the other hand, the variety Hassleri is not at all constant, and as
specimens from Marion Island resemble it very closely, it is clear it has no fixed
geographical limits. There are more than a dozen bare tests of what seem to be
magellanicus from Australia and New Zealand in the M. C. Z. Four of these
are labelled albocinctus and were received from Captain Hutton himself. After
a careful study of these tests and Benham's description (1908, Ann. Mag. Nat.
Hist., (8), I, p. 107), we have failed to find any character by which albocinctus
can be distinguished from magellanicus, for even the globiferous pedicellarise,
upon which Mortensen bases his placing of the two species in separate families,
are quite unreliable. It may well be granted that most New Zealand speci-
mens have the valves of the globiferous pedicellariae with a lateral tooth only
on one side, although no personal observations support such a view; that such
is the case uniformly seems very doubtful. In the light of all the evidence
available, it is probable that magellanicus is a circumpolar, subantarctic sea-
urchin, wanting on the South African coast, but without doubt occurring on
most of the other continental coasts, and on the islands, between 30° and 60° S.
lat. On the isolated island of New Amsterdam, it has become differentiated
into a well-marked variety, and possibly a New Zealand variety (albocinctus)
may be entitled to recognition. In New Zealand, moreover, a species has arisen,
Huttoni, which though nearly allied to magellanicus, seems to be well distinguished
by the much finer tuberculation of the test and some associated characters.
It is to be hoped that some New Zealand zoologist will soon determine, from the
study of good series of fresh material, the true relationships of albocinctus,
Huttoni, and magellanicus.

The four valid species of Parechinus may be distinguished as follows: —

Buccal membrane with more or less numerous plates (besides the primordial ambula-
crals) some of which are often thickened and bear pedicellariae.

Primary spines not banded angulosus.

Primary spines with 2-3 brown rings annulatus.

Buccal membrane thin and except for primordial ambulacrals, perfectly bare, or rarely
with a few minute plates.

Primary tubercles large and distinct, easily distinguished from secondaries, not
pink or pinkish orange; coloration variable magellanicus.

Primary tubercles very small, not occupying half the height of interambulacral
plates at ambitus, forming with the similar secondaries a horizontal row of 8-10
tubercles on each interambulacral plate at ambitus; outer half of this row
often double; test white or light yellowish brown; tubercles pink or pinkish
orange; primary spines pink or red at base, distally white Huttoni.

PARECHINUS MAGELLANICUS. 275

Parechinus magellanicus, comb. nov.

Echinus magellanicus Philippi, 1857. Arch. f. Naturg., Jahrg. XXIII, 1, p. 130.

In addition to the more than two hundred and forty specimens in the M. C. Z.
collection, the large series taken by the "Albatross" when on her voyage from
New York to San Francisco, in the Straits of Magellan in the winter of 1887-88 has
been available. There is a range in size from 2 to upwards of 40 mm. h. d., the
largest specimens being from the Falkland Islands. There is great diversity in
the length of the primary spines and in color. It is rather striking that although
the "Albatross" covered essentially the same ground as the "Hassler," no speci-
mens of magellanicus were taken having the short, blunt reddish spines so charac-
teristic of a large proportion of the "Hassler" collection. All of the "Albatross"
specimens have whitish or light colored primary spines. The color of the test
on the other hand is very variable. As a rule the median ambulacral and inter-
ambulacral areas are grayish or dull violet but they may be bright violet or
reddish, and in one specimen they are bright red. The poriferous areas are
usually rather light and often have a greenish tinge and in some specimens they
are, abactinally at least, distinctly yellow-green.

The "Albatross" specimens were taken at the following places:—

Mayne Harbor, Patagonia, Chile.

Port Otway, Patagonia, Chile.

Station 2768. Off San Jose Peninsula Patagonia, Argentina; 42° 24' S.,
61° 38' 30" W. Bott. temp.? 43 fathoms. Dk. s., bk. sp.

Station 2769. Off Cape Two Bays, Patagonia, Argentina; 45° 22' S., 64° 20'
W. Bott. temp.? 51.5 fathoms. Gn. m., fne. s.

Station 2776. Straits of Magellan ; 52° 41' S., 69° 55' 30" W. Bott. temp.?
21 fathoms. S., g.

Station 2777. Straits of Magellan; 52° 38' S., 70° 10' 30" W. Bott. temp.?
19.75 fathoms. G.

Station 2778. Straits of Magellan; 53° 1' S., 70° 42' 15" W. Bott. temp.
47.9°. 61 fathoms. Gy. s., bk. sp.

Station 2779. Straits of Magellan; 53° 6' S., 70° 40' 30" W. Bott. temp.
46.9°. 77.5 fathoms. Gn. oz.

Station 2780. Near Port Otway, Patagonia, Chile; 53° 1' S., 73° 42' 30" W.
Bott. temp. 46.9°. 369 fathoms. Gn. m.

Station 2785. Gulf of Penas, Chile; 48° 9' S., 74° 36' W. Bott. temp. 46.9°.
449 fathoms. Bu. m.

276 HAWAIIAN AND OTHER PACIFIC ECHINI.

Station 2787. Off Taytao Peninsula, Chile; 46° 47' 30" S., 75° 15' W.
Bott. temp. 53.9°. 61 fathoms. Gn. m.

Bathymetrical range, 19.75—449 fathoms. Extremes of temperature, 53.9°-
46.9°.

One hundred and thirty-one specimens.

NUDECHINUS, gen. nov.

Type-species, Nudechinus scotiopremnus, sp. nov.

The group of species to which the name of Nudechinus is given is probably the
least known of any similar group of Echinidse. They are small species having in
common the two characters of a thin, bare bucoal membrane and Lytechinus-like
globiferous pedicellariaj. That they form the connecting link between Lytechi-
nus and Gymnechinus seems unquestionable, but I cannot follow Mortensen
in putting them in the latter genus. The highly specialized abactinal system
of Gymnechinus is of far greater value for limiting a natural genus than any
character shown by buccal membrane or pedicellariae. Not having seen speci-
mens of darnleyensis Woods, inconspicuous Mortensen, or Gravieri Koshler, I
have selected for the type of the genus the larger of the two new species herein
described, though its status is not any more satisfactorily settled than that of the
other species. We have placed multicolor Yoshiwara in this genus after examin-
ing the type-specimen, which was most courteously loaned by Professor Goto of
the Imperial University, Tokyo. The specimen (PI. Ill, figs. 7-5) is 14 mm. in
diameter and 8 mm. high. The actinostome is large, 7 mm. across, while the
abactinal system is very small, 3 mm. in diameter. There are 14 or 15 inter-
ambulacral and 17 or 18 ambulacral plates in each column. The buccal mem-
brane is thin and except for the ten small, well-separated buccal plates, is per-
fectly bare. Oculars I and V reach the periproct. The globiferous pedicell arise
have valves about .35 mm. long, with a very conspicuous, straight terminal tooth,
.15 or .16 mm. in length. The ophicephalous pedicellarise are abundant and not
peculiar. No tridentate could be found, though prolonged search was made.
The coloration is slightly different from Yoshiwara's description, for we find no
brown markings on the test; all such markings seem to be deep green. The
spines are essentially as described by Yoshiwara, the violet bands being very
distinct though not sharply defined. Kcehler's species Gravieri is very
near multicolor but if the color of the spine-bands in the former is really "clear
rose," the general appearance must be very different from multicolor. It seems

NUDECHINUS SCOTIOPREMNUS. 277

that Gravieri may also be very near Lytechinus verruculatus and rufus, but if the
buccal membrane is really free from calcareous matter, the resemblance is not
significant. Unfortunately Kcehler does not state what the condition of the
buccal membrane in Gravieri is, as regards plates, and it is only because he places
the species in Gymnechinus that it is inferred the buccal membrane is thin and
naked. None of the species are known from more than one or two localities,
but all are from the Indo-Pacific region.

So far as our present knowledge goes, the six species of Nudechinus may be
distinguished from each other as follows : —

Primary spines not distinctly banded with violet or rose.

Test green or yellowish green or light with green blotches; spines white or whit-
ish, green or brown at the very base ... scolioprcmnus.

Test and spines not as above.

Secondary spines not swollen at tip; valves of tridentate pedicellarise wide,

not compressed at base of blade .... darnleijensis.

Secondary spines somewhat swollen at tip; valves of tridentate pedicel-

lariae strongly compressed in basal part of blade inconspicuus.

Primary spines distinctly banded with violet or rose.

Test whitish blotched with deep violet; primary spines white, each with a broad

band of rose-violet (K. & V. 597) . sticius.

Test and spines not as above.

Test gray greenish with some deep green spots; primaries grayish white

with 2 or 3 bands of clear rose Gravieri.

Test variegated with white and light and dark green; primaries with 2 or

3 violet bands near tip multicolor.

Nudechinus scotiopremnus,1 sp. nov.
Plate 97, figs. 4-6.

The type-specimen (PI. 97, fig. 4) is 17 mm. in diameter and 10 mm. high;
the actinostome is 7 mm. in diameter while the abactinal system is only 3.5 mm.
and the periproct 1.5 mm. There are 18 interambulacral plates in each column
and 22 ambulacrals. The primary spines are about 3.5 mm. long at the ambitus.
In the largest specimen, the diameter is 21 mm., the height 13 mm., the actino-
stome 10 mm., the abactinal system 4.5 mm., the periproct, 2 mm., and the
primary spines about 4 mm. ; there are 18 interambulacral, and 23 ambulacral
plates in each column. In the smallest specimen, the figures are, 13 mm. h. d.
9 mm. v. d.; 6 mm. actinostome, 3 mm. abactinal system; 1.33 mm. periproct;
2.5 mm., primary spines; 14 interambulacrals ; 17 ambulacrals.

= dark + irpep.vov = the lowest part of the trunk of a tree, the base.

278 HAWAIIAN AND OTHER PACIFIC ECHINI.

The test is well arched with a circular, or in the largest specimen somewhat

pentagonal ambitus. It is very completely covered with tubercles, but in the

type and in the large specimen, a sunken zigzag line is more or less marked in the

abactinal part of both the ambulacral and interambulacral areas. Seen from

above, the small specimen and the type show distinctly, twenty radiating series

of primary tubercles, two series in each area; these are of approximately equal

size and are decidedly larger than the numerous accompanying secondaries.

In the large specimen these twenty series are much less conspicuous as some of

the secondaries particularly at and near the ambitus are almost as large as the

primaries. Each ambulacral plate at the ambitus carries a primary tubercle, a

large secondary near the inner end of the plate and three to five small secondaries.

Each interambulacral plate has a large secondary tubercle on each side of the

median primary and there are four to six small secondaries also. Scattered

miliary tubercles occur in both areas. The abactinal system (PI. 97, fig. 6) is

noticeably small and the periproct is covered by few plates. The madreporic

genital is decidedly swollen and larger than the others; in the largest specimen

it has a single tubercle on the proximal margin but in the other specimens it has

none. Each of the other genitals carries two to four large tubercles. The

oculars are rather small, each with a large tubercle. In the type they are all

exsert, though I is only slightly so, and V is more nearly insert than any of the

other three. The same condition is found in the small specimen, but in the

large one, I is broadly insert and V is only barely excluded. The poriferous

areas are broad, the rather large pores being arranged in quite oblique arcs of

three pairs ; at the peristome the areas are narrower and the arcs are more nearly

vertical. The actinostome (PI. 97, fig. 5) is large, twice the diameter of the

abactinal system. The buccal membrane is thin and perfectly bare, except

for the small primordial ambulacrals. The gill-cuts are very well marked and

are moderately deep and wide. The primary spines are short, rather stout,

blunt, and not very conspicuous.

The pedicellarise are fairly common but are not peculiar. The globiferous
have valves about .25 mm. long, the narrow tubular blade about as long as the
rather wide base, and terminating in a single long tooth. The ophicephalous
have valves of nearly the same length, with the loop adding about a third as
much more. The tridentate are chiefly actinal in position and have valves,
half a millimeter long, more or less; these valves are somewhat curved, com-
pressed at the base of the blade and somewhat expanded distally, where they are
in contact with each other. The triphyllous are very small, the valves measuring

NUDECHINUS STICTUS. 279

only .10 mm. in length, the width of the blade being a trifle less. No calcareous
spicules were found.

The color of the three specimens displays some diversity. The smallest
has a light gray test, with a few scattered greenish blotches abactinally; the
spines are white, dark brown at the very base. The type is greenish yellow
with the primary spines whitish, dull green at the very base. The large speci-
men is yellowish green with the primaries whitish, deep green at the very base.

The type and the small specimen are labelled " de Suez, Vaillant. Psammechi-
nus de Liitken." They were obtained in Paris in 1869 by Mr. Agassiz, but were
never identified by him. The large specimen was purchased in Hamburg in
1870 by Mr. Agassiz and bears the label "New Zealand." It is probable that
one, if not both, of these locality labels is erroneous and as the latter appears
to be the less reliable of the two, the Indian and East African coasts will
possibly prove to be the home of this species. It seems to be nearly related
to inconspicuus Mortensen but is easily distinguished by the genital plates,
the tuberculation and the coloration. The pedicellarise are remarkably similar
to those of inconspicuus.

Nudechinus stictus,1 sp. nov.
Plates 93, fig. 1; 97, figs. 1-3.

The type and only available specimen (PI. 97, fig. 1) of this new form is
undoubtedly very young, although the pores in the genital plates are perfectly
distinct. It measures 6 mm. in diameter and rather more than 3 mm. in height;
the actinostome is more than 3 mm. across and the abactinal system, more than
1 mm. The primary spines are about 1.3 mm. long. There are 9 or 10 inter-
ambulacral, and 10 or 11 ambulacral plates in each column. Aside from the
primary tubercles, only a few small scattered secondaries or miliaries have been
observed; on the interambulacral plates at the ambitus an imperfect circle
around the primary tubercle is formed by these secondaries. The genital plates
(PL 97, fig. 3) are nearly of a size and each carries a single tubercle or none; the
pore is in the distal angle. The oculars are relatively large, all exsert, though I
is nearly insert; each with a single tubercle but apparently with no pore. The
periproct is rather larger than a genital plate and is covered by four plates of
which the suranal is largest. The poriferous areas are narrow, with the pores

s = spotted, dappled.

280 HAWAIIAN AND OTHER PACIFIC ECHINI.

small, in nearly vertical arcs of three pairs. The ambulacra are nearly as wide as
the interambulacra. The actinostome (PI. 97, fig. 2} is relatively large but the
gill-cuts are very shallow and inconspicuous. The buccal membrane is thin
and perfectly naked save for the small primordial ambulacral plates. The
primary spines are short, rather stout, and not very acute.

The pedicellariae are common but very small. The globiferous are remarkable
for the very short, hook-like blade of the valves (Plate 93, fig. 1) which are only
about .15 mm. long, with the base about .10 mm. wide. The ophicephalous
have somewhat larger valves, about .20 mm. in length, with an additional .03 to
.05 for the loop; they are slightly constricted at the base of the blade. No
tridentate nor triphyllous pedicellariae were found, nor were sphaeridia or calcareous
spicules seen.

The test is almost white with about a dozen irregularly scattered blotches
of dull purple at and above the ambitus. The spines are white but the primaries
are encircled at the middle by a broad, poorly defined band of rose-violet (K. &
V. 597).

The specimen upon which this new species is based was taken by the "Siboga"
in the Sulu Archipelago in 7-8 fathoms. (Station 109.) It was received by
the Museum of Comparative Zoology from the Amsterdam Museum, through
the kindness of Dr. Weber in 1907, bearing the label " Psammechinus verruculatus
L. Agass." In the report on the "Siboga" Echini, (p. 89), de Meijere records
nine specimens of "Psammechinus verruculatus Liitken" from Station 109 and
doubtless this specimen is one of them. But it does not seem possible to regard
it as verruculatus for in specimens of that species which are as small as, or smaller
than, stictus, the buccal membrane is fully plated. Moreover the coloration
is very different from any specimens of verruculatus, which I have, seen, and the
globiferous pedicellariae are different, resembling those of "Echinus darnleyensis1'
as figured by Mortensen ("Ingolf" Ech., pt. 1, PI. 21, fig. 36) very closely.
Indeed there can be little doubt that darnleyensis is the nearest ally of stictus,
the difference in coloration being the most obvious distinction. Larger series
of the two species may show that they are identical.

EVECHINUS. 281

EVECHINUS.

Verrill, 1871. Trans. Conn. Acad., I, p. 583.
Type-species, Echinuschloroliciis Valenciennes, 1846. Voyage "V6nus," Zoophytes, PI. VII, fig. 2.

The revival of the name Heliocidaris by Mortensen (1903, "Ingolf" Ech.,
pt. 1, p. 116) leads to some rather interesting results as to the proper use of that
name. Mortensen's remarks on the subject are as follows: —

"In 'Cat. rais.' the species variolaris Lamk., paucituberculatus Blainv., and
chloroticus Val. are enumerated under the genus Heliocidaris. — For the first of
these species the older name of Stomopneustes must be used; according to Agassiz
(Rev. of Ech.) paucituberculatus is synonymous with this. As far as I can see,
chloroticus must then be the type of the genus Heliocidaris; the name Evechinus
Verr. (1871) must then be dropped as being a much younger one, and I cannot
but wonder, why Agassiz, who otherwise takes great care to reestablish the
oldest names, has here preferred the name of Evechinus." So far as the data
given in this paragraph go there is no escape from Mortensen's conclusion but
why has he omitted to mention the five other species included in Heliocidaris
by Agassiz and Desor? Presumably it is because he considers the first named
valid species the type of the genus and hence does not consider it necessary to
even mention those which follow chloroticus. It is hardly necessary to say that
the selection of the type of a genus is not so simple a matter and in this case is
by no means easy. Agassiz and Desor included in their genus, besides the three
species mentioned by Mortensen, the following five: — Echinus margaritaceus
Val., E. erythrogrammus Val., E. ornalostoma Val., Heliocidaris mexicana Agass.,
and Echinus mirabilis Agass. The last named species was subsequently made
the type of Phymechinus by Desor, while Heliocidaris mexicana is now known
to be based on a specimen of Echinometra lucunter L. In 1863, A. Agassiz
established the genus Toxocidaris, naming as the first species Echinus Delalandi
Val., which was subsequently shown to be identical with E. erythrogrammus Val.
In 1871, therefore, when Verrill proposed Evechinus, Heliocidaris contained
chloroticus, margaritaceus, and omalostoma, and Verrill acted quite within his
right in making the first of these the type of a new genus, if he chose. The type
of Heliocidaris therefore must be either margaritaceus or omalostoma and fortu-
nately, whether one accepts for the former, Koehler's genus Sterechinus or rejects
it, as I do, as not distinguishable from Echinus, margaritaceus is eliminated,
and omalostoma is therefore the type of Heliocidaris. There is no question that
omalostoma of Valenciennes is identical with tuberculatus Lamarck and thus

HAWAIIAN AND OTHER PACIFIC ECHINI.

tuberculatus becomes the type-species of Heliocidaris. This leads to some
interesting but not very important nomenclatural changes which will be found
fully discussed under Heliocidaris (p. 350).

The genus Evechinus is characteristic of the New Zealand region and so far
as known there are no authentic records from elsewhere. There are specimens
in the M. C. Z. collection labelled "Fiji," but they are old and bare. They were
the gift of a sea-captain, and if obtained in Fiji, they were probably taken
there from New Zealand. The genus is as isolated structurally as it is geo-
graphically but, as already suggested, it is probably a highly specialized Echinus
and may find its nearest living relative in E. armatus (de Meij.). The genus
appears to be monotypic. In spite of Farquhar's very strong evidence to
the contrary, Mortensen thinks Bell's species rarituberculatus may be valid,
because the tridentate pedicellaria? are slightly different from those of chloroticus.
As I have found the form figured by Mortensen for rarituberculatus, mingled
with the ordinary form of chloroticus on the same specimen, there is no doubt
that Farquhar's view is correct; i. e. that rarituberculatus is the young of
chloroticus. As for Wood's Evechinus australice, a careful reading of his descrip-
tion, with specimens of Tripneustes in hand, shows that the bare tests upon
which his species is based, are simply young Tripneustes. The arrangement of
the pores, the shape of the auricles, and especially the fact that the ambulacra
are broader than the interambulacra constitute evidence scarcely to be doubted.

TOXOPNETJSTES.

L. Agassiz, 1841. Int. Mon. Scut., p. 7.
Type-species, Echinus pileolus Lamarck, 1816. Anim. s. Vert., Ill, p. 45.

In spite of the fact, referred to on p. 245, that Toxopneustes elegans Dod.
forms a connecting link with Lytechinus, this is a very natural and easily recog-
nized group. The broad poriferous areas, the short primaries, the abundant
tubercles, and the deep gill-slits combined with the absence, on many ambulacral
plates, of primary tubercles adjoining the poriferous area, give the members of the
genus an easily recognized appearance. Indeed the characters of the test are so
uniform, that the most obvious and perhaps the best specific characters are found
in the coloration. The difference in tuberculation between roseus and pileolus,
to which Mortensen refers ("Ingolf" Ech., pt. 1, p. 112) is not constant, and
some specimens of pileolus from Japan cannot be distinguished in this respect
from roseus. Since color seems to be the best criterion in this genus for determin-

TOXOPNEUSTES CHLORACANTHUS. 283

ing specific limits, Mortensen's restoration of roseus A. Ag. as a valid species
may be accepted and a form (chloracanthus) , of which several specimens of very
distinctive coloration from Samoa are in the M. C. Z. collection, is here described
as new. The most wide-ranging species is pileolus, which extends from Mauritius
to New Caledonia and Japan, and perhaps even to the Hawaiian Islands. The
little known and apparently rare species maculatus Lam'k. is from Bourbon and
Christmas Island, Pacific Ocean, which indicates as wide a range practically, as
that of pileolus. The American species roseus seems to be isolated on the west
coast of Mexico and Central America, while elegans is known only from Japan.
The characters distinguishing these species are as follows : —

Test not conspicuously marked with a large abactinal blotch and an ambital band of

bright violet.
Primary spines varied but without any subterminal blackish ring.

Tubercles more or less pink; spines red at base and for more or less of their

length ....................... pileolus.

Tubercles green or white; primary spines green at base, white at tip, usually

indistinctly banded with green and white; small spines white; no red chloracanthus.
Test, tubercles and spines more or less unicolor, rosy, purplish, or some shade

of brown ...................... roseus.

Primary spines with a conspicuous subterminal blackish band ....... elegans.

Test with a conspicuous abactinal blotch and an ambital band of bright violet (K. & V.

........................... maculatus.

Toxopneustes chloracanthus,1 sp. nov.
Plate 93, figs. 6, 7.

In size, proportions, and tuberculation this species is so much like pileolus
that it would be superfluous to give a detailed account of these features. There
are some slight differences in the pedicellarise which may be mentioned though
their constancy is doubtful. The globiferous are very abundant and have
extraordinarily long valves, 1.20-1.60 mm. (PI. 93, fig. 6) of which the blade is
about two thirds. They are mostly colorless, so far as the cleaned calcareous
substance is concerned, but many are more or less tinged with yellow-green,
while others have the base of the valve bright purple. The ophicephalous
pedicellaria? are also abundant but vary greatly in size. The valves measure
from .20 to 1 mm. in length; they are more or less triangular with the tip rounded
or truncate and the blade very much filled with calcareous matter. The tri-
dentate appear to be quite rare. They were found only on the actinal side of

green + anavOa. = spine.

284 HAWAIIAN AND OTHER PACIFIC ECHINI.

the largest specimen. The valves (PI. 93, fig. 7) are only about a millimeter
long and are strongly compressed. The triphyllous are rare and hard to find;
the valves are about .30 mm. long and .25 wide. The dumb-bell shaped spicules
in the globiferous pedicellarise are not peculiar.

The ground color of the test is light drab, approaching white. In the smaller
specimens this is very distinctly marked with broad, horizontal bands of green.
There is such a band at or near the ambitus and one or two above it; there may
be one on the actinal surface. These bands may be either continuous or broken
at the poriferous areas. The abactinal system, excepting the periproct, is also
more or less green. In the largest specimen all these green markings are either
wanting or only faintly indicated. All of the smaller spines are white but the
primaries are more or less green. At the ambitus, each primary has a single,
broad, poorly defined band of green, leaving both base and tip white. Actinally
this band becomes broader and more or less clearly divided into two or three
narrower bands, the base and tip of the spine remaining white. Abactinally
the green coloration becomes more or less completely diffused throughout the
whole spine, though it is often most marked at the base. There is no hint of
red anywhere.

The four specimens on which this species is based were received by the Mu-
seum of Comparative Zoology from the Godeffroy Museum in 1870, and are
labelled "Samoa." The largest is 106 mm. in diameter and 56 mm. high while
the smallest is only 40 mm. h. d. The latter and the next to the largest have the
green markings very deep and well defined. It is quite possible that this species
will prove to have the same relation to typical pileolus that the Florida form of
Lytechinus variegatus has to the typical form of that species, and if such proves
to be the case, then chloracanthus should rank only as a subspecies. In the
absence of connecting forms however it is preferable to give it full specific rank.
The occurrence of this green form of Toxopneustes at Samoa is interesting in
connection with the existence of a well-marked green variety of Mespilia among
the same islands (see p. 322). The latter however seems to occur with the usual
form, while so far as known typical pileolus does not occur at Samoa.

TRIPNEUSTES. 285

TRIPNEUSTES.

L. Agassiz, 1841. Int. Mon. Scut., p. 7.

Type-species, Echinus ventricosus Lamarck, 1816. Anim. s. Vert., Ill, p. 44, = Cidaris esculenta Leske,

1778. Add. ad Klein, p. XVII.

As there seems to be practical unanimity in accepting Tripneustes, and as
there is little room for doubt that Hipponoe of Gray is technically a nomen
nudum, while it is also preoccupied, Agassiz's name for this genus should be
accepted. There can be little doubt that it represents the extreme development
of the Echinidae, the specialized abactinal system, deep gill-slits, wide ambulacra,
and great reduction of ambulacral primary tubercles all pointing to the same
conclusion. There are nominally three species in the genus — -gratilla L. (long
known as variegatus) from the Indo-Pacific region, esculentus Leske from the
West Indies, and depressus A. Agassiz from the Western coast of Mexico. Re-
cent comparison of considerable numbers of specimens from various parts of
the world shows that the characters by which the species are supposed to be
distinguished are of slight importance and there is probably but a single species,
which is very variable in form, proportions, tuberculation, character of spines,
and color. It is true that all of the specimens with dark, slender spines are from
the Indo-Pacific region, but unfortunately the converse is not true for many
specimens from that region are as white and the spines are as coarse as in typical
esculentus. Owing to lack of sufficient material from the west coast of Mexico
the real status of depressus cannot be determined but it is doubtful whether it is
properly distinguishable from the West Indian species. The only character,
which has proved constant in the study of the material at hand, is the amount
of plating on the buccal membrane. With the aid of that feature, and the
comparison of tubercles and color, the three nominal species may be distin-
guished as follows: —

Buccal membrane with few, small scattered plates; tubercles rather large and numer-
ous, covering well the whole abactinal surface; spines white or yellowish .... esculentus.
Buccal membrane with numerous thick, moderately large plates; tubercles smaller

and more irregularly scattered.

Median interambulacral areas abactinally usually quite bare and dark colored;
tubercles usually very small; primary spines slender often very dark, but in

some specimens, white gratilla.

Median interambulacral areas abactinally not bare nor conspicuous; tubercles

moderate; spines fairly stout, white depressus.

286 HAWAIIAN AND OTHER PACIFIC ECHINI.

Tripneustes gratilla Loven.

Echinus gratilla LinnS, 1758. Syst. Nat., ed. 10, p. 664.

Tripneustes gratilla Loven, 1887. Bih. K. Sv. Vet.-Akad. Handl., XIII, afd, 4, no. 5, p. 77.

This well-known species was taken by the "Albatross" at the following
places, but none of the specimens call for any special comment, save the largest,
which is 145 mm. in diameter and pure white like esculentus.

Fakarava, Paumotu Islands.

Puako Bay, Hawaii, H. I.

Honolulu Reef, Oahu, H. I.

Honolulu Market, Oahu, H. I.

Clarion Island, Eastern Pacific Ocean.

Station 3876. Off Lahaina Light, Maui. Bott. temp. 74°. 28-43 fathoms.
S, g.

Twenty-seven specimens.

GYMNECHINUS.

Mortensen, 1903. "Ingolf" Ech., pt. 1, p. 115.

Type-species, Echinus RoUllardi de Loriol, 1883. Mem. Soc. Phys. et Hist. Nat. GeneVe, XXVIII,

no. 8, p. 23.

The remarkable abactinal system of Echinus Robillardi is easily sufficient
ground for establishing a new genus. Mortensen considers this feature over-
shadowed by certain peculiarities of the pedicellarise and the bare buccal mem-
brane, and consequently includes in his genus two species (darnleyensis and
inconspicuus) which I have placed in Nudechinus. He has however described
two species (pulchellus and versicolor) which agree with Robillardi in the
extraordinary abactinal system (see Dan. Exp. Siam: Ech., p. 113 et seq.) and
there are two others in the M. C. Z. collections. There are thus five species
which belong in Gymnechinus, all characterized by the fact that unlike any
other known Echini, oculars I and II are in contact with the periproct. As
correlated characters, it may be mentioned that the buccal membrane is thin
and bare (save for the primordial ambulacrals) , the globiferous pedicellarise have
a tubular blade ending in a single prominent tooth and the test is distinctly
flattened. In connection with the excentric position of the periproct, which
lies almost wholly on the right hand side of the antero-posterior axis, it is not
uncommon to find either genital 3 or genital 4 excluded from the periproct by
the two adjoining genitals.

GYMNECHINUS MEGALOPLAX. 287

The five species of this genus occur in the Indian region ranging from Mauri-
tius and the Persian Gulf to Australia, Siam, and Macclesfield Bank. They are
all of small size, rarely exceeding 25 mm. in diameter. They may be distin-

»

guished from each other as follows: —

Spines white; auricles meet and form an arch.

Suranal small, without a tubercle, in contact with only two genitals; gill-slits

shallow Robillardi.

Suranal large (much larger than any one of the oculars) in broad contact with four

genitals and, like each of them, carrying a tubercle; gill-slits deep megaloplax,

Spines more or less colored; auricles not united.

Interambulacral plates at ambitus with the tubercles arranged around the primary

or scattered, not arranged in definite horizontal series.
Primary spines rose-purple or reddish, with light tips, not banded .... pulchellus.

Primary spines with 1-3 bands of red versicolor.

Interambulacral plates at ambitus, each with at least four large secondaries which
form a horizontal series with the primary tubercle; sutures between inter-
ambulacral plates, especially the horizontal ones, well marked; primary spines
violet or pale reddish with a broad indistinct violet band, light tipped epislichus.

Gymnechinus megaloplax,1 sp. nov.
Plate 102, figs. 2, 3.

The single specimen at hand is 17 mm. in diameter and 8 mm. high. The
test is thus somewhat flattened and the actinostome is slightly sunken. The
abactinal system measures 6 mm. in diameter along the axis IV-1 but only a
trifle over 5 mm. along the axis V-2. The actinostome is symmetrical and is
about 7 mm. across. The longest primaries are 4 or 5 mm. in length and occur
at and below the ambitus. There are 15 interambulacral and 19 ambulacral
plates in each column. The ambulacra are very wide, only a little narrower
at the .ambitus than the interambulacra. There is a well-developed primary
tubercle on every coronal plate, those of the ambulacra being slightly smaller
than those of the interambulacra. The secondary tubercles are small and
scattered; only at the ambitus, in the interambulacra, do they become notice-
able; there, there is one on each end of the plate, which is much larger than the
others, but not nearly as large as the primary. The genital plates (PL 102, fig. 3)
vary much in size, 3 is the largest, 2 and 4 are a little smaller and of about equal
size, 5 is distinctly smaller, while 1 is much the smallest, being no larger than
ocular I. Each of the genitals, except 1, carries a single conspicuous tubercle.
The oculars are of about equal size and each of them carries a small secondary

1 fJif\o.<i (fif\aX) = big + 7r\d£ = a plate.

288 HAWAIIAN AND OTHER PACIFIC ECHINI.

and several miliary tubercles. Oculars I and II are broadly in contact with the
periproct but the other three are completely excluded. The periproct is very
large and of peculiar shape. The end opposite the anal opening is covered by a
suranal plate which is larger than an ocular, and resembles that of many Saleni-
idse in being in contact with four genitals; but the arrangement is unlike any
known in that family in that it is genital 1 which does not touch the suranal.
The suranal bears a tubercle similar to those on the genitals. The remainder
of the periproct is covered by ten or a dozen plates of which the largest adjoin
the suranal. The poriferous areas are broad (each about equals half the inter-
poriferous area) for the pores are rather large and the arcs are quite oblique.
Near the peristome the areas become narrower as the arcs become more vertical.
The actinostomal membrane (PL 102, fig. 2) is thin and perfectly naked, save
for the very small buccal plates. The gill-cuts are deep and sharply defined.
The auricles meet across the ambulacra forming a closed arch. The spines are
fairly long, moderately stout, and rather blunt, the primaries extending far
beyond the secondaries.

The pedicellarise resemble those of G. Robillardi. The globiferous are common
and variable in size and proportions. The valves range from .35-.50 mm. in
length, of which from one third to one half is the tubular blade. The ophicepha-
lous are very common and show little diversity, except in the amount of
constriction at the base of the blade. This may be distinct or scarcely noticeable.
The valves are about .25 mm. in length, besides the loop which adds .05-.08
more. No tridentate pedicellarise were found. The triphyllous are not very
common nor variable; the valves are about .10 mm. long. No spicules or
sphseridia were noted.

The specimen is perfectly white save for the tube-feet and buccal membrane
which are light brown ; on neither test nor spines is there a trace of color.

This specimen was dredged in the Persian Gulf in 1894 or 1895 by Capt.
F. W. Townsend, but there are no data to show either the exact locality or the
depth. If the remarkable suranal plate proves to be a constant character, it
will serve to distinguish this species at once from all other known Echini. If
however it is a variable feature, carried to an extreme in this particular indi-
vidual, then the coloration and the deep gill-cuts become more important specific
characters. Should the gill-cuts also prove to be a variable feature, megaloplax
might be regarded as a synonym of Robillardi as it would be impracticable to
distinguish between them.

GYMNECHINUS EPISTICHUS. 289

Gymnechinus epistichus,1 sp. nov.
Plates 93, figs. 22, 23; 102, figs. 4, 5.

The specimens are so nearly of a size that the measurements of the type will
answer for all. It is 26 mm. in diameter and only 12 mm. high, showing clearly
the flattened form of the test characteristic of the genus. The abactinal system
is 7 mm. in diameter along the axis IV-1 and nearly as much along the axis
V-2. The actinostome is a little larger, measuring 9 mm. across. There are
18 interambulacral plates in each column and 24 ambulacrals. The primary
spines are about 5 mm. long at the ambitus. The tuberculation of the test is a
characteristic feature for while there is, as usual, a primary tubercle on every
coronal plate, the secondaries are also prominent and form with the primaries,
at least on the mid-zone, nearly horizontal series. Thus on each interambulacral
plate at or just above the ambitus we have a series, consisting of a primary
tubercle in the middle and a pair of secondary tubercles on each side; at the
outer (radial) end of the plate, there are often two pairs of secondaries. In the
ambulacra, which are not nearly as wide at the ambitus as the interambulacra,
the secondary tubercles are few and irregularly scattered. The poriferous area
is moderately wide in the mid-zone, where the pores are fairly large, and the arcs
quite oblique, but actinally they become very narrow, the pores being small and
the arcs more nearly vertical. The genital plates (PI. 102, fig. 5) except 1 are
nearly of a size, decidedly higher than wide, and each carries two or three small
tubercles. Genital 1 is very small and carries no tubercle. The oculars are
about equal to each other and carry one to three tubercles apiece. Oculars I
and II are fully in contact with the periproct but the others are broadly exsert.
The periproct is not specially peculiar, but is covered by a dozen or more plates
of which the largest adjoin genital 3. The actinostomal membrane (PL 102,
fig. 4) is thin and perfectly bare, save for the buccal plates which are so small,
they scarcely exceed in diameter the tube-feet which they bear. The gill-cuts
are deep but are rather broad and not very sharply defined. The auricles are
low and do not meet across the ambulacra. The primary spines are moderately
stout, while the secondaries, though not much shorter, are very much more
slender.

The pedicellariae are very similar to those of the other members of the genus.
The globiferous are rather rare and have valves (PI. 93, fig. 22) which measure

1 iiri = in + trrfyas = a row, rank.

290 HAWAIIAN AND OTHER PACIFIC ECHINI.

about .40 mm. in length. The ophicephalous are common but small; the valves
are only .20-.25 mm. besides the loop, and are very slightly constricted at the
base of the blade. The tridentate seem to be rare and the only ones found were
on the abactinal surface. The valves (PI. 93, fig. 23) measure about .55 mm.
in length and resemble those of Nudechinus darnleyensis as figured by Mortensen
but are somewhat flatter and more compressed at base of blade.

The test is dull gray with a purplish cast which becomes deeper on the abacti-
nal median areas of both ambulacra and interambulacra, which thus stand out
more or less clearly as darker regions; actinally the whole test is nearly white.
The spines are light, particularly the secondaries. The primaries are violet or
pale reddish with a broad indistinct violet band, and with the tip light. The
general appearance of the preserved specimens is very dull.

The specimen selected as the type of this species was collected by Semper at
Bohol, Philippine Islands, in 6-10 fathoms of water. It came into the M. C. Z.
collection in 1873 but was never identified. There are also two other specimens,
which bear the label "Echinus darnleyanus Wood. Australia," but there is no
clue as to when they were received or whence they came. The label would
seem to confirm Mortensen's view that Wood confused more than one species
under his Echinus darnleyensis, for these specimens are certainly very different
from those in the British Museum which bear that name.

TEMNOPLEURID^E Desor.
GENERAL CONSIDERATIONS.

The description of the Temnopleuridse as Echinidae with sculptured or pitted
tests expresses briefly the only difference between the two families. It is difficult
to determine just what the real importance of this character is, and the recogni-
tion of the Temnopleuridae as a separate family is merely a matter of convenience.
Where the sculpturing of the test is conspicuous or the pits are deep and large,
the general appearance of the animal seems to warrant the separation from the
Echinidae, but where the sculpturing is very faint as in many specimens of
Prionechinus, or the pits and grooves are so small or shallow as to be seen only
with difficulty as in some specimens of Salmacis (and other genera), the line
between the families is virtually blotted out and one can but wonder whether
it really exists. It is not strange therefore that specimens of Echinidse have
often been identified as Salmacis, and specimens of Genocidaris and related

TEMNOPLEURID/E. 291

genera as young Echinidse. However such identifications have usually resulted
from superficial or careless examination of the specimens and it is doubtless true
that a careful study of any specimen will enable one to determine to which family
it properly belongs.

THE SPINES, PEDICELLARLE, SPH/ERIDIA, AND SPICULES.
Plate 93, figs. 16-21, 24-31, 33-36.

The resemblance of the Temnopleuridse to the Echinidse in all these charac-
ters is very marked and characteristic features are not found in any of them.
Our researches add little of importance to the detailed accounts already published
by Mortensen. The spines sometimes furnish helpful characters for distinguish-
ing species though little reliance can be placed on them. In Amblypneustes,
for example, Mortensen makes the form of the tip of the spine, whether club-
shaped or not, an important specific character. This is not very satisfactory,
as it is difficult to decide in the case of some spines whether the tip is club-
shaped or not and in some specimens both club-shaped and simply blunt spines
occur. In Temnotrema, the smoothness or roughness, and the form of the tip,
of the spines is often a useful means of separating species, but even in this genus
the lines between the different kinds of spines are sometimes difficult to observe.

The pedicellarise are so similar to those of the Echinidse, that a description
of their characteristic features would be but a repetition of what has already
been stated (p. 236-238). The spicules are uncommon and hard to find; they
are triradiate in Hypsiechinus, according to Mortensen, but elsewhere are
bihamate or simply bow-shaped. The sphseridia show no peculiarities.

THE GENERA AND SPECIES OF RECENT TEMNOPLEURID.E.

As first pointed out by Duncan, the Temnopleuridse fall into two groups
which may well be ranked as subfamilies. One of these has been designated as
Temnopleurinse and this name has been generally agreed to, but for the other
group several names have been suggested. Of these the one proposed by Morten-
sen but later rejected by him, Trigonocidarinse is preferable. Our choice is due
to the fact that it is based on a recent genus and for obvious reasons, it is desir-
able that whenever possible names should be based on recent types. No doubt
fossil species are of equal importance with the recent in determining the history
and natural relationships of Echini; indeed they have often proved of far greatsr

292 HAWAIIAN AND OTHER PACIFIC ECHINI.

importance, but since it is only the recent species of which the entire structure
and the complete life-history can be discovered they should, when possible,
furnish the types upon which names are based. It is to be regretted that the
number of fossil Temnopleuridse, which have been named, is so large as to make
it impossible to take them into account in the present work, and the following
review of the recent species of the family is given with the full understanding
that a similar review of the fossil forms may make it necessary to modify some
of the statements and to alter the generic limits herein laid down.

In no other family of Echini is there more general agreement as to the classifi-
cation of the recent species than is found here, no doubt owing to the fact that
all students of the group have sought for generic characters in the test and have
made use of the spines and pedicellariae only in a very subordinate way. The
family contains at least fifty living species and probably many more, for specific
limits in some genera are as yet unsatisfactorily known, and moreover the
smallest species of regular Echini belong here and these are often overlooked by
collectors. Even if collected, their identification being difficult, they are often
erroneously labelled, frequently being considered the young of large species to
which they have no close relationship. It is therefore to be expected that more
species will be added hereafter to the Temnopleuridse than to any other family.

As stated above the family falls into two sections according to the structure
of the test. In one group (Trigonocidarinse) the coronal plates, especially in
the interambulacra and abactinally, are more or less sculptured, i. e. the surface
of the plate is furrowed or is ornamented with elevated ridges or knobs or both.
The amount of this sculpturing is very variable even within a given species,
some specimens showing it plainly while in others it is very faint. Thus we have
specimens from the West Indies, which connect Trigonocidaris albida with
Lytechinus euerces so closely that it is to be feared the line drawn in this Memoir
is quite arbitrary. Specimens showing any "sculpturing," have been assigned
to Trigonocidaris and those where it was lacking to Lytechinus, but many of the
latter have a most close resemblance to the former. It may be added that most
of these doubtful cases are immature, though occasionally a full grown Trigono-
cidaris has the " sculpturing " very faint. In Prionechinus again, the amount
of sculpturing is oftentimes very slight, no more than is found in young
Echinidse of the same age, and the placing of such specimens in the Temno-
pleuridse is on the strength of their general appearance rather than on anything
tangible. Sometimes the abactinal system is sculptured even when the test is
not but this is not usually the case. In the other group (Temnopleurinae) the

TEMNOPLEURID.E. 293

test is not sculptured (with rare exceptions) but along the horizontal sutures or
at the angles of the plates in the median line of each area are found deep pits,
the form and size of which are very variable. They may be so large as to furnish
the most conspicuous character of the test or they may be so small as to be easily
overlooked, and they may be practically wanting in old specimens. In addition
to these pits, the Temnopleurinse have another characteristic feature in the
union of the coronal plates by dowelling. No very extensive study of this
feature has been made to determine its significance or its relation to youth and
maturity. While it may be of more morphological importance than seems likely,
it is so difficult a feature to detect that it is of little practical importance in
distinguishing species and genera.

Aside from the structure of the corona, the character of the abactinal system,
the condition of the buccal membrane, and the arrangement of the buccal plates
all furnish characters of importance in classification. It is rather remarkable
that while the abactinal system of the Temnopleuridse retains, with few exceptions,
the primitive character of having all the ocular plates small and broadly exsert,
the actinostomal membrane is with few exceptions thin and bare and even
the buccal plates may show indications of resorption or the buccal tube-feet be
reduced to five. The crenulation of the primary tubercles is a matter of some
importance in this family as it furnishes the most obvious character for dis-
tinguishing certain genera. The arrangement of the pore-pairs is of no little
importance in distinguishing some species of the more highly specialized genera,
and the number of coronal plates is frequently of value as a character. Size
and color are often of very great significance and certain species are only to be
distinguished by some characteristic feature of their coloration. As a rule the
genera are not difficult to distinguish, if reasonable care is used, but in certain
genera, notably Prionechinus and Amblypneustes, the separation of the species
is very difficult and the results are not always satisfactory. The following table
shows those characters which have proved of most use and reliability in arranging
the forty-five species, which have been personally examined, all of the sixteen
genera except Lamprechinus being represented.

Test without pits on sutures or at angles of coronal plates, but more or less exten-
sively ornamented with grooves, depressions, ridges or knobs; coronal plates

not united by dowelling TRIGONOCIDARIN.E.

Sculpturing of test, when well marked, consists of vertical or diagonal ridges,
usually connecting tubercles, but it may be so faint as to show little
character; it is sometimes best marked on abactinal system.
" Actinostomal membrane fully plated; buccal plates not conspicuous
among the others.

294 HAWAIIAN AND OTHER PACIFIC ECHINI.

Sculpturing of test more or less well marked; madreporic genital

with 10-15 pores or more; no sexual dimorphism Trigonocidaris.

Sculpturing of test slight or wanting; madreporic genital with only
2 or 3 large pores; marked sexual dimorphism; abactinal system

of female greatly elevated and modified Hypsiechinus.

Actinostomal membrane naked, at least outside the ring of buccal plates

which are usually conspicuous.
Periproct more than half covered by a single plate; actinostomal

membrane with only the buccal plates Genocidaris.

Periproct not half covered by any one plate; peristomal membrane

with at least a few plates proximal to the buccal circle.
Periproctal plates few, large, and more or less glassy.

Actinostomal membrane with few plates proximal to the
buccal circle; coronal plates usually with distinct sculp-
turing; ocular plates more or less swollen and genitals

usually more or less sculptured Orechinus.

Actinostomal membrane well plated within circle of buccal
plates; coronal plates only faintly sculptured; ocular
and genital plates flat, smooth, and not sculptured . . Lamprechinus.

Periproctal plates numerous and not glassy Prionechinus.

Sculpturing of test consists of a series of depressions along horizontal inter-

ambulacral sutures; there are 4-6 of these on each suture at ambitus . . Opechinus.
Test not sculptured (or rarely so) but with pits on horizontal sutures or at angles

of coronal plates; latter united by dowelling TEMNOPLEDRIN^B.

Primary tubercles distinctly crenulated.

Coronal plates, at least abactinally, with deep and more or less conspicu-
ous pits at their sutural angles; these pits are usually extended more
or less distinctly along the horizontal sutures; each interambulacral
plate at ambitus with 1 or 3 (very rarely 4) primary tubercles . . . Temnopleurus.
Coronal plates usually with small sutural pits or none; if the pits are large
enough to form a furrow along the horizontal suture, the inter-
ambulacral plates at ambitus each carry 4-9 subequal primary

tubercles in a horizontal series Salmacis.

Primary tubercles small, not crenulated (faintly so in some large specimens).
Every ambulacral plate with a primary tubercle close beside the porif-
erous area.

Coronal plates with deep, conspicuous, usually oblong pits at the
sutural angles and on their horizontal sutures; poriferous areas
narrow, the pore-pairs in an approximately vertical series . . . Temnotrema.
Coronal plates with small sutural pits or none.

Median abactinal interambulacral areas (and usually ambulacral
also) more or less extensively bare and free from spines
and tubercles.

Interambulacral plates low and numerous (21 or 22 in each
column in specimens 20 mm. h. d.)i pore-pairs dis-
tinctly biserial Mespilia.

Interambulacral plates high and few, not more than 18,
even in largest specimens (25 mm. h. d.); pore-pairs
usually'monoserial .

TRIGONOCIDARIS ALBIDA. 295

Actinal interambulacral plates not lower than abactinal
nor noticeably different in shape; their primary
tubercles not arranged in horizontal series of 3
or 4; bare interambulacral space extended more
or less conspicuously along the horizontal sutures;
oculo-genital ring not conspicuously dark .... Microcyphus.
Actinal interambulacral plates distinctly lower and
more oblong than abactinal; primary tubercles on
each one in a horizontal series of 3 or 4; bare inter-
ambulacral spaces not extended noticeably along
horizontal sutures; oculo-genital ring more or less

black or blackish Salmacopsis.

Median abactinal interambulacral areas not bare but more or

less covered by secondary or miliary spines Amblypneustes.

Only every second or third ambulacral plate at and above ambitus with

a primary tubercle close beside the poriferous area.
Ocular plates typically all exsert; interambulacral plates low and
numerous (27-30 in specimens 23 mm. h. d.), each with several
primary and some secondary tubercles; actinostomal membrane

naked save for buccal plates Holopneusles.

Oculars I and V insert; interambulacral plates high and few (18 in
specimen 23 mm. h. d.) each with 1 small primary, and 2-5 well
scattered, small secondary tubercles; actinostomal membrane
with numerous small plates, especially inside the buccal circle . Goniopneustes.

TRIGONOCIDARIS.

A. Agassiz, 1869. Bull. M. C. Z., I, p. 263.
Type-species, Trigonocidaris albida A. Agassiz, 1869, 1. c.

Although a second species of this genus, from the Hawaiian Islands, was
described in 1907 (A. Agassiz and Clark, Bull. M. C. Z., L, p. 242), comparison
of those specimens with a large series from the West Indies shows that the
peculiarities supposed to distinguish the Hawaiian form are not reliable but
are quite inconstant and the genus is thus monotypic.

Trigonocidaris albida A. Ag.

Trigonocidaris albida A. Agassiz, 1869. Bull. M. C. Z., I, p. 263.
Trigonocidaris albidoides A. Ag. and Cl., 1907. Bull. M. C. Z., L, p. 242.

Examination of a large series of specimens from the West Indies shows so
great diversity in the sculpturing of the test, in the number of spines on the
abactinal system and in the amount and distribution of the red coloring, that
it is impossible to satisfactorily distinguish the specimens collected by the
"Albatross" in the Hawaiian Islands. This would seem to indicate that de Mei-
jere was quite correct in identifying as albida, the Trigonocidaris collected by

296 HAWAIIAN AND OTHER PACIFIC ECHINI.

the "Siboga" in the Sulu Archipelago in 153 fms. Very likely this species
will prove to be intertropical in its distribution in water 100-500 fathoms deep.

The specimens collected by the "Albatross" were taken at the following
places: —

Station 3859. Off Mokuhooniki Islet, Pailolo Channel, Hawaiian Islands.
Bott. temp. 60.5°-60.2°. 138-140 fathoms. Fne. s., m.

Station 3863. Off Mokuhooniki Islet, Pailolo Channel, H. I. Bott. temp.
61°-60°. 127-154 fathoms. Brk. co., crs. g., r.

Station 3892. Off Mokapu Islet, north coast of Molokai, H. I. Bott.
temp. 42.5°. 328-414 fathoms. Fne. g. s.

Station 4045. Off Kawaihae Light, west coast of Hawaii, H. I. Bott.
temp. 49°. 147-198 fathoms. Co. s., for.

Bathymetrical range, 127-414 fathoms. Extremes of temperature. 61°-
42.5°.

Five specimens.

HYPSIECHINUS.

Mortensen, 1903. "Ingolf" Ech., pt. 1, p. 81.
Type-species, Hypsiechinus coronatus Mortensen, 1903, 1. c.

Mortensen (1904, Siam Ech., p. 57) holds that this genus is "the most primi-
tive of all the Temnopleuridge " and refers to the globiferous pedicellariae and cal-
careous spicules as evidence of the fact. There may be suggested as confirmative
facts, the structure of the abactinal system of the male, the plating of the buccal
membrane, the rudimentary condition of the auricles and the paucity of spines.
It is interesting to note that Trigonocidaris is only a little more specialized, the
auricles being a little higher, the valves of the globiferous pedicellarise having
only a single lateral tooth on the left side and the spicules being bihamate.

Hypsiechinus is monotypic and is known only from deep water (450-799
fathoms) southwest and west of Iceland.-

GENOCIDARIS.

A. Agassiz, 1869. Bull. M. C. Z., I, p. 262.
Type-species, Genocidaris maculata A. Agassiz, 1869, 1. c.

The discovery of additional species of this genus, agreeing in certain interest-
ing details with the type species, seems to confirm the view of Pomel and Mor-
tensen that the generic name originally bestowed on the West Indian specimens
had better be retained. The exceptionally large suranal plate, which is no

GENOCIDARIS APODA. 297

doubt a primitive character, is combined with a perfectly naked buccal
membrane, a specialized feature, and quite specialized globiferous pedi-
cellarise; an unusual combination which makes the genus easy to recognize.
The species originally described from the West Indies (maculata) is now known
from the eastern Atlantic and from the Mediterranean, according to Mortensen
and Kcehler. A second species (decipiens) was collected by the "Siboga" at
Saleyer, Flores and Sumbawa, D. E. I., in shallow water. De Meijere's descrip-
tion of the abactinal system of this species would lead one to suppose that the
arrangement of the plates is very similar to that which is characteristic of Gym-
nechinus. Examination of two specimens, received from the Amsterdam
Museum, shows that the periproct is not noticeably excentric, but that the anus
is crowded far to the right by the huge suranal plate. Genital 1 is rather low
and genital 3 rather high but there is nothing like the disproportion shown in
Gymnechinus. None of the ocular plates reach the periproct but I and II are
appreciably nearer than the others, which suggests the condition of Gymnechinus.
Owing to the deep sculpturing of the plates, it is not easy to make out the sutures
and de Meijere was probably either misled in supposing two oculars to be insert
or else the specimen on which his statement is based, was exceptional. That it
is not a question of size is evident from the fact that his largest specimen was
only seven millimeters h. d. while our larger one is over six. Comparison with
the other species of the genus shows that the excentricity of the periproct and
of the anus are little if at all greater in decipiens than in the others. The third
species (apoda) was taken by the "Albatross" off southern Japan in rather
deeper water than that in which its East Indian neighbor was found. The
three species may be distinguished from each other as follows : —

Buccal tube-feet 10, one for each buccal plate.

Buccal plates large forming a nearly closed ring around mouth maculata.

Buccal plates small, not forming a ring around mouth, as one plate in each pair is

decidedly more distal than its fellow decipiens.

Buccal tube-feet 5, wanting on one buccal plate of each pair apoda.

Genocidaris apoda A. Ag. and Cl.

Genocidaris apoda A. Agassiz and Clark, 1907. Bull. M. C. Z., LI, p. 126.

Plates 93, figs. 16, 17; 100, figs. 1-3.

The largest specimen is 7 mm. in diameter and rather more than 3 mm. high.
The abactinal system is well over 4 mm. across while the actinostome is not over

298 HAWAIIAN AND OTHER PACIFIC ECHINI.

3.5. There are about eight coronal plates in each column in both ambulacral
and interambulacral areas. The primary spines are all broken but in a smaller
specimen, 5 mm. in diameter, they are 4-5 mm. long, and thus equal to h. d.
The sculpturing of the test is well marked in the mid-zone and actinally but
abactinally the plates become quite smooth, particularly the plates of the
abactinal system itself. The latter are also remarkable for their relatively large
size and freedom from tubercles. The genital plates (PI. 100, fig. 3) are nearly
of a size, though 1 is somewhat smaller than the others. Each plate carries a
minute tubercle on the distal margin, and the madreporic genital may have
one or two smaller ones in addition. The genital pore is conspicuous and is
surrounded by a raised ring. The madreporic pores are very few, only 6-10 in
number; they are situated in an elevation of the plate surface. The ocular
plates are nearly triangular and carry several minute tubercles, one of which
may be distinctly larger than the others. In the largest specimen, the oculars
are all excluded from the periproct, but in the others ocular I is more or less
broadly insert. The periproct is very large and is covered by a huge suranal,
nearly as big as a genital, and a number of very small plates near the anus,
which lies very close to genital 1 . None of the periproctal plates carry tubercles
of any kind. The ambulacra are decidedly narrower than the interambulacra
at the ambitus. The poriferous areas are very narrow and the arcs of pores
are vertical or nearly so. The primary tubercles are of quite unequal sizes,
particularly in the ambulacra.

The actinostome (PL 100, fig. 2) is decidedly smaller than the abactinal
system. The membrane is thin and perfectly naked save for the minute buccal
plates. The latter are very remarkable for their relative size and arrangement.
At first sight there seem to be but five, as shown in fig. 2, PL 100, but more
careful examination shows that instead of a single plate in each ambulacrum
there are really two lying closely side by side, one of normal size with a well-
developed tube-foot and the other very small and with no trace of a tube-foot.
Placing the specimen with the mouth up and ambulacrum III anterior, it is seen
that, beginning in ambulacrum I, the small and large plates are arranged in the
following sequence — small, large; small, large; large, small; small, large;
large, small; that is the large plates are 16, 116, Ilia, IV6, Va. It will be noted
that while this is the reverse of Loven's formula, it accords exactly with what
Loven shows to be the condition in a very young Strongylocentrotus ("Echino-
logica," PL IV, fig. 25) and what Jackson found in young Echinus (Mem. Boston
Soc. Nat. Hist., VII, p. 119), for in these cases only one plate of each pair of

GENOCIDARIS APODA. 299

primordial ambulacrals has a tube-foot and that is the one which according to
LoveVs formula should be the smaller. In Temnotrema Doderleini .(Mortensen)
there are but five primordial ambulacral plates, as the original description states.
Examination of a specimen shows that, judged by their position with reference
to the poriferous areas, these plates are likewise Ib, lib, Ilia, IV6, Va. It seems
probable therefore that when a tube-foot fails to become associated with one
of the primordial ambulacral plates, that plate fails to develop and remains
rudimentary as la, Ha, lllb, IVa, and V6 are in Genocidaris apoda or even be-
comes quite resorbed as the same series does in Temnotrema Doderleini. But
why tube-feet fail to become associated with these particular plates still remains
to be explained. It may be noted here that in Genocidaris decipiens de Meij.
the arrangement of the primordial ambulacral plates accords with Loven's law,
for while each plate bears a tube-foot, the two plates of each pair are unequally
distant from the mouth, and the distal plate is the larger; the sequence of these
distal plates is la, Ha, III6, IVa, Vb as would naturally be expected.

The primary spines of apoda are remarkably long and slender and so easily
broken, that whole ones have been observed only in the smallest specimen.
Pedicellarise are not common. Only one globiferous was found; the valves
(Plate 93, fig. 16) are about .15 mm. long, the blade is not as tubular as in macula-
tus and the end tooth is not particularly long. The ophicephalous pedicellariae
have the valves (PI. 93, fig. 17) about .20 mm. in length and distinctly con-
stricted at base of blade. No tridenlate or triphyllous were found, nor were
either sphseridia or calcareous spicules noted.

The test and spines are white with a grayish tinge and in the small specimen
the terminal half of the longer primaries is red.

This little sea-urchin differs so decidedly from the other species of Genoci-
daris that it cannot fail to be recognized. The plates of the abactinal system
are so much larger, thinner, and smoother than in decipiens or maculata that even
aside from the peculiarly characteristic buccal plates, it would be easily dis-
tinguished.

It was taken by the "Albatross" at only two stations: — -

Station 4891. Southwest of Goto Islands, Japan; 32° 27' N., 128° 34' E.
Bott. temp. 50.2°. 181 fathoms. Gy. s., brk. sh., r.

Station 4904. Southwest of Goto Islands, Japan; 32° 31' 20" N., 128° 32' 40"
E. Bott. temp.? 107 fathoms. Fne. gy. s., brk. sh.

Three specimens.

300 HAWAIIAN AND OTHER PACIFIC ECHINI.

ORECHINUS.

Doderlein, 1905. Zool. Anz., XXVIII, p. 622.
Type-species, Trigonocidaris monolini A. Agassiz, 1879. Proc. Amer. Acad., XIV, p. 203.

In his report on the "Siboga" Echini, de Meijere removed T. monolini from
Trigonocidaris and placed it in Genocidaris but it is not clear on just what
grounds he made the change. Doderlein has showed however that the characters
of both the abactinal system and buccal membrane prevent its association with
Genocidaris and so he established the genus Orechinus for its reception. This
genus is so near Trigonocidaris that it might be regarded as superfluous, but
the lines of development are probably a little easier to follow if the two genera
are retained. So far as known Orechinus is a monotypic genus confined to the
deep waters of the Indo-Pacific region.

Orechinus monolini D6d.

Trigonocidaris monolini A. Agassiz, 1879. Proc. Amer. Acad., XIV, p. 203.
Orechinus monolini Doderlein, 1905. Zool. Anz., XXVIII, p. 622.

Plate 93, figs. 27-31.

The "Challenger" took a single specimen, 8 mm. in diameter, in 520 fathoms
near the Kermadec Islands and the "Valdivia" also took a single, somewhat
larger specimen in 278 fathoms off South Africa. The "Siboga" found the
species fairly common, in the Dutch East Indies, taking it at ten widely scattered
stations in from 262 to 1 156 fathoms. These specimens ranged from 4 to 14 mm.
in diameter. Finally the "Albatross" found monolini not uncommon among
the Hawaiian Islands, and brought home a series of specimens ranging from 6 to
22 mm. in diameter. In spite of the larger size of some of these specimens, there
is little to add to the details that have been published by de Meijere and Doder-
lein. Globiferous pedicellarise are rare in some specimens but fairly common in
others; the valves (PI. 93, figs. 27, 28) are about .60 mm. long, with the base
.30 mm. wide and the terminal tooth about .15 mm. long. The ophicephalous
are common; the valves (which have been figured by Doderlein) are .20-.25 mm.
long, besides the loop which adds .05-. 10 more. The tridentate are always rare
and often wholly wanting, which is unfortunate as they are rather characteristic ;
the valves (PI. 93, figs. 29, 30) are .40-.70 mm. long, with the blade .12-.18 mm.
wide; they are strongly curved and meet only at the tip. The triphyllous are
rare or at least hard to find; the valves (figured by Doderlein) are about .13 mm.

ORECHINUS MONOLINI. 301

long. The primary spines at the ambitus, if unbroken, often exceed the diameter
of the test. The form of the test is very variable, the vertical diameter ranging
from .45 to over .60 h. d. The distinctness of the sculpturing on both the coronal
plates and the abactinal system, is also very variable and in some specimens is
quite faint. The ocular plates are, without exception in these specimens,
excluded from the periproct, which is covered by a variable number of plates.
The buccal membrane is thin and very bare, for although the buccal plates are
large, the small plates proximal to them are few and scattered. The auricles
touch over the ambulacra, or nearly so, but they are not united to form an arch.
The color of test and spines is more or less white, very uniform, but sometimes
strongly tinged with brown.

The "Albatross" took this species at the following stations: —

Station 3839. Off Lae-o Ka Laau Light, Molokai. Hawaiian Islands. Bott.
temp. 46.3°. 259-266 fathoms. Lt. br. m., s.

Station 3865. Off Mokuhooniki Islet, Pailolo Channel, H. I. Bott. temp.
44.8°-45°. 256-283 fathoms. Fne. vol. s., r.

Station 3914. Off Diamond Head, Oahu, H. I. Bott. temp. ? 289-292
fathoms. Gy. s., m.

Station 3918. Off Diamond Head, Oahu, H. I. Bott. temp. 44.5°. 257-
294 fathoms. Wh. s., m.

Station 4085. Off Puniawa Point, Maui, H. I. Bott. temp. ? 267-283
fathoms. S., sh.

Station 41 17. Off Kahuku Point, Oahu, H. I. Bott. temp. 45.6°. 253-
282 fathoms. Co. s., for.

Station 4125. Off Kahuku Point, Oahu, H. I. Bott. temp. 36.4°. 963-
1124 fathoms. Br. m., for., r.

Station 4126. Off Kahuku Point, Oahu, H. I. Bott. temp. 35.5°. 743-
1278 fathoms. Gy. s., for.

Station 4131. Off Hanamaulu, Kauai, H. I. Bott. temp. 43.7°. 257-309
fathoms. Fne. gy. s.

Bathymetrical range, 253-1278 fathoms. Extremes of temperature, 46.3°-
35.5°.

Twenty-nine specimens.

302 HAWAIIAN AND OTHER PACIFIC ECHINI.

LAMPRECHINUS.

Doderlein, 1905. Zool. Anz., XXVIII, p. 622.
Type-species, Lamprechinus nitidus Doderlein, 1905, 1. c.

Not having seen a specimen of this genus, I am in doubt as to its validity.
Judging from Doderlein's description and figures it must be very near Orechinus,
and so far as can be determined the points which separate it from that genus are
quite trivial. The genus is monotypic and since the only specimens were taken
off South Africa in 276 fathoms, at the same place where Orechinus monolini
was taken, its supposed characters may represent only individual diversity of
Orechinus.

PRIONECHINUS.

A. Agassiz, 1879. Proc. Araer. Acad., XIV, p. 202.
Type-species, Prionechinus sagittiger A. Agassiz, 1879, 1. c.

Of all the genera of Temnopleuridae this is not only the least well defined,
but it is also the one within which specific limits are hardest to draw. There
are several reasons for this, particularly the small size, the occurrence only in
deep water and the lack of sufficient material for careful comparative study.
Test-sculpturing reaches its lowest limit in this genus; in some of the species,
it is quite wanting and in others it is chiefly on the abactinal system. What
characters are really of specific importance, it is hard to determine for so little
of their significance is known. The position of the genital pore on the plate
is used as a specific character but it may be only sexual, and the presence
or absence of tube-feet on the buccal plates may be a matter of age. And the
question whether color has any significance is also of importance. The result
of my efforts to distinguish the described species follows but it is doubtful if all
these species are valid, or if valid whether they are best distinguished by the
characters used. It is probable that de Meijere is correct in assigning Cottaldia
Forbesiana A. Ag. to this genus, although lack of material makes it very un-
certain what its true relationships are. It seems quite probable, as Dr. Morten-
sen has pointed out with rather undue emphasis ("Ingolf" Ech., pt. 1, p. 82),
that more than one species was included under the name Prionechinus sagittiger
in the report on the "Challenger" Echini and that unfortunately the drawings
on Plate VIa were not all taken from the same specimen. It seems best to
follow Mortensen and de Meijere in taking the specimen from "Challenger"
Station 218 (shown in "Challenger" Rept., PI. VIa, fig. 11) as the type of the

PRIONECHINUS. 303

species, as any other course would result in great confusion; and this decision
is strengthened by the fact that the specimens collected by the "Siboga" and
determined by de Meijere as sagittiger seem to be identical with it. This state-
ment is made from the study of a specimen from the "Siboga" material now in
the M. C. Z. collection. Mortensen states that the specimen from "Challenger"
Station 207 is not a Prionechinus, "it is, no doubt, a quite different genus."
As he gives no facts in support of this blunt assertion, except that "the pores
are really very large and form a straight line," the matter cannot be regarded
as settled. A reexamination of a specimen from "Challenger" Station 164
shows that it has ten buccal tube-feet and several large plates covering the
periproct, in both these respects differing from sagittiger. Apparently it was
from such a specimen that fig. 13, PI. VIa of the "Challenger" Report was
drawn and so far as can be seen such specimens belong to the species Agassizii.
But it is certainly very doubtful how important as a specific character the
absence of a tube-foot from one or more of the buccal plates, really is, and the
same is true of the presence or absence of several large plates in the covering
of the periproct. In view of the fact that when Prionechinus sagittiger was
described, very little indeed was known of the living members of the group now
called Trigonocidarinse, and that little was based on West Indian material, it
seems quite natural that the "Challenger" specimens from Stations 164, 207,
and 218 were all referred to one species, especially since in the character of the
spines and in their superficial appearance they are very similar. And it should
be borne in mind that it is not yet demonstrated that they do not belong to a
single species, for should Agassizii prove to be identical with sagittiger as is by
no means impossible, it will be difficult to determine how the specimens from the
three stations are to be separated. The specimens collected by the "Siboga"
and identified as Forbesiana by de Meijere, do not seem to belong to that species,
so far as can be judged from the small specimen at hand. But the agreement in
coloration is rather striking and the differences in tuberculation which we note
may be largely a question of age. Doderlein's species, Chuni, appears to be
very well characterized and this seems equally true of the Hawaiian species
sculptus. The species depressus from the Hawaiian Islands and ruber from Japan
are much more dubious.

The following table will show the characters that seem to separate these
seven species from each other: —

Test high, vertical diameter distinctly more than one half horizontal; buccal plates 10,
subequal, each with a tube-foot, the 5 pairs well separated from each other; peri-
proct covered by numerous, nearly equal graniform plates Chuni.

304 HAWAIIAN AND OTHER PACIFIC ECHINI.

Test more or less depressed; anal plates of variable number and size, one or more dis-
tinctly larger than the others.

Some of the buccal plates, usually 5, lack a tube-foot sagittiger.

All of the buccal plates carry tube-feet.

Genital pores near centre of genital plates.

Buccal plates large, all 10 usually more or less in contact with each other,
though in some cases adjoining pairs may be well separated; area
within the buccal circle more or less plated.
Color yellow-orange or whitish; spines white, yellow-orange at base Forbesianus.

Color red or reddish, the primaries lighter at tip ruber.

Buccal plates small, more or less widely separated from each other;
area within their circle not well plated; color dull purplish red,

often light sculptus.

Genital pores at distal tips of genital plates.

Buccal plates large, the two of each pair more or less in contact; area
within their circle more or less fully plated; genital plates little

sculptured, if at all Agassizii.

Buccal plates small, well separated; area within their circle with only a
few scattered plates; genital plates prettily sculptured depressus.

Prionechinus Chuni D6d.

Prionechinus Chuni Doderlein, 1906. Ech. Deuts. Tiefsee Exp. (" Valdivia" Ech.), p. 192.

This seems to be the most sharply defined and easily recognized species in
the genus. There is nothing to add to Doderlein's very complete and satis-
factory discussion of the "Valdivia" material.

The "Albatross" specimens, ranging from 2.5 to 11 mm. in diameter, were
taken at only a single place.

Station 4126. Off Kahuku Point, Oahu, Hawaiian Islands. Bott. temp.
35.5°. 743-1278 fathoms. Gy. s., for.

Seven specimens.

Prionechinus ruber A. Ag. and Cl.

Prionechinus ruber A. Agassiz and Clark, 1907. Bull. M. C. Z., LI, p. 126.

Plate 100, figs. 4-6.

The larger of our specimens is 11 mm. in diameter while the smaller is only
7 mm. The former is 5.5 mm. high and has the abactinal system and actino-
stome, each 5 mm. across, while the smaller specimen is 3.5 mm. high and its
abactinal system and actinostome are each a trifle more than 3 mm. in diameter.
In the larger specimen, the primary spines are 5 mm. long, and there are 11
interambulacral and 12 ambulacral plates in each column; in the smaller speci-

PRIONECHINUS RUBER. 305

men also, the spines are 5 mm. long, but the number of interambulacral and
ambulacral plates in each column is 9. The sculpturing of the test is almost
wholly wanting; it is simply indicated by slight depressions on the sutures and
granular swellings among the tubercles. The genital plates (PI. 100, fig. 6) are
large, subequal, and rather closely covered with small tubercles; the pores are
quite distinct near the centre of each plate. The madreporic pores are remark-
ably few (10 -12) and consequently the madreporite is not easily detected. The
oculars are of moderate size, well covered with small tubercles and are all fully
excluded from the periproct. The latter is larger than a genital plate and is
covered by about a dozen plates (or more) among which the suranal can be
distinguished by its size, though several other plates are nearly as large. The
ambulacra are very broad but do not equal the interambulacra at ambitus.
The poriferous areas are very narrow, the small pore-pairs being arranged in
nearly vertical arcs of three. The primary tubercles, of which there is one on
each coronal plate, form conspicuous vertical series on each column of plates,
those of the ambulacra about as large as the others. Secondary tubercles are
not numerous, usually two or three on each ambulacral plate and four to seven
on the interambulacral. There are also rather numerous granular elevations
which may be miliary tubercles, but their outlines are so indistinct that they may
not be tubercles at all. The actinostomal membrane (PL 100, fig. 5) is thin and
outside the circle of buccal plates is almost bare. The buccal plates are large,
each carries a tube-foot and they are nearly or quite in contact with each other,
thus forming a more or less complete closed ring, within which the membrane
is completely covered with small plates. The primary spines at ambitus are
about half as long as the test-diameter, while the abactinal ones are very much
smaller. The secondaries are short, stout, and somewhat club-shaped.

The pedicellarise are not remarkable but resemble those of the other members
of the genus, particularly those of sagittiger and Chuni as photographed by Doder-
lein. The globiferous are common and have the valves .30 -.35 mm. long; it is
not uncommon to find a lateral tooth on each side, instead of on only the left
side as is usually the case. In the ophicephalous, the valves are about .30 -.35 mm.
long. The tridentate are fairly common, but of very variable size, the valves
ranging from .35 to .80 mm. In the small ones the valves are relatively much
wider than in the large ones, but even in the latter the valves are a little wider
than in sculptus. No triphyllous were found, nor were spicules or sphseridia
observed.

In the smaller specimen the colors, are quite bright; the test is orange-red,

306 HAWAIIAN AND OTHER PACIFIC ECHINI.

brightest on the abactinal system, with the poriferous areas and the median
interambulacral areas, almost white. The spines are orange-red becoming
lighter at the tip. In the larger specimen, the colors are much less bright; the
test is paler and the whitish areas stand out in much less contrast. The spines
are darker at base, with more of a brown or purple shade in the red and the
whitish tips are more abrupt.

There is little reason to doubt that this species stands very near Forbesianus,
but it seems to differ not only in the color, but in the tuberculation and in the
plating of the buccal membrane. Unfortunately no reliable specimens of
Forbesianus are available for comparison, so that there is much doubt as to the
true relation of the two species. The specimens collected by the "Siboga"
and identified by de Meijere as Forbesianus seem, so far as can be judged from a
single specimen from Station 173, to be nearer to Agassizii. Were this specimen
a little larger and more deeply colored it would be very difficult to distinguish
it from the smaller specimen of ruber. On comparing the larger specimen of
ruber with the figures given in the "Challenger" Report of Forbesianus, it may
be noted that the tuberculation of both the abactinal system and the coronal
plates is much less crowded in ruber, and in view of this fact and the striking
difference in color, it seems better to await further material before finally de-
ciding whether the two species are identical.

The "Albatross" took ruber at the following stations: —

Station 4933. Off Kagoshima Gulf, Japan; 30° 59' N., 130° 29' 50" E.
Bott, temp. 56°. 152 fathoms. Rky.

Station 4967. Between Kobe and Yokohama, Japan; 33° 25' 10" N., 135°
37' 20" E. Bott, temp. 45.9°. 244-253 fathoms. Br. m., s., for.
Two specimens.

Prionechinus sculptus A. Ag. and Cl.

Prionechinus sculptus A. Agassiz and Clark, 1907. Bull. M. C. Z., L, p. 243.

Plates 93, figs. 24-26; 101, figs. 4-6.

The large series of specimens shows so little diversity in form that the measure-
ments of the type-specimen will be sufficient to show the proportions of the
species. The test is 10 mm. in diameter and 5 mm. high, while the actinostome
is about 4.5 and the 'abactinal system a trifle over 4 mm. The primary spines
are 3-5 mm. long at the ambitus but are decidedly shorter abactinally. There
are 11 interambulacral and 12 ambulacral plates in each column. The test is

PRIONECHINUS SCULPTUS. 307

well arched and only slightly flattened abactinally. The sculpturing is confined
to the oculo-genital ring and there is great diversity in the degree to which it is
developed there. The whole test is well covered with granules and tubercles
and the median interambulacral suture is sufficiently depressed to be rather
conspicuous. On the ocular and genital plates there are not only numerous
small tubercles but irregular or radiating furrows and low ridges. In some speci-
mens these are almost wholly wanting while in others, they are so conspicuous
(under a lens) as to give the whole abactinal system a highly ornamented ap-
pearance. The genital plates (PI. 101, fig. 6) are moderate and subequal, the
madreporite, with its minute group of a very few water-pores, being distinguish-
able only with difficulty. The genital pore is near the centre of each plate. The
ocular plates are rather large, but are all broadly exsert. The periproct is
covered by a considerable number of small, rounded plates but among them the
suranal is easily distinguishable. The ambulacra are rather narrow, little
more than half as wide at ambitus as the interambulacra. The poriferous areas
are narrow but rather conspicuous, as the pores are large; the pore-pairs are
arranged in nearly vertical arcs of three. The actinostomal membrane (PI. 101,
fig. 5) is very thin and distal to the buccal plates is perfectly naked. Within
their circle, there are small plates, sometimes abundant enough to cover that
part of the membrane very fully but in other cases imperfect and scattered.
The buccal plates themselves are of moderate size and all carry normal tube-
feet ; the two plates of each pair are very close together, but the pairs are well
separated from each other. The gill-cuts are insignificant. The spines are
rather thick and blunt or bluntly pointed. The primaries have about 10 longi-
tudinal striations.

The pedicellarise are abundant and not specially characteristic, as they are
very similar to those of Chuni. While the globiferous (PI. 93, fig. 24) and ophi-
cephalous are common enough, only two tridentates and no triphyllous were
found. The tridentate have very narrow, compressed valves (PI. 93, fig. 25)
about .60 mm. long, meeting only at the tip where they are somewhat expanded
and have minutely serrate margins.

The color of the test and smaller spines is dull purplish red, often very light
and becoming nearly white in some specimens. The primaries are white, but
the longitudinal striations are purplish. As a result of this coloration, the
general aspect of the specimens is rather diverse, the exact shade ranging from
distinct, though dull, purplish red to almost white.

This is a well-marked species and it is not likely to be confused with any

308 HAWAIIAN AND OTHER PACIFIC ECHINI.

except depressus; the position of the genital pores distinguishes it at once from
that species. It appears to be quite common among the Hawaiian Islands on
sandy bottoms, in moderately deep water. It apparently does not reach a
large size as our specimens range only from 2 to 10 mm. in diameter.

The "Albatross" collected sculptus at the following stations : —

Station 3818. Off Diamond Head, Oahu, Hawaiian Islands. Bott. temp.
44.3°. 293-295 fathoms. Fne. co. s., bl. sp.

Station 4028. Off Ukula Point, Kauai, H. I. Bott. temp. 40°. 444-478
fathoms. Gy. s., glob.

Station 4039. Off Kawaihae Light, Hawaii, H. I. Bott. temp. 38.7°.
670-697 fathoms. Gy. m., for.

Station 4083. Off Puniawa Point, Maui, H. I. Bott. temp. ? 238-253
fathoms. Gy. s.

Station 4086. Off Puniawa Point, Maui, H. I. Bott. temp. 44.6°. 283-
308 fathoms. S., sh.

Station 4087. Off Mokuhooniki Islet, Pailolo Channel, H. I. Bott. temp.
43.6°. 306-308 fathoms. Fne. gy. s.

Station 4088. Off Mokuhooniki Islet, Pailolo Channel, H. I. Bott. temp.
43.8°. 297-306 fathoms. Fne. gy. s.

Station 4115. Off Kahuku Point, Oahu, H. I. Bott. temp. 55.1°. 195-
241 fathoms. Co. s., for.

Bathymetrical range, 195-697 fathoms. Extremes of temperature, 55.1°-
38.7°.

Sixty-seven specimens.

Prionechinus Agassizii Wood-Mas, and Ale.

Prionechinus Agassizii Wood-Mason and Alcock, 1891. Ann. Mag. Nat. Hist., (6), VIII, p. 441.

The status of this species is uncertain. The original specimens were un-
usually large (13.8 mm. h. d.) for representatives of this genus and the characters
on which special stress was laid in the original description are probably due to
age; in all our largest specimens of Prionechinus the median interambulacral
suture tends to be sunken and the poriferous area tends to assume a zigzag
appearance near the actinostome. The specimens referred to this species,
from the "Albatross" collection are all small, 5-7 mm. in diameter, and though
they show much diversity in color, they agree in the position of the genital
pores and in the characters of the actinostome. One is pure white, another has

PRIONECHINUS DEPRESSUS. 309

the test pale brown and the base of the spines tinged with olive, while the third
has the tubercles and the basal half of all the larger spines pale red.

These specimens were taken at the following stations: —

Station 4965. Between Kobe and Yokohama, Japan; 33° 35' 20" N.,
135° 10' 50" E. Bott. temp. 49.4°. 191 fathoms. Dk. gn.-gy. s., sh.

Station 4967. Between Kobe and Yokohama, Japan; 33° 25' 10" N.,
135° 37' 20" E. Bott. temp. 45.9°. 244-253 fathoms. Br. m., s., for.

Station 5086. Sagami Bay, Japan; 35° 8' 15" N., 139° 20' E. Bott. temp.
43.7°. 292 fathoms. Gn. m., crs. bk. s.

Bathyrnetrical range, 191-292 fathoms. Extremes of temperature, 49.4°-
43.7°.

Three specimens.

Prionechinus depressus A. Ag. and CI.

Prionechinus depressus A. Agassiz and Clark, 1907. Bull. M. C. Z., L, p. 244.

Plate 101, figs. 1-3.

The largest specimen of this species is 10 mm. in diameter but the height of
the test is hardly 4 mm. The abactinal system and actinostome are about
equal to each other, measuring 5 mm. across, which is practically the length of
the primary spines at the ambitus. There are 11 coronal plates in the inter-
ambulacral as well as in the ambulacral columns. In another specimen, 7.5 mm.
in diameter, the test is 3.5 mm. high, while in a still smaller one, 6.5 mm. h. d.
the vertical diameter is fully half the horizontal. The depressed test therefore
seems to be acquired with age. There is practically no sculpturing on the test
and very little, if any, on the genital plates (PI. 101, fig. 3). The genital pores
are at the extreme distal tips of the genital plates. The pore is placed in a notch
which is continuous with the sunken, median interambulacral suture, which
in this species is very marked clear to the ambitus, at least in specimens over
7 mm. h. d. The ambulacra are rather wide, decidedly more than half as wide
as the interambulacrum at the ambitus. The pores are larger and the porif-
erous areas wider than in specimens of sculptus of the same size.

In all other particulars (tuberculation, character of actinostome (PI. 101,
fig. 2), spines, pedicellarise, and color) this species is so similar to sculptus as to
be practically indistinguishable. The difference in the genital pores is very
noticeable and in large specimens the shape of the test affords an obvious means
of distinction. An examination of the reproductive organs fails to show that the

310 HAWAIIAN AND OTHER PACIFIC ECHINI.

difference in the genital pores is a sexual one. So far as could be determined
from alcoholic material, which is not very well adapted to such an examination,
there is no difference between the reproductive organs of the two species.

The "Albatross" took depressus at the following five stations, at all of which
sculptus also occurred; a fact which certainly adds weight to the view that the
two forms are simply sexes of the same species. If the two species really are
distinct, their occurrence at the same stations is very interesting as tending to
show that Echini, like ophiurans,1 do not follow "Jordan's law."

Station 3818. Off Diamond Head, Oahu, Hawaiian Islands, Bott. temp.
44.3°. 293-295 fathoms. Fne. co. s., bl. sp.

Station 4028. Off Ukula Point, Kauai, H. I. Bott. temp. 40°. 444-478
fathoms. Gy. s., glob.

Station 4083. Off Kawaihae Light, Hawaii, H. I. Bott. temp. 38.7°.
670-697 fathoms. Gy. m., for.

Station 4086. Off Puniawa Point, Maui, H. I. Bott. temp. 44.6°. 283-
308 fathoms. S., sh.

Station 4088. Off Mokuhooniki Islet, Pailolo Channel, H. I. Bott. temp.
43.8°. 297-306 fathoms. Fne. gy. s.

Bathymetrical range, 283-697 fathoms. Extremes of temperature, 44.6°-
38.7°.

Forty-five specimens.

OPECHINTJS.

Desor, 1855. Syn. Ech. Foss., p. 107.
Type-species, Temnopkurus costatus d'Archiac et Haime, 1853. Anim. Foss. de 1'Inde, p. 204.

There seems to be no escape from the fact, unfortunate as it may be, that the
type of this genus was fixed by Pomel. Since he removed all of Desor's six
species except costatus from Opechinus, and since he was clearly acting within
his rights as first reviser, it is hard to see how any species except costatus can be
the genotype. It is by no means certain that the two recent species from the
East Indian region, referred to Opechinus by Mortensen, are really congeneric
with costatus, but in order that the difficulties of the problem be not increased
it is proposed to let the genus stand as Mortensen left it, so far as the recent
species are concerned. The two species may be distinguished as follows: —

Each horizontal interambulacral suture in mid-zone, with 4 large shallow depressions; a

distinct suranal plate distinguishable on periproct variabilis.

Each horizontal interambulacral suture in mid-zone, with 6 large, shallow depressions;

no distinct suranal plate spectabilis.

'See Clark, H. L., 1911, Bull. 75, U. S. N. M., p. 22-23.

TEMNOPLEURUS. 311

Opechinus variabilis Mrtsn.

Pleurechinus variabilis Doderlein, 1885. Arch. f. Naturg., Jahrg. LI, 1, p. 93.
Opechinus variabilis Mortensen, 1904. Dan. Exp. Siam: Ech., p. 92, 94.

Examination of the test of this species seems to confirm Mortensen's view
that its relationship is with Temnechinus rather than with Temnopleurus, for
there is no evidence of dowelling between the plates and they are rather thinner
than would be expected from the outer surface. Our specimens are small
(8-11 mm. in diameter) and the pits at the angles of the plates are often very
small or wanting while the large depressions on the sutures are very conspicuous.

The "Albatross" took this species at the following stations: —

Station 4893. Southwest of Goto Islands, Japan; 32° 32' N., 128° 32' 50" E.
Bott. temp. 55.9°. 95-106 fathoms. Gy. s., brk. sh., p.

Station 4894. Southwest of Goto Islands, Japan; 32° 33' N., 128° 32' 10" E.
Bott. temp. ? 95 fathoms. Gy. s., brk. sh., p.

Station 5068. Suruga Gulf, Japan; 35° 2' 25" N., 138° 46' 55" E. Bott.
temp. 63°. 77-131 fathoms. Bk. s., brk. sh.

Bathymetrical range, 77-131 fathoms. Extremes of temperature, 63°-
55.9°.

Three specimens.

TEMNOPLEURUS.

L. Agassiz, 1841. Int. Mon. Scut., p. 7.
Type-species, Cidaris toreumalica Leske, 1778. Add. ad Klein, p. 155.

This genus has been so recently and so fully revised by Mortensen that
only agreement with his general conclusions need be expressed. There seems
to be no reason to doubt that T. Reynaudi is really synonymous with toreumaticus
and therefore the species long known as Reynaudi must take Gray's name
Reevesii. The other species, Hardwickii, is near toreumaticus but with the spines
on, the two are very easily distinguished. All three species occur in Japanese
waters, but toreumaticus extends westward to Arabia and southward probably
to Queensland, while Reevesii ranges westward to Ceylon, and Hardwickii so
far as known, is peculiar to Japan, the record from Unalaska being exceedingly
dubious.

The three species 1 may be distinguished from each other regardless of the
differences in the sculpturing of the test, which shows great individual diversity,
as follows: —

1 For Perezi, see below under toreumaticus.

312 HAWAIIAN AND OTHER PACIFIC ECHINI.

Primary spines very dark (almost black) at base, not banded; poriferous area narrow,

pore-pairs small in a vertical series close to margin of ambulacrum Hardwickii.

Primary spines not very dark at base; poriferous area not very narrow; pore-pairs
medium or large, in an interrupted vertical series, not close to margin of ambu-
lacrum, one or more miliary tubercles being present between some pore-pairs
and the margin.

Primary spines, at least actinally, banded ; all ocular plates excluded from periproct toreumaticus.
Primary spines light colored, not banded; in adults (over 20 mm. h. d.) ocular

I is usually insert Reevesii.

Temnopleurus Hardwickii A. Ag.

Toreumatica Hardwickii Gray, 1855. Proc. Zool. Soc. London, p. 39.
Temnopleurus Hardwickii A. Agassiz, 1872. Rev. Ech., pt. 1, p. 166.

The specimens taken by the "Albatross" range in diameter from 13 to 31 mm.,
but show very little diversity in coloration. There are some differences in the
size, depth, and form of the pits but they do not seem to be very remarkable.
It is a little odd that the "Albatross" did not meet with this species during her
extended third visit to Japan, but only on her second voyage and then only at
the following adjoining stations: —

Station 3723. Off Yokkaichi Light, Honshu Island, Japan. 13-16 fathoms.
M., s., p., sh.

Station 3725. Off Noma Saki, Honshu Island, Japan. 13 fathoms. S.,
sh., g.

Nine specimens.

Temnopleurus toreumaticus Agassiz.

Cidaris toreumatica Leske, 1778. Add. ad Klein, p. 155.
Temnopleurus toreumaticus L. Agassiz, 1841. Int. Mon. Scut., p. 7.

A single specimen, 42 mm. in diameter, is the only representative of this
well-known species in the "Albatross" collections. It is dark colored and the
banding of the primaries is only noticeable actinally. This is in striking contrast
to the color of a fine series of specimens from the Persian Gulf, which the M. C. Z.
received from Capt. F. W. Townsend in 1895. In these the ground color is very
light, usually a pale cream-color, in some specimens with a more or less marked
olive-green cast, and the primary spines are very conspicuously banded with
brownish or purplish red. These specimens appear to be identical with those
from the Red Sea (Arabian coast) described by Krehler (1906, Bull. Paris Mus.,
XI, p. 460) as a new species, T. Perezi. Kcehler fails to mention a single charac-
ter by which Perezi is to be distinguished from toreumaticus, except perhaps

TEMNOPLEURUS REEVESII. 313

the height of the test. This is of little importance however as toreumaticus is
very variable in that particular. The coloration seems of more importance,
but the material at hand shows that the light and dark extremes intergrade.
Perhaps the very light western specimens may be recognized as a subspecies
or variety, Perezi. Aside from the remarkable difference in color, the Persian
and Arabian specimens do not differ constantly in any particular from specimens
from the East Indies, China, and Japan.

The specimen in the "Albatross" collection was taken at Nanao Beach,
Japan.

Temnopleurus Reevesii Mrtsn.

Toreumatica Reevesii Gray, 1855. Proc. Zool. Soc. London, p. 39.
Temnopleurus Reynaudi A. Agassiz, 1872. Rev. Ech., pt. 1, p. 166.
Temnopleurus Reevesii Mortensen, 1904. Dan. Exp. Siam: Ech., p. 62.

V.

Although a large series of this species is at hand, they show so little diversity,
that little need be said about them. Since this species is one of the very few
Temnopleuridse in which an ocular is insert, it will be of interest to record the
conditions of the oculo-genital ring as shown by the fifty specimens, ranging from
7 to 43 mm. h. d., which have been examined. Of the fifty, eleven have ocular I
insert, and eight others have it almost in. This would seem like a small per-
centage on which to base the statement that it is a species character to have
ocular I insert, but when the specimens are grouped according to size, interesting
information in regard to this character is obtained. There are eighteen speci-
mens less than 14 mm. in diameter; i. e., since our largest specimen is 43 mm.
h. d., less than one third grown. Of these eighteen, only three have ocular I
insert, or say 17%; one other has it almost in, so it may be noted that 78%
show no indication of the character. There are twenty-seven specimens from a
third to a half grown (14-21 mm. inclusive) and of these, five have ocular I
insert, or say 18? %; but six others have it almost in, so that less than 60% show
no indication of the character. There are only five specimens more than half
grown; of these three (60%) have ocular I insert and one other has it almost in,
and thus only 20% show no indication of the character. The largest specimen
(43 mm. h. d.) has ocular I very broadly insert. Is it not fair from these data
to assume that having ocular I insert is a species character, assumed late in
development and not usually acquired until the individual is more than half
grown?

This species was taken by the "Albatross" at the following stations: —

314 HAWAIIAN AND OTHER PACIFIC ECHINI.

Station 3717. Off Ose Zaki, Honshu Island, Japan. 63-100 fathoms.
Vol. s., sh., r.

Station 4815. Between Hakodate and Sado Island, Japan; 38° 16' N.,
138° 52' E. Bott. temp. 51°. 70 fathoms. Dk. gn. s.

Station 4832. Between Nanao and Tsuruga, Hondo, Japan; 36° 14' 30" N.,
135° 56' 30" E. Bott. temp. 53.2°. 76-79 fathoms. Dk. gy. s.

Station 4893. Southwest of Goto Islands, Japan; 32° 32' N., 128° 32' 50" E.
Bott. temp. 55.9°. 95-106 fathoms. Gy. s., brk. sh., p.

Station 4894. Southwest of Goto Islands, Japan; 32° 33' N., 128° 32' 10" E.
Bott. temp. ? 95 fathoms. Gn. s., brk. sh., p.

Station 4895. Southwest of Goto Islands, Japan; 32° 33' 10" N., 128° 32'
10" E. Bott. temp. ? 95 fathoms. Gn. s., brk. sh., p.

Station 4902. Southwest of Goto Islands, Japan; 32° 30' 50" N., 128° 34'
40" E. Bott. temp. 52.9°. 139 fathoms. Gy. s., brk. sh.

Station 4904. Southwest of Goto Islands, Japan; 32° 31' 20" N., 128° 32'
40" E. Bott. temp. ? 107 fathoms. Fne. gy. s., brk. sh.

Station 4931. In Colnett Strait, Japan; 30° 12' N., 130° 43' 40" E. Bott.
temp. 75.4°. 83 fathoms. Brk. sh., p., co.

Station 4933. Off Kagoshima Gulf, Japan; 30° 59' N., 130° 29' 50" E.
Bott. temp. 56°. 152 fathoms. Rky.

Station 5074. In Suruga Gulf, Japan; 34° 40' 45" N., 138° 18' 30" E.
Bott. temp. 74.9°. 47 fathoms. Gy. m.

Station 5095. Off Gulf of Tokyo, Japan; 35° 5' 34" N., 139° 38' 36" E.
Bott. temp. 57.8°. 58 fathoms.

Bathymetrical range, 47-152 fathoms. Extremes of temperature, 75.4°-51°.

Thirty-two specimens.

SALMACIS.

L. Agassiz, 1841. Preface to Valentin's Anat. Genre Echinus, p. VIII.
Type-species, Salmacis bicolor L. Agassiz, 1841, 1. c.

The specific limits within this genus are at present imperfectly drawn,
although the work of Doderlein and Mortensen has greatly increased our knowl-
edge. The great variability of some species, in test sculpture and form, and also
in color, coupled with the lack of large series of specimens, has made it very
difficult to decide just what species really are valid. Fortunately the color of
the type-species, bicolor, appears to be both constant and distinctive, and as it
is particularly mentioned by Agassiz, it makes his diagnosis of both species and

SALMACIS. 315

genus recognizable and valid. A closely related form has long been distinguished
as rarispina, but the examination of our M. C. Z. material confirms Mortensen's
statement that there is no constant difference between the two forms. Indeed
the difference which his specimens showed in the number of ambulacral and
interambulacral plates and which led him to retain rarispina as a variety of
bicolor is not shown by our specimens, and it is not practicable to distinguish
the variety. So far as can be determined from the available material, and the
published notes on the different species, the form of the test is strikingly variable
in bicolor, virgulata, Alexandri, and perhaps also in sphceroides, the vertical diame-
ter ranging from little more than one half to nearly three fourths of the horizontal.
We note also that most specimens fall into one of two groups, the one with the
vertical diameter from .55 to .60 h. d. and the other with it about .70 h. cl.
Ramsay, who appears to be the only writer to have examined large series of
living Salmacis, speaks emphatically of the diversity shown by S. Alexandri
in the form of the test (1885, Cat. Ech. Australian Mus., p. 48). The question
of the relation of A lexandri to virgulata appears to be a debatable one; the former
is at least a subspecies characteristic of the Australian region, and as a con-
necting series of specimens is lacking, those at hand being easily recognized, it
has seemed well to let Alexandri stand as a valid species. These two species
(virgulata and Alexandri) are distinguished from the other members of the genus
by the absence of bands on the primary spines, but this is of course, not so
important a character as those which distinguish Dussumieri. In this species,
the test appears to be always flattened, so much so that the vertical diameter
is less than one half the horizontal, and ocular I is insert or very nearly so.
Moreover the ambulacra are highly modified, so that there is a primary tubercle
close to the poriferous zone, only on every other ambulacral plate, excepting
only the oldest (near actinostome) and youngest (near ocular plate). Morten-
sen (1904, Dan. Exp. Siam: Ech., p. 73) refers to "a very dark colored form"
of Dussumieri in the collection of the British Museum from Tuticorin (southern
India) and he adds that Bell's S. sulcata from Zanzibar is similar. He says
further that "it can only be regarded as a color variety." A fine specimen of
this form from Zanzibar (PL 111, figs. 4-6) is in the collection of the M. C. Z.
and its color characters are so strikingly different from Dussumieri that it must
rank as a new species, for which we suggest the name erythracis (epvOpos, red +
OLKLS, a pointed instrument) in reference to the more or less conspicuously ver-
milion-red bases of the secondary and miliary spines. There is little question
about the status of the form, which Loven has showed Linne designated as

316 HAWAIIAN AND OTHER PACIFIC ECHINI.

sphtxroides, but there is some doubt regarding Belli Doderlein. We have a
beautiful specimen of the latter, or one at least from "Challenger" Station 188,
and Mortensen (1904, Dan. Exp. Siam: Ech., p. 68) asserts that the Salmacis
from that station is Belli. Moreover it answers well to the published descrip-
tions, save for the following details: — the globiferous pedicellariae do not "agree
exactly with" those of bicolor as Mortensen says but, are easily distinguishable
(PI. 93, figs. 33-36) ; the coronal (interambulacral) plates are not essentially
different in proportions from those of sphoeroides, for they are scarcely five times
as wide as high, while in one of our sphceroides they are more than five times as
wide as high, instead of less, as usual. Whether the differences in color, tri-
dentate pedicellarise, actinal primaries, and gill-cuts really have any significance
remains to be determined. So far as can be determined, the following species
may be accepted and distinguished from each other thus : —

Vertical diameter of test exceeds one half horizontal; all oculars exsert; each ambula-

cral plate with a primary tubercle close to the poriferous area.
Primary spines not banded, greenish or light colored at base, becoming reddish or

purple distally (or for most of their length) but sometimes with light tips.
Coronal plates separated from each other by distinct sutures, the edges of

which may be slightly bevelled virgulata.

Coronal plates separated from each other by deep horizontal furrows, the sides

of which are more or less vertical Alexandri.

Primary spines banded, usually conspicuously so.

Primary spines more or less bright red, especially abactinally, at least at base;

small spines red bicolor.

Primary spines green at base, at least abactinally.

Primaries, distal to green base, not rose-red; valves of tridentate pedicel-
lariae very narrow; actinal primaries not specially widened at tip;

gill-cuts deep sphceroides.

Primaries, distal to green base, rose-red; valves of tridentate pedicel-
laria;, broad; actinal primaries noticeably widened at tip; gill-cuts

insignificant Belli.

Vertical diameter of test less than one half horizontal; ocular I insert or nearly so; only
every other ambulacral plate (at least in mid-zone) with a primary tubercle close
to the poriferous area.

Primary spines white or greenish with 2 or 3 broad bands of purple or purplish red,
or the dark color may predominate, the white disappearing and the bands
becoming indistinct; secondary and miliary spines not at all vermilion-red . . Dussumieri.
Primary spines green or greenish; actinal ones with 2 or 3 indefinite broad bands of

purplish brown; secondaries and miliaries vermilion-red at base erythracis.

TEMNOTREMA. 317

TEMNOTREMA.

A. Agassiz, 1863. Proc. Acad. Nat. Sci. Philadelphia, p. 358.

Type-species, Temnotrema sculpta A. Agassiz, 1863, 1. c.

(= Pleurechinus A. Agassiz, 1872, and later writers; non Pleurechinus L. Agassiz, 1841, Int. Mon.

Scut., p. 7.)

The genus Pleurechinus was established by L. Agassiz in 1841 (1. c.) and he
definitely selected Cidaris bothryoides Leske as the type-species (1841, Int.
Valentin's Anat. Genre Echinus, p. viii). In 1846 however he abandoned the
generic name and placed bothryoides in Temnopleurus. Although he says his
Temnopleurus bothryoides is equivalent to Leske's Cidaris bothryoides, it is clear
that such is not the case for his diagnosis does not apply at all to the species
figured by Klein and described by Leske. Fortunately Agassiz labelled as
bothryoides a fine bare test in the Michelin collection in Paris, to which his
diagnosis of Temnopleurus bothryoides does apply, and this was taken by A.
Agassiz in the "Revision" and by later writers as the type-specimen of Pleurechi-
nus bothryoides. In the "Revision" however, it is distinctly stated that this
specimen is entirely different from Cidaris bothryoides Leske, which is perhaps
a Microcyphus. Clearly then the Paris specimen cannot be the type of Pleur-

^

echinus bothryoides, which according to Agassiz in 1841, was nothing more nor
less than Leske's species. What names then should be applied to the Paris
specimen and to the genus to which it belongs? Since it is not labelled "Cidaris
bothryoides Leske," there seems to be no reason why its name may not stand as
Temnopleurus bothryoides Agassiz, but of course, it cannot be the type of Pleur-
echinus which is the entirely different Cidaris bothryoides Leske. Since this
latter is with little doubt, quite unrecognizable (for even if it is a Microcyphus
the species is not determinable with certainty), the name Pleurechinus must be
abandoned, and the group which has borne it for so many years must take
another. Fortunately such a name already exists, having been proposed in
1863 by A. Agassiz, for a small sea-urchin from Japan, Temnotrema sculpta.
Later the specimen was believed to be a young Temnopleurus Hardwickii and
in the "Revision" Temnotrema sculpta is placed in the synonymy of the Temno-
pleurus. The type-specimen of Temnotrema sculpta (PI. 112, figs. /, 2) is in
the M. C. Z. collection and comparison with other Japanese material shows that
it is identical with the species described by Mortensen (1904, Dan. Exp. Siam:
Ech., p. 84) as Pleurechinus variegatus. As a specimen of variegatus was re-
ceived from Dr. Mortensen himself, there can be no doubt of the identification.

318 HAWAIIAN AND OTHER PACIFIC ECHINI.

The difference in the descriptions of the color of the spines, is probably of degree
rather than of kind, but in the type of Temnotrema the spines are now missing
or broken, so that it is impossible to speak positively on this point. Since
sculpta (= Mortensen's variegatus) seems to be undoubtedly congeneric with
Agassiz's bothryoides, it is clear that the generic name Temnotrema must replace
Pleurechinus for this group of Temnopleuridse, and sculpta becomes the type in
place of bothryoides.

As the genus has been so recently (1904) and so fully revised by Mortensen,
there is no need for a discussion of the species. Only four of the seven species
he tabulates are at hand, but there appears to be no reason for questioning any
of the others and his list with the addition of the new species found by the
"Albatross" among the Hawaiian Islands, may be accepted. All of the species
are found in the Indo-Pacific region, especially among the East Indian Islands;
perhaps sculpta is confined to Japanese and Formosan waters, as haivaiiensis
appears to be to the Hawaiian region. A specimen of Doderleini in the M. C. Z.
collection from Fiji may be recorded here; this however is not surprising as it
was already known from Samoa.

As Mortensen has made considerable use of the pedicellarise in his table,
and as their characters seem of very little importance and not altogether reliable,
the following table, showing additional characters by which the eight species
may be distinguished, is offered.

Buccal plates 10; anal plates not very numerous, a suranal usually evident.

Pits large, the distance between the two of same horizontal interambulacral suture

less than the length of one.

Anus central or nearly so; periproct without a conspicuous suranal; test high
(v. d. may equal .75 h. d.) uniformly dark; primary spines light with 2-4 red
bands; valves of globiferous pedicellarue with a lateral tooth on each side

near tip bothry aides.

Anus more or less excen trie; suranal distinct; test not uniformly dark; valves

of globiferous pedicellariic without lateral teeth.
Test not very high, v. d. = .50-.60 h. d.; suranal plate very large, covering

half or more of periproct.

Spines, at least secondaries, thorny, not swollen at tip; poriferous area
\ to i as wide as interporiferous; coronal plates conspicuously

sculptured around primary tubercles scillce.

Spines smooth; primaries often swollen at tip; poriferous area J to f
as wide, as interporiferous; coronal plates little sculptured around

primary tubercles siamensis

Test rather high, v. d. = 60-.70 h. d.; suranal plate moderate not covering

half the periproct.

Genital plates marked by a transverse line which appears to divide the
distal from the proximal part; tubercles on coronal plates in
horizontal rows; no green in coloration . . . . : maculata.

TEMNOTREMA HAWAIIENSIS. 319

Genital plates not marked by a transverse line; tubercles on coronal
plates not in horizontal rows; more or less green on test, or if

green is lacking, test is bright red hawaiiensis.

Pits small, the distance between the two of same horizontal interambulacral suture

greater (often much greater) than the length of one.

Ambulacral plates equal interainbulacral in height and number; test variegated
gray and whitish; spines reddish at base, often with a small dark band

distally; genital plates marked with a transverse line sculpla.

Ambulacral plates lower and somewhat more numerous than interambulacral;

test red; spines red with white bands; no transverse line on genital plates . . ruber.
Buccal plates only 5; anal plates very numerous, subequal, with no distinct suranal . . Doderleini.

Temnotrema hawaiiensis, comb. nov.

Pleurechinus hawaiiensis A. Agassiz and Clark, 1907. Bull M. C. Z., L, p. 244.

Plate 99, figs. 1-3.

The specimens at hand of this pretty species range in horizontal diameter
from 5 to 9 mm. The vertical diameter ranges from about two thirds to three
fourths of the horizontal. The test is therefore quite high and it is well arched.
The actinostome of the largest specimen is 4 mm. across, while the abactinal
system is only 3 mm. The longest primary spines are only a little over 3 mm.
long. There are 12 interambulacral plates in each column and the number of
ambulacrals appears to be the same. In the mid-zone, the interambulacral
plates carry at the centre a small primary tubercle, on each side of which is a
distinct secondary, and on the upper half of the plate, in an irregular, curved
series are 4-6 smaller secondaries. The pits are very large and occupy most of
the lower half of each plate. Of course, the tubercles are fewer and the pits
smaller as the ocular plate or the peristome is approached. In the ambulacra
the number of tubercles on each plate is less and the pits are smaller. The pit
at the outer corner of each plate is insignificant or wanting. The genital plates
(PI. 99, fig. 3) are of moderate size, pentagonal, and a little elevated; there is a
tubercle (rarely two) on the proximal margin, and the rather large genital pore
is just distal to the centre of the plate. The madreporite is very inconspicuous
as the pores are comparatively few. The oculars are small and completely
exsert ; each one carries a couple of very small tubercles. At the proximal end
of each ocular is a more or less conspicuous pit. The periproct is covered by a
small number of plates (4-10) among which the suranal is usually easily dis-
tinguishable. The anus is approximately central. The actinostomal membrane
(PI. 99, fig. 2} is very thin and bare. The buccal plates are very small, and

320 HAWAIIAN AND OTHER PACIFIC ECHINI.

although the two of a pair are placed side by side, the pairs themselves are well
separated from each other. Each plate carries a tube-foot. The spines are
slender, rather blunt, and perfectly smooth. The small ones may end with a
distinct central point, and some at least are slightly swollen at the tip.

The pedicellarise are very much like those of siamensis. The globiferous are
fairly common; the valves measure about .16 mm. in length. The ophicephalous
are abundant and quite variable in size, the valves ranging from .13 to .19 mm.
besides the loop. In the larger ones, the valves are relatively longer and have a
more sinuate margin than Mortensen figures for siamensis. No tridentate
pedicellarise were found. The triphyllous seem to be rare and the valves are
only .10 mm. in length. Neither spicules nor sphseridia were noted.

The color of the test is prevailingly green, with the abactinal interambulacra
lighter and often pure white in striking contrast. The primary spines are
whitish with more or less red. The tendency towards a bright red coloration
is noticeable and two specimens are almost uniformly bright red, test as well as
spines. Around the actinostome the test often becomes whitish, while abactinally
it is frequently marked with purplish brown. While the color is thus very
variable, there is no tendencjr to approach that of siamensis, except that each
species has a bright red variety.

Although this species appears to be very near siamensis, comparison of
specimens of the same size reveals some important differences. The primary
spines are noticeably longer and more slender and are not banded, in hawaiiensis,
while the test is relatively higher and the suranal is smaller. We do not find
any trace of a membrane surrounding the pits, as Mortensen figures for siamensis.
The characters by which hawaiiensis is distinguished from maculata are given
in the table above. It seems probable that with a specimen of maculata at hand
for comparison, other and weightier differences might be found, for with only
the published description and figures of maculata, some important points may
have been overlooked.

The "Albatross" took hawaiiensis at the following stations: —

Station 3823. Off Lae-o Ka Laau Light, Molokai, Hawaiian Islands. Bott.
temp. 69°. 78-222 fathoms. Fne. s., p.

Station 3847. Off Lae-o Ka Laau Light, Molokai, H. I. Bott. temp. ?
23-24 fathoms. S., st.

Station 3871. Off Mokuhooniki Islet, Auau Channel, H. I. Bott. temp.
74.6°? 13-43 fathoms. Fne. wh. s.

Station 3872. Off Mokuhooniki Islet, Auau Channel, H. I. Bott. temp.
74.6°. 32-43 fathoms. Yl. s., p., co.

TEMNOTREMA SCULPTA. 321

Station 3876. Off Lahaina Light, Maui, H. I. Bott. temp. 74°. 28-43
fathoms. S., g.

Station 3962. Off Laysan Island, H. I. Bott. temp. ? 16 fathoms. Wh.
s., co.

Station 3978. Off Modu Manu (Bird Island), H. I. Bott. temp. ? 32-
46 fathoms. Co. s., for., r.

Station 4148. Off Modu Manu (Bird Island), H. I. Bott. temp. 77.9°.
26-33 fathoms. Co. s., for.

Station 4150. Off Modu Manu (Bird Island), H. I. Bott. temp. 74°.
71-160 fathoms. Co.

Bathymetrical range, 13-222 fathoms. Extremes of temperature, 77.9°-
69°.

Sixteen specimens.

Temnotrema sculpta A. Ag.

Temnotrema sculpta A. Agassiz, 1863. Proc. Acad. Nat. Sci. Philadelphia, p. 358.
Pleurechinus variegatus Mortensen, 1904. Dan. Exp. Siam: Ech., p. 84; PI. 1, figs. 5, 6, S, 19;
PI. 2, fig. 6.

Plate 112, figs. 1, 2.

The "Albatross" specimens range from 6 to 10 mm. in diameter, and are
thus smaller than the type of sculpta which is 11 mm., but the latter is smaller
than Mortensen's specimens which ranged from 11.5 to 17mm. The "Albatross"
specimens have scarcely a trace of red on the primaries but are not otherwise
peculiar. Examination of the periproct in the specimens examined shows that
the description in the original diagnosis of Temnotrema is misleading. The
suranal plate is prominent and there may be two, or even three, other large
plates, but there is no close resemblance to the quartet of equal plates seen in
Arbacia.

This species was taken by the "Albatross" at the following stations: —

Station 4893. Southwest of Goto Islands, Japan; 32° 32' N., 128° 32' 50" E.
Bott. temp. 55.9°. 95-106 fathoms. Gy. s., brk. sh., p.

Station 4895. Southwest of Goto Islands, Japan; 32° 33' 10" N., 128° 32'
10" E. Bott. temp. ? 95 fathoms. Gn. s., brk. sh., p.

Station 5095. Off Gulf of Tokyo, Japan; 35° 5' 34" N., 139° 38' 36" E.
Bott. temp. 57.8°. 58 fathoms. Fne. bk. s., brk. sh.

Bathymetrical range, 58-106 fathoms. Extremes of temperature, 57.8°-
55.9°.

Three specimens.

322 HAWAIIAN AND OTHER PACIFIC ECHINI.

MESPILIA.

Agassiz and Desor, 1846. Ann. Sci. Nat., (3), VI, p. 357.
Type-species Echinus globulits Linne, 1758. Syst. Nat., ed. 10, p. 664.

Little need be said of this well-known genus, which still contains only the
species on which it was based. It is true Yoshiwara (1897, Ann. Zool. Jap., I,
p. 58) has described a second species but his description applies perfectly well
to globulus except that he says the spines are "longitudinally striated with orange
stripes and tipped with white." As he doubtless means transversely banded with
orange, his name becomes a simple synonym of globulus.1 Some Samoan speci-
mens have the bare spaces on the test densely covered with globiferous pedi-
cellarife but this condition seems to be unusual. In the M. C. Z. collection
are a number of specimens from Samoa, Fiji, the Caroline Islands, and the
Philippines, which differ strikingly from the ordinary globulus in the complete
absence of red from both test and spines. The primary spines are green banded
with blackish or with blackish and whitish. The largest specimen is 47 mm.
h. d. Though this form is not entitled to specific rank it is suggested that it be
known as Mespilia globulus variety pellocrica (TreAAos = dark-colored + KpiKos =
a circle). It is certainly not a geographical form, for typical globulus occur in
the same localities, with specimens of var. pellocrica.

MICROCYPHUS.

Agassiz and Desor, 1846. Ann. Sci. Nat., (3), VI, p. 358.
Type-species, Microcyphus maculalus Agassiz and Desor, 1846, 1. c.

Since Agassiz's references to this genus in 1841 are absolute nomina nuda
for both genus and species, the names cannot date back earlier than 1846. No
type has been definitely given the genus since it was first published, but there
appears to be no objection to adopting maculatus, the first species mentioned
by Agassiz. Mortensen (1904, Dan. Exp. Siam: Ech.) has added two species
to the genus as known to Agassiz, but no other additions have been made.2

1 Examination of Yoshiwara's type-specimen shows this conclusion to be correct.

2 Unfortunately Mortensen selected elegans as the name of one of his species, overlooking the fact
that a Microcyphus elegans (= Temnopleurus hardwickii juv.) was described by A. Agassiz in 1863.
In place of elegans, I would propose for Dr. Mortensen's species, the Greek equivalent (
compsus.

MICROCYPHUS. 323

These two species are from southeastern Australia, the home of M. zigzag, while
maculatus ranges from Mauritius, where it appears to be common, to Japan and
Samoa. It is a curious fact that adults of maculatus seem to be rare. Of three
hundred and forty-three specimens in the M. C. Z. collection, three hundred and
forty are less than 16 mm. h. d., one is 16 mm., one is 26.5 mm. and one is 30
mm. The two last are simply bare tests; the larger is the one figured in the
"Revision." So far as can be judged these large ones are not different from
the small ones except in such points as would be involved in their larger size.
In coloring, tuberculation, and pitting of the test, maculatus appears to be very
variable, but the coloring of the spines is fairly constant. For this reason,
it cannot be doubted that the species mentioned by Mortensen, from Tor, on
the Red Sea, is perfectly distinct, and may be called cricacanthus (K/HKOS = a
circle + aKavOa. = a thorn, spine) . It is unlikely it will prove to be Rousseaui, as
the latter is most probably the adult of maculatus. As for Anthechinus roseus
A. Ag., a reexamination of the type-specimen, now in the M. C. Z. collection,
leaves no room for question that it is maculatus. The test of young maculatus
is usually somewhat greenish or yellowish, with the bare abactinal spaces, rang-
ing from almost white to pure deep rose-color. Of the large specimens, one has
the bare spaces dull rose and the other very dark gray. Among our three
hundred and forty small specimens, no less than four are more or less per-
fectly tetramerous.

The five species of Microcyphus may be distinguished as follows : —

Test low, vertical diameter .50-. 70 h. d.; interambulacral plates few, 6-10 in each col-
umn; ambulacral plates nearly or quite twice as numerous; poriferous areas
broad; pore-pairs distinctly biserial actinally in large specimens, the inner series,
of pore-pairs only, the outer, of pore-pairs alternating with small tubercles.
Spines green, sometimes with light tips, and rarely with one or two distal bands . maculatus.

Spines red-brown, banded with 2—1 lighter rings cricacanthus.

Test high, vertical diameter .75-.90 h. d.; interambulacral plates 10-18 in each column;
ambulacrals only a little more numerous; poriferous areas narrow, without
tubercles and not biserial actinally.
Spines light, greenish at base, white at tip, broadly encircled by a band of red; bare

interambulacral spaces broad and very light colored annulatus.

Spines not banded.

Bare interambulacral spaces narrow and dark colored (yellowish brown in

young, almost black in adults) ; spines deep red zigzag.

Bare interambulacral spaces narrow, rose-red; spines pale reddish .... compsus

324 HAWAIIAN AND OTHER PACIFIC ECHINI.

SALMACOPSIS.

Doderlein, 1885. Arch. f. Naturg., Jahrg. LI, 1, p. 93.
Type-species, Salmacopsis olivacea Doderlein, 1885, 1. c.

Little need be said regarding this genus, which appears to be well marked
and characteristic of the Japanese fauna. Examination of the type-specimens
of Yoshiwara's species " pulchellimus " shows that that species is based wholly
upon specimens of Mespilia globulus. One of the specimens is peculiar in that
the spines are almost uniformly green and the outer series of pore-pairs, usually
conspicuous in Mespilia and in this specimen present at the ambitus, disappears
entirely abactinally so that most of the younger ambulacral plates are made up
of only two elements. Since Yoshiwara's species has no standing the genus
remains as it was when established, monotypic.

Salmacopsis olivacea D6<t

Salmacopsis olivacea Doderlein, 1885. Arch. f. Naturg., Jahrg. LI, 1, p. 93.

Plates 103, figs. 6, 7; 112, figs. 3, 4.

The specimens at hand range from 19 to 26.5 mm. in diameter and are
thus somewhat larger than Doderlein' s. They are also somewhat greener in
color, though none of them is conspicuously green, even on the bare interambu-
lacral areas. As the species has never been figured, photographs, as well as
drawings of the abactinal and actinal (in part) systems are given.

The "Albatross" took this handsome species at the following stations: —

Station 3708. Off Ose Zaki, Honshu Island, Japan. Bott. temp.? 60-70
fathoms. Gn. m., vol. s., a.

Station 4894. Southwest of Goto Islands, Japan; 32° 33' N., 128° 32' 10"
E. Bott. temp.? 95 fathoms. Gn. s., brk. sh., p.

Station 4937. In Kagoshima Gulf, Japan ; 31° 13' N., 130° 43' 10" E. Bott.
temp. 64.8°. 58 fathoms. M., lav., p.

Bathymetrical range, 58-95 fathoms.

Seven specimens.

AMBLYPNEUSTES. 325

AMBLYPNEUSTES.

L. Agassiz, 1841. Int. Mon. Scut., p. 7.
Type-species, Echinus griseus Blainville, 1825. Diet. Sci. Nat., XXXVII, p. 81.

In the present state of our knowledge, this is one of the most perplexing
genera of recent Echini. The species have always been in confusion, and in
spite of Mortensen's excellent work (1904, Dan. Exp. Siam: Ech.), there is
still great difficulty in identifying specimens and in determining specific limits.
The difficulty has arisen largely from the facts that no trained echinologist
has been able to examine these Echini in life, and practically all Museum material
has consisted of bare tests, which show great individual diversity in color and
proportions. It is only within recent years that our Museums have obtained
material with the spines on and even now such material is not common. Mor-
tensen's view that no reliance can be placed on the identifications of Amblyp-
neustes hitherto is wholly justified and while his work, combined with the results
here published, will not make identifications perfectly reliable, they will, it is
hoped, increase the possibility of distinguishing the various species. It seems
certain however that the truth in regard to the species of Amblypneustes can only
be known, when some zoologist in Australia carefully determines the valid species
and the limits of individual diversity. This work must be done in southeastern
Australia and Tasmania, which is the principal, if not the exclusive home of the
genus. I have seen no authenticated specimens from elsewhere and although
Amblypneustes has been recorded from the Cape of Good Hope, from the Fiji,
and Santa Cruz Islands, from New Caledonia and New Zealand, and from
western Australia, all these records need confirmation. There is no doubt that
the Cape of Good Hope record is wrong, and that the western Australian record
is correct, but there is great doubt as to the others. It is a curious fact that
although two species (not to mention Studer's grossularia) have long been listed
from New Zealand, the zoologists of the present day there, apparently do not
know the genus at first hand. In none of the valuable papers on New Zealand
echinoderms by Farquhar or Benham, is there any new information regarding
the occurrence of Amblypneustes in New Zealand.

So far as can be judged from Mortensen's key and from the material in the
M. C. Z. collection there are probably half a dozen valid species of Amblyp-
neustes; of the eight recognized in the table below, at least two seem very
dubious. It is more than possible that ovum and griseus are identical, for
although typical examples of each are obviously different, specimens that might

326 HAWAIIAN AND OTHER PACIFIC ECHINI.

with almost equal propriety, go in either one species or the other, are in the series
studied. As Mortensen has pointed out, the name " Amblypneustes ovum"
has been used very carelessly. It is borne on the original labels of more than
two thirds of our specimens of Holopneustes ! But typical examples of the true
ovum are easily determined. The status of grossularia is troublesome but it
is included in our table in accordance with the information given by Studer
and Mortensen. The latter seems to think it may not be an Amblypneustes
at all and this seems quite probable, but facts in support of the belief are want-
ing. It is very difficult to find tangible characters to distinguish it even as a
species and we are obliged to rely upon the peculiarities of the pedicellarke, as
given by Mortensen, though unable to find in them constant characters. It is
unfortunate that Mortensen does not include grossularia in his key to the
species of Amblypneustes, so that its really diagnostic features might be made
clear. The M. C. Z. specimens of formosus and pallidus appear to agree with
those in the Copenhagen Museum.

In addition to the species already known, there are three others in the
M. C. Z. collection which appear to be undescribed, so that eight species are
distinguished in the following manner: -

Ambulacra narrower than interambulacra; pore-pairs in distinct arcs of three.

Spines red; coronal plates handsomely marked in median interambulacral area with
a broad zigzag, furrowed band, lighter colored than the lozenge-shaped areas
which separate its outer angles; poriferous areas usually light-colored .... formosus.
Spines not red; coronal plates not marked as above.

Interambulacral plates high and few (18 in a specimen 20 mm. h. d.); second-
ary and miliary spines and tubercles very few; valves of ophicephalus pedi-
cellariae constricted near middle; those of triphyllous expanded at tip . . grossularia.
Interambulacral plates numerous (more than 20 in specimens 20 mm. h. d.);
secondary and miliary spines and tubercles more or less abundant;
valves of ophicephalous pedicellariae not specially constricted nor those
of triphyllous expanded at tip.
Vertical diameter more than .90 of horizontal; test usually ovoid and little

flattened actinally.

No secondary tubercles on interambulacral area between the two
vertical series of primaries but the area is crossed by a zigzag line
of miliary tubercles which connect the primary tubercles of oppo-
site sides; primary spines green, secondaries pale purple . . . pallidus.
Secondary tubercles common on interambulacra; no zigzag line of

miliaries at all obvious ovum.

Vertical diameter less than .90 of horizontal; test sphseroidal or depressed,

more or less flattened actinally.

Poriferous areas rather wide, the two together forming nearly or quite
half the ambulacrum.

AMBLYPNEUSTES PACHISTUS. 327

Ambulacral plates not very numerous (31 or 32 in specimens
34 mm. in diameter) about 30-50 % more numerous than in-
terambulacral; primary tubercles very small, those of am-
bulacra near ambitus occupying about one half the height
of plate and those of interambulacra occupying about one

third the height of plate griseus.

Ambulacral plates numerous (40-43 in specimens 34 mm. h. d.)
about 50-70% more numerous than interambulacral; pri-
mary tubercles large, those of ambulacra occupying almost
the entire height of plate, especially actinally, where they
form a rather crowded vertical series; those of interambul-
acra occupy about one half the height of plate .... pachislus.
Poriferous areas narrow, the two together forming only one third
of ambulacrum; vertical diameter .ti5-.75 of horizontal; tuber-
culation relatively coarse; test and primaries pale brown . . gram/is.
Ambulacra broader than interambulacra; pore-pairs in three very distinct vertical

series . Iriseriatus.

Amblypneustes pachistus,1 sp. nov.

Plates 104, fig. 6; 112, figs. 10-11; 121, figs. 1-3.

There are in the M. C. Z. collection eleven specimens of Amblypneustes which
are easily distinguished from griseus and ovum by their much stouter tests and
coarser tuberculation. While it is possible that they are only a form of griseus,
they are so easily distinguishable, it seems better to give them a name, and
describe and figure them. The largest (PI. 112, figs. 10, 11) is 48 mm. in diame-
ter and 34 mm. high. The actinostome is 15 mm. across, while the abactinal
system is only two thirds as much. There are about 34 interanibulacral and 54
ambulacral plates in each column. The spines are practically wanting. A
somewhat smaller specimen is 30 mm. in diameter and 25 mm. high, and has the
actinostome 11 mm. across. There are 25 interambulacral plates in each column
and no less that 39 ambulacrals. The spines are present, in some numbers,
and the largest primaries, just below the ambitus, are 4 mm. long. In a speci-
men 21 mm. h. d., there are 22 interambulacral plates, and more than 40 ambula-
crals, in each column.

The test is well arched and rather high, the vertical diameter ranging from
.70 to .80 of the horizontal. The coronal plates are all low and wide. At the
ambitus, each interambulacral plate carries a primary tubercle, somewhat nearer
the ambulacral than the median suture. On the inner side of this tubercle

very coarse or stout.

328 HAWAIIAN AND OTHER PACIFIC ECHINI.

and on the same horizontal line are three or four secondaries, while on the ambula-
cral side are a pair or two pairs of similar tubercles. One or more of these
secondaries are nearly or quite as large as the primary and none are very small,
but each plate also carries about twenty miliaries. The areola of the primary
tubercle occupies one half or more of the height of the plate. In the ambulacra,
each plate has a primary tubercle close to the poriferous area, two secondaries
little smaller on the inner half, and six to eight miliaries, of which two are in
the poriferous area. The areola of the primary occupies nearly or quite the
full height of the plate, particularly below the ambitus, but in large specimens
even far above it. The ambulacra are .80-.90 of the interambulacra, in width
at the ambitus, while each poriferous area is about .45 of the interporiferous.
The genital plates are of moderate size, approximately equal, with the pores
in the distal ends; the pores are often so large as to encroach on the coronal
plates, cutting clear through the margin of the genital. The oculars are small
and all are broadly exsert. The periproct is relatively rather large and is covered
with numerous small plates. The entire abactinal system is completely covered
with small tubercles. The actinostomal membrane is thin and bare, as usual in
the family. The buccal plates are small, and though the two of a pair are near
together, the pairs are widely separated. The spines are as usual short and
rather stout. The primaries taper slightly or not at all and are truncate at the
tip; many of the secondaries are distinctly swollen distally. The pedicellarise
and spicules show nothing of interest.

Although this species is similar to griseus in the form of the test, the much
stouter and coarser appearance distinguish it easily. The coloration appears
to be different also. The bare tests of the largest specimens are yellowish, the
poriferous areas lighter than elsewhere. The three smallest specimens are dull
gray rather than yellow and in one specimen a very evident greenish tinge is
present actinally. The finest specimens, about 35 mm. in diameter (PI. 121,
figs. 1-3 is the type) collected at Westernport, Victoria, March, 1911, have the
test very light, almost cream-color; the tube-feet and all the small spines are
of about the same shade, but the primaries are deep greenish brown in striking
contrast. There are two other specimens with some spines; in one, these are
dull purple becoming light at the tip especially around the mouth where the
distal half is almost white; in the other, the spines are very pale, almost white,
but with a more or less evident purple tinge, especially near base. It is hard to
decide whether this diversity in color is characteristic of the species or whether
we have several different species included under the name pachistus.

AMBLYPNEUSTES GRANDIS. 329

So far as can be determined this is the species figured by Valenciennes on
Plate 2 of the Zoophytes of the "Voyage of the Venus," as Echinus (Amblyp-
neustes) pallidus. As Mortensen has pointed out that figure is certainly not
pallidus, as that species is now understood.

Excepting those recently collected at Westernport, Victoria, not one of our
specimens has a reliable locality label. These were purchased in Europe in
1870; five are labelled "New Zealand," two are labelled "Australia" and one
has no label. It is rather remarkable that other specimens of this species have
not been included among the many specimens of Amblypneustes and "Holo-
pneustes" received during the past few years at the M. C. Z. from New South
Wales and Victoria. It is apparently much less common at Westernport than
griseus, of which we received a good series, showing practically no diversity in
form or color.

Amblypneustes grandis,1 sp. nov.

Plate 121, figs. 4-6.

This is the largest species yet known, the type specimen having a horizontal
diameter of 70 mm. It is 49 mm. high, with the actinostome 18 mm. in diameter
and the abactinal system 13 mm. across. There are 33 interambulacral and 50
ambulacral plates in each column. The interambulacra are 26 mm. wide at
the ambitus. The ambulacra are 17 mm. wide but of this the two poriferous
zones together only occupy 6 mm. The pores are rather large and the arcs of
pores are quite oblique especially abactinally. The tuberculation of the test
is rather coarse for an Amblypneustes and it is virtually impossible to distinguish
between the primary and secondary tubercles. Each interambulacral plate
carries a horizontal series of these larger tubercles; on most of the plates, the
series bifurcates so that on the outer (adradial) half of the plate there are two
series, the whole group forming a narrow, elongated Y lying horizontally on
the plate; the stem of the Y contains 6 or 7 tubercles while each of the branches
consists of about 4. Besides these larger tubercles, there are some forty miliaries
scattered about all over the plate. Each ambulacral plate has a horizontal
series of four or five large tubercles, of which the innermost is smallest while the
one adjoining the poriferous zone is largest, its areola occupying about one half
the height of the plate; there are also a dozen or more miliaries on each plate;
of these two are placed side by side just above the largest tubercle, but their

1 grandis = large and fine.

330 HAWAIIAN AND OTHER PACIFIC ECHINI.

distance from each other is variable. The whole abactinal system is covered
by numerous tubercles of which four or five on the proximal margin of each
genital are large secondaries. The genital pores are large and occupy the distal
tips of the genital plates; the oculars are small and are all broadly exsert. The
buccal membrane is of course thin and perfectly bare; the buccal plates are very
small and not only are the pairs well separated from each other but the two plates
of a pair are some distance apart. The primary spines are 7 or 8 mm. long;
they taper little to the truncate tip, which usually has a conspicuous (under the
lens) central thorn. Though some of the small secondaries are a little thickened
at the tip, none of the spines are noticeably club-shaped. Pedicellarise are
exceedingly abundant all over the test.

The general color of the test is brown, darkest on the interambulacra where
there is a distinct purple shade; the ambulacra are more greenish. In both
areas there is a distinct band of a light color, 2 or 3 mm. wide, along the median
vertical suture line, but these bands disappear actinally. The poriferous areas,
with the adjoining margins of the interambulacral plates, are somewhat lighter
than the remainder of the test. The small spines are almost white but the larger
spines are pale brown, though actinally they are tipped with lighter, and around
the mouth they are nearly white for their distal half. There is little indication
of either green or violet in the coloration but along the abactinal portions of the
poriferous areas, there are traces of dull rose-red.

A second specimen measures 43 mm. in horizontal diameter and is 31 mm.
high. There are 30 interambulacral and 45 ambulacral plates in each column.
The interambulacra are 16 mm. wide and the ambulacra, 11; the poriferous
areas are each scarcely 2 mm. wide. The tuberculation of the test differs from
that of the type only as might be expected from the difference in size; there
are not so many large tubercles and the horizontal Ys on the interambulacral
plates are much less regular and noticeable; the pair of miliary tubercles above
the largest ambulacral primary (the one next the poriferous area) are very
regular in position and form a noticeable feature of the ambulacra. The colora-
tion is similar to that of the type but the median light colored bands in each area
are less distinct, and the dull rose shade on the poriferous areas is more marked
and is evident on all the abactinal coronal plates; it tends to form a more or
less distinct pattern on the interambulacra. The large spines are all much
lighter than in the type, but they are a little more brownish than the miliaries.

These two specimens were collected by Mr. J. Gabriel, in March, 1911, at
Westernport, Victoria, in 2-5 fathoms. The coloration and form of the test

AMBLYPNEUSTES TRISERIATUS. 331

distinguish them at a glance from griseus, while they are equally different from
the specimens of pachistus which came with them. The very narrow poriferous
areas, the peculiarities of the tuberculation and the color combine to make
them representatives of an undescribed species. Among the specimens in the
M. C. Z. collection, received from the "Challenger" material as A. formosus
and taken in Bass Strait, 38^40 fms., are two which seem to be the young
of grandis. One is 22 mm. in diameter and except for the fact that it is
much lighter colored and has the dull rose markings more distinctly indicated,
it resembles the larger specimens quite closely; the pair of miliaries above each
of the outer ambulacral primaries is noticeable; the primary spines are white,
but of course they may have become bleached with the passing of forty years.
The other specimen is only 16 mm. in diameter and is quite possibly not grandis;
the narrow poriferous areas and the tuberculation of the ambulacra are charac-
teristic but the coloration is peculiar, for the test has a dull greenish cast and the
large spines are bright green with white tips.

Amblypneustes triseriatus,1 sp. nov.
Plates 104, fig. 5; 112, fig. 5.

The specimen upon which this new species is based is a bare test without
actinostomal membrane or abactinal system, but so unique in the characters
it shows that it is without doubt specifically distinct from any described Amblyp-
neustes. This test is 30 mm. in diameter and 26 mm. high, with the actino-
stome 11 mm. in diameter and the abactinal system (wanting) 6.5 mm. across.
There are 28 interambulacral, and 43 ambulacral plates in each column. The
ambulacra (PL 104, fig. 5) are 9.5 mm. wide at the ambitus, while the interambul-
acra are scarcely 9, so that the former are obviously the wider. The two porif-
erous areas together are about equal in width to the interporiferous space. The
pores are small and the pore-pairs are so uniformly arranged and the arcs are so
nearly horizontal that each poriferous area shows three distinct vertical series of
pore-pairs. Corresponding to these but much more irregular and imperfect are
three vertical series of miliary tubercles. Of these tubercles those just within
and below the middle pair of pores are the largest. Each interambulacral plate
at the ambitus bears a primary tubercle, a little outside of the centre, the areola
of which occupies scarcely one half the height of the plate. On the outer end
of the plate is a secondary tubercle (or often there are two) and on the inner half,

1 triseriatus = in three series.

332 HAWAIIAN AND OTHER PACIFIC ECHINI.

there are two (or often only one) , and one (or more) of these secondaries is almost,
if not quite, as large as the primary. On each of these plates there are also
15-20 miliaries, of diverse sizes but all small. On each ambulacra! plate, there
is a primary tubercle close to the poriferous area and on the inner half of the
plate, there may be a second tubercle nearly as large. These inner tubercles
are of somewhat variable size and are altogether wanting on many plates. Be-
sides the miliaries in the poriferous area, already described, each plate carries
five or six more on its inner half, irregularly scattered but chiefly near the margins.
The areola of the primary tubercle occupies practically the whole height of the
plate, so that this series of tubercles bordering the poriferous area is very similar
to that of pachistus.

The color of the test is grayish olive, becoming lighter actinally so that around
the actinostome it is nearly cream color. The poriferous areas are distinctly
lighter than the spaces between them. The larger tubercles are all white. Under
a lens, the coronal plates at least on the inner half are seen to be variegated with
a lighter shade, and in the median interambulacral areas these lighter lines tend
to form figures similar to those seen in formosus. The medial vertical suture
in both areas is indicated by a lighter line and the pits at the angles of the plates,
along this line, though very small, are still visible.

This specimen was received into the M. C. Z. collection from Louis Agassiz,
in 1859. It bore no other label than "Australia," but was marked by A. Agassiz
" Nov. gen." No description of it has ever been published, as it has been hoped
that additional specimens might come into the collection. The hope has been
in vain however and attention is now called to this noteworthy species. While
it appears to be allied to formosus, as nearly as to any species of the genus, the
appearance of the ambulacra, when the specimens are placed side by side, is
strikingly different in the two.

HOLOPNEUSTES.

Agassiz and Desor, 1846. Ann. Sci. Nat., (3), VI, p. 364.
Type-species, Holopneustes porosissimus Agassiz and Desor, 1846, 1. c.

There can be no question that if the law of priority is to be rigidly enforced,
this genus is without a name while the equally well-known Mespilia becomes
Holopneustes. To prevent such a disaster it is preferable to ignore the first
suggestion of Holopneustes by Agassiz in 1841, when he mistook one of Leske's
species and selected it as the type, and date the genus from the time of its first

HOLOPNEUSTES. 333

real description. This is justified because it is clear that Agassiz intended, when
he suggested the name in 1841, that Holopneustes should be the name of these
Echini with highly specialized ambulacra; this is unmistakable from his ac-
companying remarks. It is not reasonable that his intent should be nullified
by a mistake in identifying one of Leske's inadequately described, and poorly
figured species, especially when that mistake was rectified at the first opportunity.

The simple method suggested by Mortensen (1904, Dan. Exp. Siam: Ech.,
p. 102) for distinguishing Holopneustes from Amblypneustes by means of the
arrangement of the ambulacral tubercles, works very well and is a great con-
venience in distinguishing specimens of the two genera, when confused in the
same lot, as they commonly are !

Three species of Holopneustes have usually been recognized but Mortensen
has suggested that inflatus and purpurascens may be identical, and this is the case,
as the material now at hand shows that the characters, supposed to be specific,
are not constant. Mortensen (1904, Dan. Exp. Siam: Ech.) writes both inflatus
and purpurascens with Liitken as the authority, and states that the types are
in Copenhagen. While it is true that the names were suggested by Liitken, they
were published by A. Agassiz and so far as we know were not even used by Ltitken
in print. It is hard to see therefore why they should be credited to Liitken.
Dr. Mortensen is however consistent in this, for he quotes several of Agassiz
and Desor's species, as from Valenciennes, although they were simply manuscript
names of the latter, on museum labels. As for the location of the types, it is a
question which cannot be settled, for Mr. Agassiz never definitely designated
type-specimens. But there are specimens in the M. C. Z. collection, on which
the descriptions and figures published were based, and these, by some at least,
would be considered the type-specimens. In uniting the two species, the name
inflatus is retained because it has "page-precedence" in publication, because it
is shorter, and because purpurascens is often inappropriate.

More than half the specimens of Holopneustes at hand seem to be neither
porosissimus nor inflatus, and a third species, pycnotylus is therefore described.
The three species are all from southeastern Australia and Tasmania and are not
even reported from elsewhere.

They may be distinguished as follows : —

Interambulacral plates fairly numerous (22-25 in specimens 22 mm. h. d., about 46
in specimens, 48 mm. h. d.), their tubercles small; areolse of primaries only
occupying about half height of plate, and forming a well-spaced vertical series
on each side of interambulacrum.

334 HAWAIIAN AND OTHER PACIFIC ECHINI.

Ambulacra decidedly wider than interambulacra; primary spines greenish, more

or less extensively tipped with red porosissimus.

Ambulacra not appreciably wider than interambulacra and usually distinctly

narrower; primary spines vary from very pale brown to reddish purple . . . inflatus.
Interambulacral plates very numerous (30 in specimens 22 mm. h. d., 40-50 in speci-
mens over 36 mm. h. d.), their tubercles large, areolae of primaries occupying
nearly whole height of plate and hence forming more or less crowded vertical series
on each side of interambulacrum; primary spines whitish or cream color, or pale
purplish or greenish, often tipped with lighter or darker; ambulacra in adults
usually distinctly wider than interambulacra pycnolylus.

Holopneustes pycnotylus,1 sp. nov.
Plates 104, fig. 4; 112, figs. 6-9.

~ The form of the test in this species shows great individual diversity, appar-
ently not associated with age. The type (PI. 112, fig. 6) is not quite 36 mm. in
diameter, and is 32 mm. high, the height thus equalling .90 h. d.; the ambitus
is circular. Another specimen (PI. 112, figs. 8, 9) also with a circular ambitus,
is 33 mm. in diameter and only 25 mm. high; the height is thus only about .75
h. d. A specimen 23 mm. in diameter is 19 mm. high (v. d. = .83 h. d.) while
another almost as large is only 16 mm. high (v. d. = .71 h. d.). A specimen
(PI. 112, fig. 7) 34 mm. in diameter and 30 mm. high has the ambitus very
distinctly pentagonal. In a specimen 19 mm. h. d., there are 26 interambulacral
and 49 ambulacral plates in each column; in another, 23 mm. h. d., the numbers
are 30 and 55 respectively; in another, 33 mm. h. d., 45 and 88; in another,
34 mm. h. d. (pentagonal) 37 and 76; in the type, 36 mm. h. d., 38 and 90; and
in another, 37 mm. h. d., 52 and 113. The relative width of ambulacra and
interambulacra shows some diversity also. Thus the same six specimens give
the following measurements, the ambulacral width being given first : — 6 and 6, 7
and 6.5; 11 and 8; 10 and 10; 11.5 and 10; 12 and 10. The ambulacra are thus
on the average about 15% wider than the interambulacra. The abactinal system
and actinostome are small, the latter averaging about .30 h. d., while the abacti-
nal system is only half as large. The poriferous areas (PL 104, fig. 4) are wide,
each one at least half as wide as the interporiferous area, and in adults the propor-
tion rises to three fourths or even four fifths. The arrangement of the pores and
tubercles is as usual in the genus; there is a more or less perfect vertical series
of pore-pairs on each margin of the poriferous area and between these are numer-

= crowded + TvXos = a tubercle.

HOLOPNEUSTES PYCNOTYLUS. 335

ous scattered pore-pairs. Most of the ambulacral plates carry primary tubercles,
but in scarcely a third of them is this tubercle adjoining the poriferous area.
The plates are very unequally developed and some are much higher than others.
The areolffi of the primary tubercles nearly or quite equal the height of the plates.
Each plate has about four secondary tubercles, at least one of which is nearly
or quite as large as the primary, and two are usually very small and are located
in the poriferous area. In the type, miliary tubercles appear to be wanting but
in other specimens, they are fairly numerous. Each interambulacral plate at
the ambitus carries a primary and 5-8 secondaries. The primaries are a little
nearer the outer than the inner end of the plates and form a regular vertical
series. In typical cases, with low plates, this series is crowded, the areolse
occupying the full height of the plates, but in the type and some other specimens
the plates are higher and the tubercles more separated. One or more of the
secondaries is nearly or quite as large as the primary and sometimes the row of
primaries is thus made to appear double for part of its length. The madreporic
genital is conspicuously larger than the others. The oculars are small and fully
exsert. Both genitals and oculars are well covered with tubercles, but the
numerous periproctal plates carry none. The actinostomal membrane is very
thin and perfectly naked. The buccal plates are very small and placed quite
near the mouth.

The primary spines of the type are about 4 mm. long, rather stout and bluntly
pointed. Most of them taper slightly to the tip, but many are distinctly swollen
there. The secondaries are much more slender than the primaries but some are
about as long. Many of them are swollen at the tip. Pedicellariae are very
numerous but excepting some of the ophicephalous, which have valves over half
a millimeter long, they are all very small. Otherwise neither they nor the
spicules afford any characters of interest.

The type is uniformly dirty cream color, spines and all, but most of the other
specimens are gray with more or less of a purple, or a red, or a yellow tinge, and
their spines show the same shades to a greater or less degree.

Most of the specimens of pycnolylus have been received from Port Jackson,
New South Wales, labelled " Amblypneustes ovum." One label further says:
"Coastal beaches outside; after gales only," but other specimens do not have
this limiting phrase on the label.

336 HAWAIIAN AND OTHER PACIFIC ECHINI.

GONIOPNEUSTES.

Duncan, 1889. Journ. Linn. Soc. London. Zool., XXIII, p. 113.
Type-species, Amblypnfiistes pentagonus A. Agassiz, 1872. Bull. M. C. Z., Ill, p. 56.

There can be no doubt that Duncan was quite right in instituting a new
genus for the remarkable and apparently very rare sea-urchin, which A. Agassiz
described in 1872 as an Amblypneustes from Mauritius. The abactinal system
with oculars I and V fully insert, the high coronal plates, the presence of scattered
plates on the buccal membrane, the scarcity of spines and the large size of the
primaries combine to make the genus unusually well characterized. The type
and only known species was based on a single specimen, in the M. C. Z. collec-
tion, and figured in the Revision. (Since the text states correctly the measure-
ments of the specimen as 22 X 21.5 mm. while the figures on PL VIIIC measure
32 X 30 mm., it is evident that the "explanation of the plate" is erroneous in
saying the figures are natural size.) The locality, whence this specimen came,
is very doubtful. It was purchased in Hamburg in 1870 and bears the label
"Brandt. Hamburg." It is entered in the M. C. Z. catalogue as from "lie de
France?" So far as known, no other specimens are to be found in museums,
and neither Pike nor Mobius, nor Robillard in their extensive collecting at
Mauritius ever met with this remarkable echinoid. It is most probable that the
specimen came from Australia.

Goniopneustes pentagonus Duncan.

Amblypneustes pentagonus A. Agassiz, 1872. Bull. M. C. Z., Ill, p. 56.

Goniopneustes pentagonus Duncan, 1889. Journ. Linn. Soc. London. Zool., XXIII, p. 113.

Plate 93, figs. 18-21.

In addition to the characters given in the "Revision," the following points
may be mentioned. There are 18 interambulacral plates in each column and
each one bears a single, conspicuous, imperforate, non-crenulated tubercle.
There are also on each plate a few (6 in the mid-zone) well-spaced secondaries,
and some very minute and widely scattered miliaries; the latter bear pedicellaria?
but not spines. There are 43 ambulacral plates in each column but not more
than 12-15 carry primary tubercles; at the ambitus and below, every other
plate has a primary but above the ambitus, there are very few indeed. There

GONIOPNEUSTES PENTAGONUS. 337

are rarely more than two well-developed secondary tubercles on each ambulacral
plate, but there are quite a number of very small ones, in addition, in the porif-
erous area. The arcs of pores are very oblique and are not at all crowded. The
pits at the angles of the plates are fairly distinct but are exceedingly small in
both ambulacral and interambulacral areas.

The periproct is covered by about 30 small plates, no one of which is notice-
ably larger than the others. Oculars I and V are broadly insert but the others
are completely shut out. The ocular pores are large and placed near the centre
of the plate, but curiously enough no genital pores are to be seen, even with a
lens. Each genital plate carries a single secondary tubercle and smaller ones
are to be found on oculars I, III and IV. There are also rather numerous
miliary tubercles on oculars, genitals, and periproctal plates. The buccal
membrane is rather thick, thicker at least than in Amblypneustes, and, sur-
prising to find, carries numerous small, scattered plates which are rather crowded
around the mouth. The buccal plates themselves are rather large and carry
large tube-feet; the two of each pair are near together but the pairs are well
separated from each other. The buccal plates and the scattered plates distal
to them carry pedicellarise. The gill-cuts are deeper and more conspicuous than
in Amblypneustes.

The primary spines are about 6 mm. long, in the mid-zone; they are very
finely striated and taper to a blunt point. The secondaries are shorter and more
slender and many appear to be pointed, though some are slightly swollen near
the tip. Pedicellarise are very common, but no tridentate could be found. The
others remind one at once of Gymnechinus. The globiferous have valves (PI. 93,
fig. 19) about .30 mm. in length, with a long end tooth (PI. 93, fig. 20) but no
lateral teeth. The ophicephalous are larger, having valves (PI. 93, fig. 18) about
.40 mm. long, but they show no distinctive characteristics. The triphyllous
are very smalt, having valves (PL 93, fig. 21) only .12 mm. long. The calcareous
spicules in the tube-feet are C-shaped. Sphseridia are few and very small.
The test is almost fawn-color but has a distinct purplish cast, except in the
bare median areas of both ambulacra and interambulacra. The primary spines
are pure white but the secondaries are gray, tinged with purple. The buccal
membrane is dark, but the plates it bears are light.

338 HAWAIIAN AND OTHER PACIFIC ECHINI.

STRONGYLOCENTROTID^E Gregory.

GENERAL CONSIDERATIONS.

The real relationship of the polyporus Echini with the Echinidse can probably
be made more evident, if the forms with a circular ambitus are separated from
those in which one axis is elongated. I therefore agree with Jackson (1912,
Mem. Boston Soc. Nat. Hist., VII) when he follows Gregory in separating
Strongylocentrotus and its nearest allies from the Echinometridse and giving
them family rank. The group contains some thirty or more species which form
so complete a series from those nearest the Echinidse to those most highly spe-
cialized that it is exceedingly hard to arrange them in genera. Of course taking
some one character as the standard, artificial sections, called genera, can be
recognized but the associations thus formed are often valueless. The connection
of the family with the Echinidse through its lowest genus Paracentrotus is so
plain it will hardly be questioned, the resemblance of Paracentrotus agulhensis
to certain species of Echinus is so great that the M. C. Z. has received specimens,
labelled "Echinus gilchristi" from the describer of the latter species, himself!
The Strongylocentrotidse are widely distributed but it is a remarkable fact that
not a single species is known from the Caribbean region. Nearly one third of
the species are found in the northwestern Pacific, mostly in Japanese waters.
All of the species are littoral and only a few extend into water exceeding one
hundred fathoms.

THE SPINES, PEDICELLARI^E, SPH^RIDIA, AND SPICULES.
Plates 94, figs. 1-33; 95, figs. 16-24.

In the character of the spines, the Strongylocentrotidse show little indications
of advance over conditions found in the Echinidse. In most of the species the
spines are relatively short and there is no marked contrast between the primaries
and the other spines. In Echinostrephus however the abactinal primaries are
conspicuous; they are long, slender, very smooth, and rather sharp. In Helio-
cidaris and in some of the most specialized species of Strongylocentrotus, the
primary spines are long and stout, usually quite smooth, and more or less pointed;
in one species of Heliocidaris they are remarkably short and thick. The charac-
ter of the primary spines may thus be of considerable use for specific distinctions.

STRONGYLOCENTROTID.E. 339

In the pedicellarice, the differentiation of special forms undergoes a series of
changes identical with that which exists in the Echinidse. Thus in the simplest
forms, those with only four pairs of pores and all oculars exsert, the globiferous
pedicellarise, like those found in the most primitive species of Psammechinus
and Echinus, have valves with lateral teeth on both sides. In forms somewhat
more specialized, a lateral tooth on only one side is found and in many species
that has disappeared and there is only the conspicuous terminal tooth, so char-
acteristic of the most specialized Echinidse. It is interesting to find that many
of the more specialized species of Strongylocentrotidse have retained the less
specialized form of pedicellaria? ; thus Loxechinus albus, a highly modified species,
has the simple form of globiferous pedicellarise. On the other hand, Strongylo-
centrotus granularis, which is very variable but without highly specialized am-
bulacra, has the most extreme form of pedicellarise. There do not seem to be
any characters in which the pedicellarise of the Strongylocentrotidse differ
regularly from those of the Echinidse. It is neither necessary nor desirable
therefore to repeat here their characteristic features. Usually all four kinds are
present on any given specimen; not infrequently the tridentate occur in two
forms and in a few species, there are two forms of globiferous. In some species
the globiferous pedicellarise are quite wanting, at least in adult specimens, but
they are usually common enough.

In their sphceridia, this family shows no special advance over what is found
in the Echinidse. There are several to many sphseridia on the actinal portion
of each ambulacrum, attached to minute tubercles, not at all sunken in depres-
sions in the test. They vary in shape from almost globular to oval or ellipsoidal.
They are usually smooth but may be quite rough at the tip.

The spicules in the tube-feet and in the glands of the globiferous pedicellarise
are usually bihamate, with the ends simple or more or less branched; often the
spicules are not developed completely but are short, slightly curved with rounded
ends. What Mortensen calls "biacerate" spicules also occur. As any two or
three of these forms may occur in a single specimen, they are without any real
systematic importance.

THE GENERA AND SPECIES OF STRONGYLOCENTROTID.E.

In attempting to arrange the species of this family in anything like a natural
sequence, the genera, with the exception of Echinostrephus and Pseudoboletia,
which are really a little off the main line of development, are exceedingly hard

340 HAWAIIAN AND OTHER PACIFIC ECHINI.

to define. Mortensen (1903, "Ingolf " Ech., pt. 1) has made clear the interesting
characters shown by the pedicellarise, and although he greatly exaggerates their
importance, it is beyond question that they are suggestive and oftentimes useful
for systematic purposes. By using them in connection with the structure of
the ambulacra, the arrangement of the oculars, the buccal membrane, the thick-
ness or thinness of the test, and the primary spines, it is possible to break up the
large and heterogeneous genus Strongylocentrotus into smaller and more homo-
geneous genera. Mortensen made seven such groups, basing his divisions
chiefly on the pedicellarise and spicules, and placing the resulting genera in three
different families. His genera are Strongylocentrotus, Paracentrotus, Lox-
echinus, Sphserechinus, Pseudocentrotus, Anthocidaris, and Toxocidaris. Of
these the first three seem natural groups, and the same is true of the last, though
it is necessary to recognize it under another name. As already shown (p. 281),
the type of Toxocidaris is Echinus erythrogrammus Val. and the type of Helio-
cidaris is Echinus tuberculatus Lamk. As these two species are indubitably
congeneric, the later name (Toxocidaris) becomes a synonym of the earlier
(Heliocidaris) , and the genus, which Mortensen calls Toxocidaris, and for which
he definitely designates erythrogrammus as the type, must be called Heliocidaris.
As regards Sphserechinus, it is impracticable to retain the genus, in spite of
Mortensen's statement that it is "very well characterized." His definition
of the genus as "large, short-spined forms, almost globular," and other references
to the "high form" of the test, show that he has not examined large series of
specimens. Many of our specimens of Sph. granularis from the Azores are
greatly flattened, the vertical diameter scarcely exceeding one half the horizontal,
and one specimen, 76 mm. in diameter is less than 37 mm. high. Many of
these specimens also have somewhat longer spines than usual, and except for
the deep gill-cuts and the insert oculars, would be easily mistaken for Paracen-
trotus limdus from the same islands. In the number of pore-pairs in an arc
these specimens of granularis from the Azores, show great diversity, many arcs
having as many as seven pairs, and six seems to be the typical number for adults.
The deep narrow gill-cuts furnish a good specific character and one that is
remarkably constant, but unfortunately it does not make a useful generic charac-
ter, for it occurs nearty or quite as well developed in many specimens of S.
depressus and S. pulcherrimus, but in these species, is very variable and of little
significance. The species hitherto known as " Sphoerechinus " australice proves
to have little in common with granularis, except well-defined gill-cuts, and it
may best be placed in a genus by itself. It cannot retain the name Sphserechinus

STRONG YLOCENTROTID.E. 341

since it was not included in the genus by the original describer of that group.
It is possible that Sphserechinus may still be used for some fossil forms but con-
sidering S. roseus Russo a synonym of granularis and placing the latter in Strongy-
locentrotus, Sphserechinus is not available for recent Echini. The differences
between the pedicellarite of granularis and those of its allies in Strongylocentrotus,
upon which Mortensen lays much weight, seem too trivial to be of any real
value. So too with regard to the characters by which Pseudocentrotus and
Anthocidaris are supposed to be distinguished. Had Dr. Mortensen examined
larger series of specimens, it is improbable that he would have attempted to
isolate those two genera.

There are two species, regarding whose status Mortensen expresses his
doubt owing to his inability to examine sufficient material, and of which good
series are at hand. One of these, Sphcerechinus australice A. Ag., becomes, as
suggested above, the type of the new genus Pachycentrotus. The other, Tox-
opneusles gibbosus Agas. and Des., while related to Paracentrotus in many ways
has such a modified abactinal system and such specialized globiferous pedicel-
larise, that it is better to make it the type of a new genus, Csenocentrotus. While
therefore rejecting three of Mortensen's genera, two new ones are here proposed
and thus the family contains only one less genus than he suggested.

The eight genera of Strongylocentrotidse adopted, are distinguished from
each other as follows; but the arrangement is very artificial and the various
characters are used arbitrarily regardless of their real importance, the only
object being to make the accepted genera tangible and easily recognized.

Test with ambitus above equator; not more than 4 pore-pairs in each arc; ocular

plates all fully exsert; primary spines longest on flat abactinal surface . . Echinostrephus.
Test with ambitus at or below equator (if above, there are more than 4 pore-pairs in

an arc, and one, at least, of the oculars is insert).

Test thin with deep gill-cuts, having a very prominent flange ("lip" or "tag")
on interradial side; plates of buccal membrane carry both spines and

pedicellarise Pseudobolelia.

Test seldom thin; gill-cuts without a prominent flange, and usually shallow;

plates of buccal membrane do not carry spines.
Pore-pairs in arcs of 4-10; if more than 5, valves of globiferous pedicellarise

with one or more lateral teeth.

Usually all oculars exsert but sometimes one and rarely more, are
insert; buccal membrane with scattered plates; valves of globif-
erous pedicellarise with one or more lateral teeth on each side.
Test not very stout; primary spines usually rather long, at least
.25 h. d.; valves of tridentate pedicellariee not peculiar; pore-
pairs 4 or 5 Paracentrotus.

Test very stout; primary spines rather short; valves of tridentate

342 HAWAIIAN AND OTHER PACIFIC ECHINI.

pedicellariae short, abruptly rounded at tip and with 6-8 con-
spicuous teeth there; pore-pairs 7-10 Loxechinus.

Usually two, not rarely three or more oculars insert; valves of globif-

erous pedicellariae with a lateral tooth only on left side.
Oculars V and IV insert, often III also, sometimes II and not rarely

all; test not very stout; buccal membrane with scattered plates Ccenocenlrotus.
Oculars I and V insert, rarely IV or II also; test very stout, thickly
covered with tubercles; buccal membrane more or less heavily

Plated Pachycentrotus.

Pore-pairs in arcs of 4-10; if fewer than 6, valves of globiferous pedicel-

laria; with tubular blade, prominent end tooth and no lateral teeth.
Pore-pairs 7-10; test stout; valves of globiferous pedicellarise with a

lateral tooth on left side near tip Heliocidaris

Pore-pairs usually 4-7, rarely 8-10; test thin or moderate; valves of
globiferous pedicellariae with no lateral teeth Strongylocentrolus.

ECHINOSTREPHUS.

A. Agassiz, 1863. Bull. M. C. Z., I, p. 20.

Type-species, Echinostrephus aciculatus A. Agassiz, 1. c.

A careful comparison of the "Albatross" material from the Hawaiian Islands
with that in the M. C. Z. collections from Mauritius, and from the Pacific Ocean,
shows that this interesting genus contains two species, aciculatus A. Ag. not
being identical with Echinus molaris Bl. Owing to the fact that in the "Re-
vision" they are regarded as identical, and the figures given are those of acicu-
latus, the description also being based on that species, Yoshiwara (1898, Ann.
Zool. Jap., II, p. 59) on finding specimens of the true molaris (PI. 105, figs. 10,11)
in Japanese waters (Bonin Islands) was misled into describing it as a new species
to which he gave the name pentagonus.1 The two species seem to be quite
distinct and are to be separated by the following characters : —

Pore-pairs 3; ambitus more or less evidently pentagonal; 'genital and ocular plates with

no tubercles except on outer margin; test usually more or less bright green . . molaris.

Pore-pairs 4; ambitus circular; genital and ocular plates with more or less numerous

tubercles; test dull purplish or greenish aciculatus.

Echinostrephus aciculatus A. Ag.

Echinostrephus aciculatus A. Agassiz, 1863. Bull. M. C. Z., I, p. 20.

Echinostrephus molare A. Agassiz and Clark, 1907. Bull. M. C. Z., L, p. 242. NON Blainville, 1825.

Plates 95, figs. 23, 24; 105, fig. 9.

The discovery that aciculatus is not identical with molaris gives added interest
to this unique genus and the specific characters have been studied in great detail.

1 This conclusion is confirmed by examination of Yoshiwara's specimens.

ECHINOSTREPHUS ACICULATUS. 343

Neither the pedicellarke, sphseridia, or spicules show any constant differences
between the two species but there is more or less individual diversity. The
globiferous pedicellarise are large and conspicuous (PI. 95, fig. 23) and it is inter-
esting to find that while the valves usually have a lateral tooth on only one side,
occasionally they occur with such a tooth on each side (PI. 95, fig. 24)- This
emphasizes the intermediate position occupied by the genus. The tridentate
pedicellarise are not usually common and they show the greatest diversity in
size, and nearly as much in form; the valves measure from .20 mm. to over a
millimeter in length. Most of our Hawaiian specimens lack tridentate pedicel-
larise, although globiferous and ophicephalous are common. When tridentate
are common, they are chiefly actinal in position, and ophicephalous are rare.
The ophicephalous on the buccal plates are much smaller than those on the test.
The latter have the valves about half a millimeter long, including the loop;
they are provided with long muscular necks; the valves are constricted near the
base of the very coarsely serrate or sinuate blade. The triphyllous pedicellarire
have valves .15-.20 mm. long and .10-. 15 mm. across the blade. Sphseridia
are numerous, 15-20 on the actinal part of each ambulacrum, but not extending
to the ambitus.

The "Albatross" specimens show great diversity in size and color. The
largest (PI. 105, fig. 9) is 37 mm. in diameter and nearly 25 mm. high; the
ambitus is about 6 mm. above the equator. The longest primaries, which are
very near the abactinal system, are 30 mm. in length. The test is dull grayish
or purplish with a slight greenish tinge on the primary tubercles. The primary
spines are very dark, almost bronze color and those below the ambitus are
abruptly tipped with deep pink. This specimen is from Laysan Island. Another
from the same vicinity differs strikingly in color, but is not otherwise peculiar.
The test is dull lavender with no trace of green; the secondaries, and the basal
half of the primaries below the ambitus, are bronze-brown; the abactinal pri-
maries and the terminal half of the actinal, are very pale lavender, almost white,
with more or less of a pink shade, especially at the very tip. The smallest of
the Hawaiian specimens is 12 mm. in diameter. It is of interest because, as
might be expected, there are only three pore-pairs in an arc actinally, but from
the ambitus, which nearly coincides with the equator, upward, each arc has four.

It is a noteworthy fact, that all of the "Albatross" specimens are from the
extreme northwestern part of the Hawaiian group, as will be seen from the fol-
lowing list of the stations where Echinostrephus was taken.

Station 3959. Off Laysan Island, Hawaiian Islands. Bott. temp. 75° ?
10 fathoms. Wh. s., co.

344

HAWAIIAN AND OTHER PACIFIC ECHINI.

Station 3960. Off Laysan Island, H. I. Bott, temp. 74° ? 10-19 fathoms.
S., sh., co.

Station 3968. French Frigate Shoal, H. I. Bott. temp. ? 14£-16£ fathoms.
Crs. s., co.

Station 3969. French Frigate Shoal, H. I. Bott. temp. ? 15-16 fathoms.
Crs. s., sh., co.

Station 3970. French Frigate Shoal, H. I. Bott. temp. ? 17-17^ fathoms.
Crs. s., sh., co.

Station 3975. Off Necker Island Shoal, H. I. Bott. temp. ? 16-171 fath-
oms. Crs. s., co., sh.

Station 4147. Off Modu Manu, H. I. Bott. temp. 77.9°. 26 fathoms.
Co., corln.

Bathymetrical range, 10-26 (171 ?) fathoms. Extremes of temperature,
77.9°-74° ? Twenty specimens.

PSEUDOBOLETIA.

Troschel, 1869. Sitzungsb. Niederrh. Gesel. Bonn, p.. 96.

Type-species, Pseudoboletia stenostoma Troschel, 1. c. = Toxopneustes indianus Michelin, 1862. Ech. et
Stel. : Annexe A, in Maillard's Notes sur Bourbon, p. 5.

The question as to the number of valid species in this interesting genus is
still open in spite of the writings of de Loriol, Bell, Mortensen, de Meijere, and
Koehler. It cannot be doubted that Kcehler's species from the Atlantic Ocean,
and which he called maculata, is really quite a new species and the name atlantica
is suggested for it. Neither is there any good reason to question that the species
so well described by de Loriol as Indiana is really that species, and the type of
granulata A. Ag. seems to be only a very large specimen (93 mm. h. d.) of the
same. But whether the species called by Troschel, maculata, is really entitled
to recognition seems doubtful. There are in the M. C. Z. collection, besides
the type of granulata, three specimens of Indiana from Mauritius, three speci-
mens of the "Challenger" species from Zamboanga, which Bell and Mortensen
call maculata, and a small bare test from an unknown locality. These speci-
mens seem to prove that the size of the peristome and the depth of the gill-cuts
do not furnish reliable specific characters and throw doubt on the value of the
coloration as a means of separating the two species. In the type-specimen of
granulata, the test is 3.2 times the diameter of the peristome, while in the
other specimens it ranges from 2.2 in the smallest, to 2.5 in the next to the
largest. As in all regular Echini, the peristome is relatively larger in young

PSEUDOBOLETIA INDIANA. 345

specimens, and is relatively smallest in the largest. In all the specimens, the
depth of the gill-cuts is one fifth to one sixth the diameter of the peristome.
In coloration none of the specimens from Mauritius show any blotches of
dark color; the spines are greenish at base and more or less pinkish or light
rose-purple distally. The specimens from Zamboanga have the brown blotches
very distinct while the spines show considerable diversity in color; in one
specimen they are green with very decidedly rose-purple tips. In granulata,
there are rather indistinct, large dark blotches, the color including the base of the
spines. Obviously this material is not sufficient to demonstrate whether the
coloration furnishes a specific character in this genus or not. On the other
hand, the size and arrangement of the buccal plates, are constant so far as our
material goes, and Bell's division of maculata from indiana may be accepted at
least until more abundant material shows the separation to be unnatural.
While indiana ranges from Mauritius to the Hawaiian Islands, maculata has as
yet been found only in the East Indian region. Koehler's (1908, Trans. Roy.
Soc. Edinburgh, XLVI, p. 641) statement that the example of maculata in the
British Museum has no indication of locality is unaccountable; Bell (1884, Ann.
Mag. Nat. Hist., (5), XIII, p. 108-111) refers to several specimens and gives the
localities from whence they came.

The three nominal species are distinguished from each other as follows: -

Pore-pairs in arcs of four; actinal spines not banded.

Buccal plates large, nearly or quite in contact; buccal membrane with many other

rather large plates; test without dark blotches indiana.

Buccal plates rather small, widely separated; some parts of test with bases of
accompanying spines, dark colored, forming more or less distinct but indefinite

blotches maculata.

Pore-pairs in arcs of five; actinal spines banded with green and white atlantica.

Pseudoboletia indiana A. Ag.

Toxopneustes indianus Michclin, 1862. Ech. et Stel.: Annexe A, in Maillard's Notes sur Bourbon,

p. 5.

Pseudoboletia indiana A. Agassiz, 1872. Rev. Ech., pt. 1, p. 153.
Psammechinus paucispinus A. Agassiz and Clark, 1907. Bull. M. C. Z., LI, p. 246.

Plate 92, figs. 12-18.

It has become evident on further examination of the Hawaiian material
that Psammechinus paucispinus is simply the youthful form of Pseudoboletia.
The specimens range from 12.5-19 mm. in diameter and all of them have the

346 HAWAIIAN AND OTHER PACIFIC ECHINI.

actinal membrane fully plated. The latter character, taken with the small size,
was misleading and the possibility of their being young Pseudoboletias did
not suggest itself. But the presence of deep actinal cuts for the gills, spines
on the buccal plates, and four pairs of pores in each arc, show their true
position beyond doubt. The buccal plates are large and near together but are
not as nearly in contact as in large specimens. One of the specimens is badly
distorted, ambulacrum II being malformed, sunken, and without pores above
the ambitus. The arrangement of the ocular plates in the other specimens is
interesting as showing that the adult character is not acquired very early in all
cases. In the two largest specimens, ocular I is broadly insert while ocular V
is still exsert, although very nearly in. In the two smallest specimens ocular I
is broadly insert while V is narrowly so. It seems fair to assume that in this
species ocular I becomes insert before the individual is 10 mm. h. d. while ocular
V comes in during the growth in diameter from 10 to 20 or perhaps 25 mm.

The coloration of these specimens is very plain. The test is whitish with a
more or less pronounced green tinge when cleaned. The plated actinostome
is almost pure white. The spines vary from white to deep pink, in four of the
specimens being distinctly pink.

The pedicellarise are remarkably numerous and varied, no less than six
different forms being found on a single specimen. The globiferous occur in two
different forms, a larger (PI. 92, fig. 12) without glands on the stalk and with
valves (PI. 92, fig. 15) .65-.8S mm. long, and a smaller, with glands on the stalks
(PI. 92, 13) and valves only about .20 mm. in length. The ophicephalous pedi-
cellarise are common everywhere but are not peculiar; the valves measure .20-
.40 mm. in length besides the "loop" which may be half as much again. The
tridentaie occur in two very different forms, one (PI. 92, fig. 14) with very slender
valves, about .70 mm. long; the other, with broad valves (PI. 92, fig. 17) .50-
.90 mm. long. The latter are common everywhere but the former are rare and
seem to occur only abactinally. The triphyllous are very small, with valves
only .12-. 14 mm. in length. There are 4 or 5 sphseridia (PI. 92, fig. 18) on the
lower part of each ambulacrum. Calcareous spicules occur in both the pedicels
and the heads of the globiferous pedicellariae. The latter are smaller but are
not otherwise different. The spicules are usually C- or (- shaped but they may
have an irregular branch or two; they are sometimes perfectly bihamate. As
Mortensen makes the form of these spicules one of the chief subfamily characters
in his classification, it seems desirable to call attention to their variability.

The "Albatross" took no adult specimens of Pseudoboletia but the young
ones were taken at the following stations : —

LOXECHINUS. 347

Station 3872. Off Mokuhooniki Islet, Auau Channel, Hawaiian Islands.
Bott. temp. 74.6°. 32-43 fathoms. Yl. s., p., co.

Station 3876. Off Lahaina Light, Maui, H. I. Bott. temp. 74°. 28-43
fathoms. S., g.

Station 4033. Off Diamond Head, Oahu, H. I. Bott. temp. ? 28-29
fathoms. Fne. co. s., for.

Station 4164. Off Modu Manu, H. I. Bott. temp. 78.1°. 40-56 fathoms.
Co. s., p., sh.

Bathymetrical range, 28-56 fathoms. Extremes of temperature, 78.1°-74°.

Five specimens.

PARACENTROTUS.

Mortensen, 1903. "Ingolf" Ech., pt. 1, p. 124, 135.
Type-species, Echinus lividus Lamarck, 1816. Anim. s. Vert., Ill, p. 50.

Although Mortensen does not designate lividus as the type of his genus, it is
without doubt the species he intended as the type, and therefore it is here
accepted as such. The similarity of this genus to Echinus is striking and
Mortensen is undoubtedly right in describing it as a polyporous Echinus. It
appears to be very near the parent stock from which the Strongylocentrotidse
have sprung. There are only three species known, lividus, which is found in the
Mediterranean and in the eastern North Atlantic from the English Channel to
the Azores and Canaries1; Gaimardi (Bl.), which appears to be confined to the
Brazilian coast; and agulhensis Dod., which occurs off South Africa, and in
much deeper water than the others, having been taken by the "Valdivia" in
278 fathoms. The three species may be distinguished from each other by the
following characters : —

Pores-pairs 4 in each arc.

TJnicolor; preserved material whitish or light yellowish agulhensis.

Variegated; preserved material green and brown Gaimardi.

Pore-pairs generally 5; color variable, but dark lividus.

LOXECHINUS.

Desor, 1856. Syn. Ech. Foss., p. 136.
Type-species, Echinus albus Molina, 1782. Saggio St. Nat. Chili, p. 175.

Whether this genus contains more than a single species seems open to grave
doubt, for it is not easy to believe that bullatus is distinct from albus. A series of

1 Bell (1892, Cat. Brit. Ech., p. 158) records this species from Brazil. This is probably due to con-
fusing lividus with Gaimardi, from which it is quite distinct.

348 HAWAIIAN AND OTHER PACIFIC ECHINI.

specimens of albus, collected at various points on the Chilean coast, from Shoal
Bay, Patagonia, northward, chiefly by the "Hassler", shows that that species
is very variable in all those characters by which bullatus is supposed to differ.
But with only a single authenticated specimen of bullatus at hand, its title to
recognition is not clear. It is supposed to be characterized by having all the ocu-
lars exsert but this is true of more than half the specimens of albus, and by
having only 7 or 8 pairs of pores in an arc instead of 9 or 10, but specimens of
albus occur which have only 8. The color of bullatus is also supposed to be less
green and more brown than albus but specimens of the latter are sometimes
more brown than green. In the face of these facts the status of bullatus cannot
be decided. Loxechinus is an interesting genus for several reasons. Its geo-
graphical isolation on the western coast of southern South America is notice-
able. It combines remarkably unspecialized globiferous pedicellaria? with a
very high specialization of the ambulacra, while the character of the abactinal
system is apparently not at all firmly fixed as yet. It reaches an unusually
large size and develops an exceedingly thick test, and finally it is one of the
very few genera of echinoderms which are of use to man, its type species being
the edible "erizo" of Peru and Chile.

gen. nov.

Type-species, Echinus (Toxopneustes) gibbosus L. Agassiz and Desor, 1846. Ann. Sci. Nat., (3), VI,

p. 367.

Plate 95, figs. 16, 17.

It is a remarkable fact that although this genus is nearer to Loxechinus geo-
graphically than to any other genus of the family, it does not seem to be very
near it structurally but differs from it very strikingly. The only known species
is common on the coast of Peru, especially at Payta and it is also known from the
Galapagos Islands. In the very great majority of adult specimens, a parasitic
crab lives in the periproct, which is thereby more or less distorted. Whether
this is the cause of the unusual arrangement of the ocular plates by which ocular
I is never insert unless all the others are also, is not yet definitely determined.
It is unknown how large a proportion of the individuals of a given season reach
adult size without being parasitized and also what the typical arrangement of
the oculars is for such adults.

new, strange + Kfvrpov = prickle.

PACHYCENTROTUS. 349

The arrangement of the pore-pairs in arcs of four is a very obvious difference
between this genus and Loxechinus and shows its relationship to Paracentrotus.
The globiferous pedicellarise are unlike either of those genera but are like those
of Heliocidaris, the valves (PI. 95, fig. 16) having a prominent lateral tooth on
the left side (PI. 95, fig. 1 7).

So far as known this genus is monotypic but attention may be called to the
fact that specimens from Albemarle Island, Galapagos, are very different in
color from those collected at James Island or on the mainland coast. They
seem to be dull red-brown, with no trace of green, so that their general appear-
ance is quite different from typical gibbosus.

PACHYCENTROTUS,1 gen. nov.

Type-speoiea, Sphcerechinus australice A. Agassiz, 1872. Bull. M. C. Z., Ill, p. 55.

Plates 94, figs. 1-6; 98, figs. 5-8.

Although Mortensen had the opportunity to examine a specimen of this
species in the British Museum, taken by the "Challenger" in Bass Strait, he
does not attempt to "decide to which genus and species" it belongs, but simply
points out the most striking feature of its pedicellarise. It appears from other
remarks which he makes that he has seen specimens of S. granularis labelled
australice and it is not strange that this has led him to doubt whether he- has ever
seen the true australice. There is however little doubt that the "Challenger"
specimen he saw in London is australice for a similar small specimen from the
same station in the collection of the M. C. Z. is undoubtedly that species.
The globiferous pedicellariae are, as Mortensen says, like those of Heliocidaris
in having a large lateral tooth near the tip only on the left side but they are quite
characteristic nevertheless in certain details (PI. 94, figs. 1,2}; the stalk about
equals the head (.90-1 mm.) and there is no neck. The tridentate vary greatly
in size, the valves (PL 94, •figs. 3, 4) ranging from .15 to .80 mm. In the ophi-
cephalous, the valves (PI. 94, fig. 5) are about .50 mm. long and in the triphyl-
lous, they (PI. 94, fig. 6) measure .12 mm. The spicules in the pedicels are
bihamate.

This genus is well characterized by its thick, heavily tuberculated test, with
ambulacra having only 4 pore-pairs (rarely 5, PI. 98, figs. 7, 8) in each arc, and
with oculars I and V (rarely IV or II, Plate 98, fig. 6} insert. The sharp, distinct

1 Tra^us = thick, stout + Kcvrpov = prickle.

350 HAWAIIAN AND OTHER PACIFIC ECHINI.

gill-cuts and more or less heavily plated buccal membrane (PI. 98, fig. 5) are
equally important and diagnostic. The short, small spines and the character-
istic globiferous pedicellarise are notable additional features. The genus appears
to be monotypic and is probably confined to the Australian region. The records
from Mauritius and New Zealand are open to very grave doubt.

HELIOCIDARIS.

Agassiz and Desor, 1846. Ann. Sci. Nat., (3), VI, p. 371.

Type-species, Echinus omalostoina Valenciennes, 1846, Voy. "Venus." Zoophytes, PI. 6, fig.2.
= Echinus luberculatus Lamarck, 1816, Anim. s. Vert., Ill, p. 50.

The reasons for considering tuberculatus the type of this genus have already
been given fully (p. 281) and need not be repeated here. As thus understood
Heliocidaris is a fairly homogeneous group of five species confined to the Pacific
Ocean, three of the species being characteristic of the Australian region. A fourth
species, stenopora (new name for mexicana A. Ag.) is very little known; the
specimens upon which it is based are supposed to have come from Lower Cali-
fornia; the species itself is well characterized but that its home is really in the
eastern Pacific needs verification. It cannot continue to bear its original name,
mexicana, for it is entirely distinct from the Heliocidaris mexicana of Louis
Agassiz and Desor described many years earlier. It may be called stenopora,
because of the narrow abactinal poriferous areas. The fifth species, crassi-
spina (A. Ag.) is Japanese and resembles Strongylocentrotus nudus (A. Ag.)
quite closely, but can always be distinguished by the number of pore-pairs in
an arc. It appears that crassispina is quite distinct from E. tuberculatus Lamk.,
and that Anthocidaris homalostoma Liitken, and Toxocidaris purpurea von Mart,
are synonyms of crassispina. It also includes Toxocidaris globulosa A. Ag., as
determined from examination of one of the type-specimens in the U. S. National
Museum, kindly loaned through Mr. Austin H. Clark. Of the three Australian
species of Heliocidaris, erythrogramma (Val.) and tuberculata (Lamk.) are quite
distinct but it would not be surprising to have it demonstrated that armigera
(A. Ag.) is simply a form of erythrogramma. Without intermediate specimens,
the two are retained for the present. The five species of Heliocidaris may be
distinguished as follows: —

Pore-pairs 7 or 8 in each arc, rarely 9.

Primary spines short (20 mm. ± ) and very stout (2 or 3 mm. in diameter) . . . armigera.
Primary spines longer and much more slender.

HELIOCIDARIS STENOPORA. 351

Actinostome .30-.35 h. d.; pore-pairs mostly in arcs of 7, of which only the

outer 3 or 4 form more or less vertical series above the ambitus.
Primary spines very dark (blackish, purplish, or brownish); test very flat

actinally with poriferous areas more or less petaloid crassispina.

Primary spines dull reddish or greenish, often with more or less of distal
half, violet; test not flattened actinally and poriferous areas not peta-
loid erythrogramma.

Actinostome .35-.40 h. d.; pore-pairs mostly in arcs of 8 (occasionally 9), of
which the outer 4-6 are in more or less vertical series above the ambitus, so

that the poriferous areas are very narrow stenopora.

Pore-pairs 9 or often 10 in each arc tuberculata.

Heliocidaris stenopora,1 nom. nov.

Toxocidaris mexicana A. Agassiz, 1863. Bull. M. C. Z., I, p. 22. NON Heliocidaris mexicana L.
Agassiz and Desor, 1846, Ann. Sci. Nat., (3), VI, p. 372.

' Plates 95, figs. 18-22; 104, figs. 1-3; 110, figs. 4, 5.

As this species appears to be rather rare, few specimens having found their
way into museums, it seems desirable to give figures of the specimen in the
M. C. Z. collection, which is almost certainly the type. Its measurements cor-
respond with those in the "Revision." All four kinds of the pedicellarise figured
were found on this specimen. The globiferous have valves about .45 mm. long
(PI. 95, fig. 18) with a prominent lateral tooth on the left side of the blade near
the tip (PI. 95, fig. 19). The ophicephalous have valves (PI. 95, fig. 20) of about
the same length, counting the "loop"; they are not constricted at the base of
the blade. The tridentate are also of about the same size, with rather wide,
somewhat curved blades (PI. 95, fig. 21). These three kinds of pedicellariae
are covered with a dark tissue, which makes them all look much alike, but the
ophicephalous and tridentate have long necks, which seem to be wanting in the
globiferous. The triphyllous also have long necks, and valves (PI. 95, fig. 22)
about .15 mm. long; the blade is the same width as the base, little constricted
and very square cut at the tip.

This species appears to be very well characterized for when compared with
tuberculata or crassispina, or with the west coast species of Strongylocentrotus
(franciscanus and purpuratus), the differences are so obvious, there can be
no question of assigning it to any of those species. The spines of the specimen
at hand are all broken, but the primaries were evidently very stout; the remain-
ing pieces are bright violet of a light shade. This specimen is supposed to have

1 orevoiropos = with a narrow way.

352 HAWAIIAN AND OTHER PACIFIC ECHINI.

been taken at Cape St. Lucas, Lower California, by John Xantus, about I860,
but if the locality is correct, it is remarkable that the explorations of the "Alba-
tross" in the same region have failed to bring other specimens to light.

STRONGYLOCENTROTUS.

Brandt, 1835. Prodrom. desc. Anim., p. 263 (or 63).

Type-species, Strongylocentrotus chlorocentrotus Brandt, 1. c., p. 264 (or 64) = Echinus drobachiensis O. F.
Muller, 1776. Prodrom. Zool. Dan., p. 235.

So homogeneous is this genus and so completely do the forms intergrade that
it is almost wholly a matter of opinion as to what species are to be considered
valid. It seems nearly certain that the widely distributed drobachiensis is the
parent stock from which the others have sprung, but how many species are at
the present time fully differentiated from this stock is very hard to decide. Of
course a typical franciscanus is very easily distinguished and that is certainly
a distinct species, yet it appears to intergrade with purpuratus which in turn
connects very clearly with drobachiensis. In the European seas, granularis
(Lamk.) has become quite an easily recognized species particularly when its
characteristic features are fully developed. The Japanese and Aleutian forms
are not so well defined and although the attempt to distinguish, from those
seas, pukherrimus A. Ag., depressus A. Ag., intermedius A. Ag., pulchellus A. Ag.
and Cl., echinoides A. Ag. and Cl., polyacanthus A. Ag. and Cl., and nudus A. Ag.,
typical examples of which are easily recognizable, is here made, it is probable
that at least two or three of them are not really specifically distinct from droba-
chiensis, and ought not to be honored with names. There is not the slightest
ground for doubting that chlorocentrotus Brandt is a synonym of drobachiensis.
Alaskan specimens show very great diversity in the length of the spines as well
as in coloration. The name carnosus A. Ag. might well be revived for the very
handsome form from northeastern Asia, but as the coloration is not always
striking and seems to intergrade with that of drobachiensis, and as any other
distinctive character is not apparent it is better to let the case stand as it is,
until some zoologist can make a special study, in the field, of the North Pacific
forms of Strongylocentrotus. The form which Doderlein (1906, Zool. Anz.,
XXX, p. 517) described from Sakhalin as a variety of drobachiensis under the name
sachalinica, seems to be very well characterized by the remarkably small number
of coronal plates and although specimens have not been seen, it is here raised to
specific rank. His proposed species hokkaidensis is however nothing more or less

STRONGYLOCENTROTUS. 353

than nudus. The recently described fragilis Jackson from off the coast of Cali-
fornia, although obviously related to drobachiensis, seems to be fully entitled
to recognition.

The following table will show the most important characters for distinguish-
ing the species herein recognized: —

Pore-pairs 4-7 in each arc, only very exceptionally 8.

Poriferous areas very broad, with pore-pairs in arcs of 4 which are nearly hori-
zontal at ambitus, with pore-pairs well separated so that they form 4 vertical

series, the two outer sometimes crowded together pulcherrimus.

Poriferous areas not as above.

Gill-cuts deep and sharply defined.

Test not depressed, v. d. rarely less than .50 h. d. and usually more than

.66 h. d. ; actinal surface not flat granularis.

Test depressed, v. d. usually less than .4.5 h. d.; actinal surface very flat depressus.
Gill-cuts neither deep nor sharply defined.

Pore-pairs 4-6, sometimes 7; abactinal primary tubercles small; primary

spines short or moderately long; color rarely very dark.
Test very thin and fragile, depressed, v. d. less than .50, sometimes
only .33 h. d.; auricles high, slender, racquet-shaped; spines ex-
ceedingly slender and fragile, bright reddish orange at base, with

more or less of tip whitish fragilis.

Test and spines not as above.

Coronal plates very few; 11 interambulacral plates in each col-
umn in a specimen 30 mm. h. d., 15 in 60 mm. specimen . . sachalinicus.
Coronal plates much more numerous; 14 interambulacral plates
in a column in 17 mm. specimen, 28 in a 56 mm. specimen.
Pore-pairs typically 5, sometimes 4, rarely 6.

Poriferous area much wider than half interporiferous,
the arcs nearly horizontal at ambitus, with pore-pairs
tending to form three vertical series, inner with two,
median with one and outer with two (rarely 3) pore-
pairs; primary spines greenish or reddish or both;
valves of globiferous pedicellarise with terminal

tooth not half as long as blade intermedius.

Poriferous area about equal to half of interporiferous,
the arcs very oblique at ambitus, the pore-pairs,
forming only an inner series of two and an outer series
of three; primary spines light violet tipped with
white, the actinal ones often faintly banded; valves
of globiferous pedicellariEe with a short blade and

terminal tooth nearly as long pulchelluf;.

Pore-pairs typically 6, often 7, very rarely 8.

Primaries moderate, 10-15 mm. long or more; inter-
ambulacral plates 18-23 in each column; color
not uniform dull rose-purple.
Pore-pairs 6 in an arc, in the mid-zone, occasion-
ally 7; arcs horizontally oblique; abactinal

354 HAWAIIAN AND OTHER PACIFIC ECHINI.

interambulacral areas well covered by secondary
tubercles with relatively few miliaries; color
more or less greenish with or without a violet
tinge, passing on the one hand into a violet test
with deep green spines, and on the other into
dull yellowish or light brown, with little trace of

either green or violet drobachiensis.

Pore-pairs 7 in an arc, in the mid-zone, occasionally
6; arcs vertically oblique; abactinal inter-
ambulacral areas with few secondaries but
closely covered with miliary tubercles, except
along the bare median line; color more or less
reddish white, darkest on abactinal interam-
bulacral areas, which may be deep reddish
purple; primaries light red or light green or both echinoides.
Primaries numerous, very short, 6-8 mm. long, scarcely
distinguishable from secondaries; interambulacral
plates 25 in each column; color uniform dull rose-
purple polyacattthus.

Pore-pairs 6 or commonly 7; abactinal primary tubercles large and
conspicuous; primary spines long; color usually dark brown or

purplish nvdus.

Pore-pairs 8-10 in each arc.

Pore-pairs 8; primary spines short; primary tubercles not conspicuous; globif-

erous pedicellaria3 common; color, when adult, purple purpuralus.

Pore-pairs 9-10; primary spines long; primary tubercles large, in 6 conspicuous
series in each interambulacrum ; globiferous pedicellariae usually very few;
color commonly red-brown, rarely purple franciscanus.

Strongylocentrotus fragilis Jackson.

Strongylocentrotus fragilis Jackson, 1912. Mem. Boston Soc. Nat. Hist., VII, p. 128.

Plates 94, figs. 28, 29; 113, figs. 3-6.

As Jackson gives only a brief description of this fine species and leaves me
the privilege of figuring it, photographs of the type and of a smaller specimen
with spines are given, together with some notes on the structural characters.
A large series of specimens at hand ranges from 9 to 90 mm. in diameter; the
largest is 40 mm. high. A specimen 55 mm. h. d. is 26.5 mm. high, while another
38 mm. h. d. is only 12 mm. high. But these represent the two extremes and
the great majority of specimens are about two fifths as high as they are across.
The test is very thin and fragile, and the auricles are so high and slender, and
expand at the tip so much like a racquet, that they make a good specific
character. There are 5 (rarely 6) pairs of very large pores in each arc, and the

STRONGYLOCENTROTUS FRAGILIS. 355

arcs themselves approach the horizontal at ambitus, yet the poriferous areas
and the ambulacra are rather narrow. The actinostome is more or less deeply
sunken and measures about .30 h. d. in diameter. The actinostomal membrane
is almost wholly free from calcareous plates, excepting the ten that make up the
buccal ring. The abactinal system is about .20 h. d. across. Oculars I and V
are broadly insert in 55-60% of the specimens, but in a number, V is exsert and
in several none of the oculars reach the periproct. The primary spines are 15-
20 mm. long and are exceedingly slender and brittle, as are the secondaries and
miliaries. The color of the test ranges from orange-red to deep purple, but the
spines are less varied, orange-red at base becoming yellowish or whitish at tip.
The pedicellariae (PL 94, fig. 28, 29), sphajridia, and spicules do not appear to be
in any way different from those of drobachiensis.

The "Albatross" took this species at the following stations off the west coast
of the United States and Canada : —

Station 2839. Off Santa Barbara Islands, California, 33° 8' N., 118° 40' W.
Bott. temp. 41.4°. 414 fathoms. Gy. s.

Station 2861. North of Vancouver Island, 51° 14' N., 129° 50' W. Bott.
temp. 42.6°. 204 fathoms. There is some doubt as to whether the specimens
really came from this station.

Station 2886. Off Oregon, 43° 59' N., 124° 56' 30" W. Bott. temp. 48.1°.
50 fathoms. Rky.

Station 2890. Off Oregon, 43° 46' N., 124° 57' W. Bott. temp. 42.2°.
277 fathoms. Gy. s.

Station 2896. Off southern California, 33° 55' 30" N., 120° 28' W. Bott.
temp. 42.8°. 376 fathoms. Yl. m.

Station 2925. Off southern California, 32° 32' 30" N., 117° 24' W. Bott.
temp. 42.9°. 339 fathoms. M.

Station 2927. Off southern California, 32° 43' N., 117° 51' W. Bott. temp.
43.3°. 313 fathoms. Gn. m.

Station 2928. Off southern California, 32° 47' 30" N., 118° 10' W. Bott.
temp. 41°. 417 fathoms. Bk. s., g.

Station 2935. Off southern California, 32° 44' 30" N., 117° 23' W. Bott.
temp. 49.2°. 124 fathoms. Fne. gy. s.

Station 2936. Off southern California, 32° 49' N., 117° 27' 30" W. Bott.
temp. 49°. 359 fathoms. M.

Station 2946. Off southern California, 33° 58' N., 119° 30' 45" W. Bott.
temp. 56.5°. 150 fathoms. Crs. gy. s.

356 HAWAIIAN AND OTHER PACIFIC ECHINI.

Station 2948. Off southern California, 33° 55' 30" N., 119° 41' 30" W.
Bott. temp. ? 266 fathoms. Gy. s., g., brk. sh.

Station 2951. Off southern California, 33° 55' 30" N., 119° 55' W. Bott.
temp. ? 48 fathoms. Fne. gy. s.

Station 3051. Off Oregon, 43° 59' 15" N., 124° 58' 30" W. Bott. temp. ?
59 fathoms. Co., brk. sh., rky.

Station 3053. Off Oregon, 44° 4' 30" N., 124° 50' W. Bott. temp. 47.3°.
64 fathoms. Co., brk. sh., rky.

Station 3076. Off Washington, 47° 46' N., 125° 10' W. Bott. temp. 43.4°.
178 fathoms. Gn. m.

Bathymetrical range, 48-417 fathoms. Extremes of temperature, 56.5°-
41°.

One hundred and twenty-one specimens.

Strongylocentrotus pulchellus A. Ag. and Cl.

Strongylocentrotus pulchellus A. Agassiz and Clark, 1907. Bull. M. C. Z., LI, p. 121.

Plates 94, figs. 2^-27; 98, figs. 1, 2; 111, figs. 9, 10.

There are only two specimens of this species at hand and one of those is so
small, only 8 mm. h. d., and so obviously immature, that it hardly reveals any
distinctive characters. The other (from St. 5003) which may be considered the
type (PI. Ill, figs. 9, 10) is 17 mm. in diameter, and about 8 mm. high. The
actinostome is about 8 mm. in diameter and the abactinal system half as much.
There are 13 or 14 interambulacral plates in each column and 16 or 17 ambulacral.
The ambulacra are only a little narrower than the interambulacra but the porifer-
ous zones are rather narrow. There are five pairs of pores in each arc from be-
low the ambitus to the ocular plate. The pairs, in the mid-zone, are in oblique,
somewhat curved arcs, divided by a secondary tubercle into an inner group of
two and an outer, lower group of three pairs. The vertical series of secondary
tubercles thus divides the poriferous area into an inner and an outer band.
Each ambulacral plate carries one primary tubercle, which is relatively rather
large, and in the mid-zone a secondary tubercle at the inner end, in addition
to the secondary tubercle in the poriferous zone. There are also smaller sec-
ondaries and miliaries. The tuberculation of the interambulacral plates is
not essentially different but the secondaries are a trifle larger and those on the
outer ends of the plates are as close as possible to the poriferous area. Oculars

STRONGYLOCENTROTUS PULCHELLUS. 357

I and V are broadly insert (PI. 98, fig. 2} and the large suranal carries a well-
developed tubercle. The anus is decidedly excentric near ocular I. The madre-
poric genital is somewhat swollen and is much larger than any of the others.
All five carry tubercles, 2 or 3 on the proximal margin of each plate. — The
buccal plates (PI. 98, fig. 1) are of moderate size and are well spaced. Proximal
to them the membrane is thickly covered with small plates but distally these
become more and more scattered and a considerable part of the membrane is
perfectly bare. The gill-cuts are well defined but are neither narrow nor deep.
The primary spines are about 3 mm. long, stouter, and distinctly longer than
the secondaries. The test is very light purplish, noticeably darker abactinally,
particularly on the median interambulacral areas; on and around the abactinal
system there is a very evident greenish tinge. The primary spines are light
purple rather abruptly tipped with whitish. The smaller spines are much lighter.

In the smaller specimen, there are only 9 interambulacral and 10 ambulacral
plates in each column. The arcs of pores contain only 4 pairs from the ambitus
adorally but 5 aborally. The arcs are not so clearly divided into an inner pair
and an outer trio by a small tubercle as in the larger specimen. Ocular V is
wholly exsert and a genital pore is visible only in genital 1. The primary spines
are purplish only at base and the terminal part is light greenish.

The pedicellarise of pulchellus are quite characteristic, particularly the globif-
erous, the valves of which (PI. 94, figs. 24, 25} are only about .50 mm. long;
the base of the valves is .27 or .28 mm. wide and the terminal tooth is about
.20 mm. long. These pedicellarise are common and the stalks are about twice as
long as the head. The ophicephalous pedicellarise have valves about .20-.35 mm.
in length, besides the loop which is .03-.12 mm. more; the valves are somewhat
constricted at base of blade, but not greatly so. The tridentate seem to be quite
rare; they have straight, narrow valves (PI. 94, fig. 26) only about half a milli-
meter long, and more or less in contact throughout. The triphijllous have the
valves about .12 mm. long and the truncate tip of the blade is about .09 or .10
mm. wide. The sphseridia (PI. 94, fig. 27) are fairly numerous and seem to be
rather elongated; they measure about .20 mm. in length but the width is scarcely
half as much. Spicules are scarce in the pedicels; they are C-shaped with the
ends branched.

While this species is certainly nearly related to intermedius and drobachiensis,
the globiferous pedicellarise are sufficient to distinguish it from those or any
other species.

It was taken by the "Albatross" only at the two following stations: —

358 HAWAIIAN AND OTHER PACIFIC ECHINI.

Station 4794. Off east coast of Kamchatka, 52° 47' 20" N., 158° 44' 30" E.
Bott. temp. ? 58-69 fathoms. S., p.

Station 5003. Off southwestern coast of Saghalin Island, 47° 32' 30" N.,
141° 45' E. Bott. temp. 42.4°. 35-38 fathoms. Fne. gy. s., gn. m.

Two specimens.

Strongylocentrotus drobachiensis A. Ag.

Echinus drobachiensis O. F. Mtiller, 1776. Prodrom. Zool. Dan., p. 235.
Strongylocentrotus drobachiensis A. Agassiz, 1872. Rev. Ech., pt. 1, p. 162.

The large series of specimens brought in by the "Albatross" taken in con-
nection with the hundreds of specimens in the M. C. Z. collection make the
true limits of this widespread and variable species more uncertain than ever.
Although recognizing the apparent inconsistency of considering chlorocentrotus
and carnosus synonyms of drobachiensis, while retaining intermedius, sachalmicus,
poly acanthus, and echinoides as valid species, it is not easy to see any better solu-
tion of the difficulty. In length of primary spines, in abundance of tubercles,
and in color, drobachiensis is exceedingly variable. In some places, Puget
Sound for example, the deeper-water specimens have long primaries and are
light colored, while those from shallow water have short thick spines and a
much darker coloration. The largest and finest specimen seen, is from "Alba-
tross" Station 4302; it is 90 mm. in diameter and is very light colored. In its
coloration indeed it approaches echinoides, but it is distinctly greener. The
test is very pale brown or dirty cream-color and the spines are light green abac-
tinally but become cream-color below. The longest primaries are nearly 22 mm.
long. The arcs of pores above the ambitus contain 6 pairs, very rarely 7. (I
have never seen 8 in drobachiensis although that number is recorded and there
is no good reason why it may not occur occasionally as an extremely progressive
variation). Oculars I and V are broadly insert but the other three are quite as
fully exsert. The suranal plate is large and hence easily distinguished. Pedi-
cellarise are exceedingly abundant.

The "Albatross" specimens were taken at the following places : —

Station 2878. Off Washington, 48° 37' N., 125° 32' W. Bott. temp. 45.5°.
66 fathoms. P.

Station 4200. Off Fort Rupert, Vancouver Island, British Columbia. Bott.
temp. 46.8°. 68 fathoms. Gn. m., s., sponge.

STRONGYLOCENTROTUS DROBACHIENSIS. 359

Station 4205. Admiralty Inlet, near Port Townsend, Washington. Bott.
temp. 50.8°. 15-26 fathoms. R., sh.

Station 4208. Admiralty Inlet, near Port Townsend, Washington. Bott.
temp. 50.5°. 83-99 fathoms. Rky.

Station 4223. Boca de Quadra, southeastern Alaska. Bott. temp. 44.6°.
48-57 fathoms. Sft. gn. m.

Station 4245. Kasaan Bay, Prince of Wales Island, southeastern Alaska.
Bott. temp. 48.9°. 95-98 fathoms. Dk. gn. m., s., sh., r.

Station 4262. Off Point Wimbleton, Dundas Bay, Icy Strait, Alaska. Bott.
temp. ? 9 fathoms. Crs. s., rky.

Station 4270. Afognak Bay, Afognak Island, Alaska. Bott. temp. ?
14-19 fathoms. Hrd. gy. s., r.

Station 4278. Alitak Bay, Kadiak Island, Alaska. Bott. temp. ? 27-29
fathoms. Dk. gn. m.

Station 4280. Chignik Bay, Alaska. Bott. temp. ? 43 fathoms. Gn.
m., fne. s.

Station 4285. Chignik Bay, Alaska. Bott. temp. ? 31-59 fathoms. Gy.
s., brk. sh.

Station 4287. Uyak Bay, Kadiak Island, Alaska. Bott. temp. 43 ? 66-67
fathoms. Gy. m.

Station 4289. Uyak Bay, Kadiak Island, Alaska. Bott. temp. 42.2. 74-80
fathoms. Gy. m.

Station 4302. Off Shakan, Sumner Strait, southeastern Alaska. Bott.
temp. 44.2°. 169-212 fathoms. Bu. m.

Port Townsend, Washington.

Union Bay, British Columbia.

Dutch Harbor, Unalaska.

Attu, Aleutian Islands.

Agattu, Aleutian Islands.

Medni, Komandorski Islands.

Bering, Komandorski Islands.

Petropavlovsk, Siberia.

Bathymetrical range, shore — 212 fathoms. Extremes of temperature,
50.8°-42.2°.

One hundred and thirteen specimens.

360 HAWAIIAN AND OTHER PACIFIC ECHINI.

Strongylocentrotus echinoides A. Ag. and Cl.

Strongylocentrotus echinoides A. Agassiz and Clark, 1907. Bull. M. C. Z., LI, p. 122.

Plates 94, figs. 13-16; 113, fig. 1.

This species is so near to drobachiensis that its separation as a distinct species
is open to grave doubt, yet when specimens of the same size are compared echi-
noides shows certain features that lead one to consider it such. The light color of
the test is very noticeable, being more or less reddish white, but specimens of
drobachiensis of essentially the same color have been seen. The light green and
reddish shades of the spines are quite different from most specimens of drobachien-
sis, but this difference is by no means constant. The difference in the granulation
of the plates is noticeable, for while in drobachiensis, an abactinal interambula-
cral plate may have 50-75 small secondary and miliary tubercles, in echinoides
there are twice as many. In the ambulacra, there is an important difference,
in that the arcs of pores are less horizontal, the ambulacral plates being corre-
spondingly higher and therefore fewer. Thus, in a drobachiensis, with about
20 interambulacral plates in each column, there are 35 ambulacral plates, but
in an echinoides with 20 interambulacrals, there are only 27 or 28 ambulacrals.
The arcs of pores in drobachiensis, in the mid-zone, have 6 pairs, occasionally 7,
while in echinoides they have 7, occasionally 6. The poriferous areas are dis-
tinctly narrower in echinoides. The primary spines of echinoides are longer and
more slender than in most specimens of drobachiensis, but specimens of the latter
from Puget Sound cannot be distinguished from echinoides in this respect.

The pedicellarise of the two species are indistinguishable so far as any char-
acters of importance are concerned. In echinoides, the globiferous pedicellarise
have valves .40-1.00 mm. in length (PI. 94, figs. 15, 16); the length of the blade
is about equal to the width of the base while the length of the terminal tooth is
about one half as much. In the tridentate (PI. 94, fig. 13) which vary enormously
in size, the valves are .20-2.00 mm. long; in the small ones the valves are narrow
and meet for most of their length but in the large ones, the blades are somewhat
expanded at the tip and meet only there. In the ophicephalous, the valves
(PI. 94, fig. 14) are about half a millimeter long, besides the loop, and are dis-
tinctly constricted at base of blade. In the triphyllous, the valves are about
.15 mm. long. All these different kinds of pedicellarise are abundant, but the
more globiferous there are, the fewer the tridentate and vice-versa. All forms
have a neck but on the globiferous it is very short. The spicules show great

STRONGYLOCENTROTUS ECHINOIDES. 361

diversity; they may be simply bihamate, or rods, more or less curved, more or
less flattened and more or less branched at the tips..

The geographical distribution of echinoides apparently coincides with that of
drobachiensis, in the North Pacific, as will be seen from the table of stations given
below. When described it was supposed to be confined to the western side of
the ocean but a specimen from Alaska and another from off Oregon is at hand.
The former is quite young but the latter is an adult in fine condition and very
typical of the species, and there seems to be no reason for doubting the correct-
ness of the label. The following list shows the conditions under which echinoides
lives: —

Station 3053. Off Oregon, 44° 4' 30" N., 124° 50' W. Bott. temp. 47.3°.
64 fathoms. Co., brk. sh., rky.

Station 4253. Off Thistle Ledge, Stephens Passage, Alaska. Bott. temp.
40.9°. 131-188 fathoms. R., brk. sh.

Station 4777. Petrel Bank, Bering Sea, 52° 11' N., 179° 49' E. Bott. temp.?
43-52 fathoms. Fne. g.

Station 4778. Petrel Bank, Bering Sea, 52° 12' N., 179° 52' E. Bott. temp.?
33-43 fathoms. Fne. blk. g.

Station 4779. Petrel Bank, Bering Sea, 52° 11' N., 179° 57' W. Bott, temp.?
54-56 fathoms. Brk. sh.. p., s.

Station 4782. Off East Cape, Attu Island, Aleutians, 52° 55' N., 173° 27' E.
Bott. temp.? 57-59 fathoms. R., g.

Station 4784. Off East Cape, Attu Island, Aleutians, 52° 55' 40" N.,
173° 26' E. Bott. temp.? 135 fathoms. Crs. p.

Station 4786. Between Medni and Bering Islands, Komandorskis, 54° 51'
30" N., 167° 14' E. Bott. temp.? 54 fathoms. Gn. s.

Station 4787. Between Medni and Bering Islands, Komandorskis, 54° 50'
50" N., 167° 13' 30" E. Bott, temp.? 54-57 fathoms. Gn. s.

Station 4788. Between Medni and Bering Islands, Komandorskis, 54° 50'
24" N., 167° 13' E. Bott, temp.? 56-57 fathoms. Gn. s.

Station 4789. Between Medni and Bering Islands, Komandorskis, 54° 49'
45" N., 167° 12' 30" E. Bott, temp.? 56 fathoms. Gn. s.

Station 4790. Between Medni and Bering Islands, Komandorskis, 54° 38'
45" N., 167° II' 45" E. Bott. temp.? 64 fathoms. P.

Station 4791. Between Medni and Bering Islands, Komandorskis, 54° 36'
15" N., 166° 58' 15" E. Bott. temp.? 72-76 fathoms. Rky.

Station 4792. Between Medni and Bering Islands, Komandorskis, 54° 36'
15" N., 166° 57' 15" E. Bott. temp.? 72 fathoms. P.

362 HAWAIIAN AND OTHER PACIFIC ECHINI.

Station 4794. Off eastern coast of Kamchatka, 52° 47' 20" N., 158° 44' 30"
E. Bott. temp.? 58-69 fathoms. S., p.

Station 4795. Off eastern coast of Kamchatka, 52° 46' 50" N., 158° 44' 30"
E. Bott. temp.? 48-69 fathoms. Gn. s., p.

Station 4796. Off eastern coast of Kamchatka, 52° 47' N., 158° 43' E.
Bott. temp.? 48 fathoms. S., p., sh.

Station 4804. Off Simushir Island, 46° 42' N., 151° 47' E. Bott. temp.
35.9°? 229 fathoms. Crs. p., bk. s.

Station 4810. Between Hakodate and Sado Island, Japan, 41° 17' 20" N.,
140° 7' E. Bott. temp. 44.7°. 90-195 fathoms. Fne. gy. s.

Station 4822. Between Nanao and Tsuruga, Hondo, Japan, 37° 8' 10" N.,
137° 8' E. Bott. temp. 39.4°. 130 fathoms. Gn. m.

Station 4982. Between Hakodate and Otaru, Hokkaido, Japan, 43° N.,
140° 10' 30" E. Bott. temp. 32.7°. 390-428 fathoms. Gn. m.

Station 4987. Between Hakodate and Otaru, Hokkaido, Japan, 43° 19' 20"
N., 140° 17' E. Bott. temp. 44.8°. 59 fathoms. Rky.

Station 4993. Between Otaru, Hokkaido, and Korsakov, Saghalin, 45° 25'
30" N., 140° 53' E. Bott. temp. 35.1°. 142 fathoms. Gy. m., s., g.

Station 4996. Between Otaru, Hokkaido, and Korsakov, Saghalin, 45° 35'
N., 140° 55' E. Bott. temp. 43.4°. 86 fathoms. Bk. s., p.

Station 5016. Off eastern coast, southern end of Saghalin, 46° 44' 30" N.,
143° 45' E. Bott. temp. 29.8°. 64 fathoms. Br. m.. fne. bk. s., r., co.

Station 5041. Off southern coast of Hokkaido, Japan, 42° 16' 30" N.,
142° 4' E. Bott. temp. 49.8°-41.1°. 61-140 fathoms. Br. m., fne. bk. s., co., s.

Station 5048. Between Hakodate and Yokohama, Japan, 38° 9' 24" N.,
141° 52' 30" E. Bott. temp. 40.7°. 129 fathoms. Dk. gy. s., brk. sh.

Station 5049. Between Hakodate and Yokohama, Japan, 38° 12' N.,
142° 2' E. Bott. temp. 37.8°. 182 fathoms. Dk. gy. s., brk. sh.

Bathymetrical range, 33^28 fathoms. Extremes of temperature, 49.8°-
29.8°.

One hundred and sixty-four specimens.

Strongylocentrotus polyacanthus A. Ag. and Cl.

Strongylocentrotus polyacanthus A. Agassiz and Clark, 1907. Bull. M. C. Z., LI, p. 123.

Plates 94, figs. 30-33; 113, fig. S.

This species is so near to drobachiensis, that it would be quite superfluous
to give a detailed description of the single specimen upon which it is based.

STRONGYLOCENTROTUS NUDUS. 363

The test is 73 mm. in diameter and 40 mm. high. When compared with a
specimen of drobachiensis of the same size, the only noticeable difference, aside
from the characteristic coloration, is the smaller size and larger number of the
primary tubercles, and spines. There are 25 or 26 interambulacral plates, and
39 or 40 ambulacral, in each column, while in a drobachiensis, 75 mm. h. d. and
42 mm. high, the numbers are 22 and 38 respectively. The primary spines of
the latter specimen are 11 or 12 mm. long, while those of polyacanthus are only
6-8 mm. The arc of pore-pairs are essentially as in drobachiensis, oblique
approaching the horizontal, and with only six pairs in each arc. The test and
spines are uniformly dull rose-purple, without a trace of green.

The pedicellarise show some characters which may prove to be of importance.
The globiferous are all small and the valves (PI. 94, figs. 31, 32) have a short
terminal tooth; they are about .50 mm. long with the terminal tooth scarcely
one fourth as much. The ophicephalous are peculiar in that the valves (PI. 94,
fig. 33) which measure .40-.60 mm. in length, are not at all constricted but are
bluntly triangular in form; the articular loops are also noticeably low and
flattened. The tridentate pedicellarise (PI. 94, fig. 30) are all rather long and
narrow; none of those, so common in drobachiensis, with broad curved blades,
were found; the valves are .40-1.60 mm. in length and are more or less fully
in contact throughout. The triphyllous are very common and show some
diversity in the relative width of the blade; the valves are about .15 mm. long.
The calcareous spicules are either bihamate or curved rods with branched ends.

The single specimen of this species was taken by the collectors of the " Alba-
tross" on the shores of Milne Bay, Simushir Island, Kuril Group, Northern
Japan, June 23, 1906.

Strongylocentrotus nudus A. Ag.

Toxocidaris nuda A. Agassiz, 1863. Proc. Acad. Nat. Sci. Philadelphia, p. 356.
Strongylocentrotus nudus A. Agassiz, 1872. Rev. Ech., pt. 1, p. 165.
Strongylocentrotus hokkaidensis Doderlein, 1906. Zool. Anz., XXX, p. 518.
Strongylocentrotus tuberculatus A. Agassiz and Clark, 1907. Bull. M. C. Z., LI, p. 122. NON
Echinus tuberculatus Lam'k.

Plate 94, figs. 17-23.

In view of the fact that this species is now known to be one of the common
sea-urchins of northern Japan but does not appear to occur in the warmer south-
ern seas of that country, it seems very probable that the original specimens,
brought home by the United States Exploring Expedition and described in
1863, were all from Hakodate and not in part from Hilo, Hawaii, as they were

364 HAWAIIAN AND OTHER PACIFIC ECHINI.

labelled. Original specimens in the M. C. Z. collection from "Hilo. W. Stimp-
son" and "N. E. end of Niphon. W. Stimpson" are indistinguishable. This
view is confirmed by the fact that the "Albatross" failed to collect any Strongy-
locentrotus in the Hawaiian Islands.

The pedicellarise of this species show some interesting features. The globif-
erous are not rare in very young specimens but are less common in those half
grown and are almost entirely wanting in adults. The reverse of this is true
of the tridentates, which are very common in adults and occur in two or three
different forms, but are less common in half grown specimens, and are rare in very
young individuals. The globiferous appear to be of two different sizes, one
with valves .30 mm. long and the other with valves about .80 mm. (PI. 94, figs.
18, 19). It is possible these forms intergrade but intermediate sizes do not
occur in the M. C. Z. material. The tridentate occur in such a variety of forms
that it is not easy to group them exactly, but three rather distinct kinds may be
recognized. In one of these the valves (PI. 94, fig. 21) are from .20 to .50mm.
in length, rather wide and flat, bluntly rounded and in contact for most of their
length; in a second, the valves (PI. 94, figs. 22, 23) are very long (1.5-2 mm.)
and narrow and in contact only for more or less of their distal half; in the third,
the valves are about 1.5 mm. long, with broad, curved blades in contact at their
tips. The ophicephalous occur in two forms, one with valves .20 mm. besides
the loop and the other about .50 mm. also besides the loop; the valves are notice-
ably constricted. The triphyllous (PI. 94, fig. 20) are neither rare nor peculiar.
The spicules are not common; they are curved rods, branched at the ends.

The specimens taken by the "Albatross" range from 23 to 75 mm. in diameter.
The smallest is dull purplish with an evident greenish tinge abactinally, while
the spines are more or less greenish. There are six pairs of pores in each arc
in the mid-zone but the uppermost arcs contain seven as a rule. In the larger
specimens, the color is uniformly dark reddish brown with no trace of green. In
the largest specimen, there are six pairs of pores in each arc at the ambitus,
but just above, there are a few arcs with seven; then follow a few with six and
the uppermost have only five, indicating a senescent change.

The "Albatross" collected this species at the following stations: —

Station 4807. Between Hakodate and Sado Island, Japan, 41° 36' 12" N.,
140° 36' E. Bott. temp. ? 44-47 fathoms. Sh., crs. g.

Station 5018. Off Cape Tonin, Saghalin Island, Japan, 46° 41' 30" N.,
143° 57' 40" E. Bott. temp 30.4°. 100 fathoms. Br. m., bk. s., p.
Hakodate.
Seven specimens.

ECHINOMETRID^:. 365

ECHINOMETRIDJE Gray.

GENERAL CONSIDERATIONS.

There can be little question that this family includes the most highly special-
ized of the recent regular Echini, for the elongation of one axis, when combined
with highly developed ambulacra, indicates an unusual complexity of structure.
And yet in the characters of the abactinal system and the globiferous pedicel-
lariae, the more specialized Echinidse, such as Tripneustes, are apparently more
advanced than any of the Echinometridae, and it is therefore merely a matter
of opinion whether Tripneustes or Heterocentrotus is considered the "highest"
of the regular Echini. They represent different lines of development and cannot
properly be considered competitors. Since, however, the ambulacra are held
to be the most important structures of an echinoid and the multiplication
of ambulacral elements the best evidence of increasing specialization, we do
not hesitate to rank the Echinometridse as the last and highest of the families
of Regulares. In addition to the characters already mentioned, Jackson (1912,
Mem. Boston Soc. Nat. Hist., VII) has brought out the interesting fact that
in the Echinometridse, ocular V is the first to become insert instead of ocular I
as in the Echinidse and Strongylocentrotidse. This seems a very significant
fact when taken in connection with the small size of the abactinal system.

The family is confined to the tropics and is made up entirely of reef-dwelling,
littoral species. With the exception of the two West Indian Echinometras,
none of the family are found outside of the Indo-Pacific region. It is interesting
to note that with only two exceptions, there are no localities north or south of
30° latitude, where members of the family are known to occur. One of these
exceptions is the Bermuda Islands, which are north of 30° N., where Echinometra
lucunter is common, and the other exception is Lord Howe Island, which is
south of 30° S., where Echinometra mathcei occurs. The old record of the occur-
rence of the latter species at the Cape of Good Hope has not been confirmed by
later collecting and is doubtless erroneous, while the reported occurrence of an
Echinometra in New Zealand waters (Index Faunge Novse Zealandse, p. 288),
supposedly based on the collections of the "Challenger," is obviously erroneous,
as the "Challenger" collected no shallow water Echini whatever, in that region.

366 HAWAIIAN AND OTHER PACIFIC ECHINI.

THE SPINES, PEDICELLARI.E, SPH^RIDIA, AND SPICULES.
Plate 95, figs. 1-15.

In no other family do the spines reach such a degree of specialization as they
do in the Echinometridse. While Parasalenia and Echinometra are not remark-
able in this respect, the remaining genera are extraordinary. In Heterocentro-
tus, the primary spines have become so long and stout, as to give the animal a
very heavy appearance, while the secondaries, though not lengthened are greatly
thickened. In Colobocentrotus and Podophora on the other hand, the spines
are all reduced to a uniform length abactinally and being correspondingly thick-
ened, they have come to form a close, smooth covering over the whole upper
surface. Although little is known of the habits of the species, contained in these
three genera, it can hardly be doubted that these remarkable modifications of
the spines are due to the conditions under which they live on surf -exposed reefs.
Even in Echinometra, a tendency for the abactinal secondary spines to become
modified in a manner somewhat similar to that shown by Podophora is found.

The pedicellariae of the Echinometridse are not sufficiently different from those
of the Strongylocentrotidae to warrant any extended description. The valves
of the globiferous have as a rule, a single lateral tooth on the left side but this
is wanting in Parasalenia and there is considerable diversity, in Podophora,
the lateral tooth sometimes occurring on the right side, and occasionally either
it or the terminal tooth is paired (see A. Agassiz, 1908, Mem. M. C. Z., XXXIX,
PI. 5, figs. 10-13). As a rule, these pedicellariae are smaller than in the Echinidse
and Strongylocentrotidse, but in some species they are well developed. The
ophicephalous, tridentate, and triphyllous pedicellarise are commonly present
in large numbers and reveal great diversities of form, even within a single species.
As they have been very fully illustrated for Colobocentrotus and Podophora
(A. Agassiz, 1908, 1. c.), it would be quite superfluous to figure or discuss them
in detail here. The tridentate in spite of their diversity sometimes furnish
excellent specific characters.

The sphseridia occur as in the related families on the actinal portion of each
ambulacrum. There are commonly 4-6 in each area but there may be 8 or
even more. They show great diversity in form, as has been very fully shown
recently, for Colobocentrotus and Podophora (A. Agassiz, 1908, Mem. M. C. Z.,
XXXIX, Pis. 31 and 32).

The calcareous spicules of the pedicels, show an unusual diversity. They are

ECHINOMETRID^:. 367

most commonly bihamate but in Echinometra they are often triradiate. In
Parasalenia, they are very peculiar in having two little projections on the con-
cave side, quite unlike anything seen in other Echini.

THE GENERA AND SPECIES OF ECHINOMETRID^E.

Generic distinctions in this family are quite sharp and it is easy to distinguish
the five which seem valid. Specific distinctions are on the contrary remarkably
ill-defined and except in Podophora, it is a matter of difficulty to decide just how
many species ought to be recognized. Doderlein's attempt to distinguish Echino-
metra oblonga as the type of a new genus seems peculiarly unfortunate, not merely
because the character is microscopic but because it is very variable and is not
distinctive. Triradiate spicules occur in the pedicels of Echinometra picta,
Van Brunti, and especially viridis, as well as in those of oblonga. It is true that
some specimens of oblonga have a much greater development of such spicules
than are ever seen in any other species, but there is great individual diversity;
I have never seen a specimen with nearly as many as those photographed by
Doderlein appear to have possessed, while on the other hand, some specimens
have very few. Far from considering oblonga as worthy of generic rank, de
Meijere's view that it is simply an extreme form of mathcei seems much more
probable, for a considerable number of specimens have been seen which could be
assigned to one species or the other only quite arbitrarily.

The characters which are of most service in distinguishing the genera of this
family, are the position of the long axis, the structure of the ambulacra, the
covering of the periproct, and the character of the spines.

The following table will show how these are grouped in the different genera
here recognized: —

Long axis through IIIb-5b; ambulacral plates of 3 elements; periproct covered

by 4 subequal plates; spines not peculiar Parasalenia.

Long axis to the left of IIIb-5b; ambulacral plates of more than 3 elements; peri-
proct with numerous plates.

Long axis through 3-1 ; ambulacral plates of 4-8 elements; spines not peculiar Echinometra.
Long axis through IVb-lb or 4a-IIa; ambulacral plates of 9-19 elements; spines

extraordinarily modified.
Primary spines very short, thick and flat-topped, forming a smooth abacti-

nal pavement.
Primary tubercles very large, at least in mid-zone, 3-6 in an irregular

horizontal row on each interambulacral plate Podophora.

Primary tubercles moderate, in mid-zone 10-12 in two horizontal

series on each interambulacral plate Colobocentrotus.

Primary spines remarkably long, stout, heavy Heterocentrotus.

368 HAWAIIAN AND OTHER PACIFIC ECHINI.

PARASALENIA.

v

A. Agassiz, 1863. Bull. M. C. Z., I, p. 22.
Type-species, Parasalenia graliosa A. Agassiz, 1863, 1. c.

This is a very interesting genus and its affinities are by no means clear. The
ambulacra are echinoid, the anal plates are like Arbacia, while the form of the
test is like Echinometra. The spines are externally like Echinometra and the
differences internally are hardly as important as Me Intosh and Mortensen seem
to think. When compared with Echinometra-spines of the same size, those
of Parasalenia are not essentially different. The lantern, teeth, and auricles are
like those of small Echinometras, except that there are apparently no projections
on the epiphyses for the support of the teeth. It is clear from these important
structures about the mouth, that there is no near relationship to Arbacia, and
the pedicellarise confirm this conclusion. Except that the lateral tooth is lacking
on the left side of the narrow blade, there is nothing about the globiferous pedi-
cellarise to distinguish them from those of Echinometra, and the other forms
are even less peculiar. The calcareous spicules in the pedic"els of Parasalenia
are very characteristic but they do not indicate relationship to any othe r group.
On the whole it seems that Parasalenia is an isolated offshoot from the echinid
stock from which the Echinometridse have sprung and that it is not properly a
part of that family from the phylogenetic point o f view, but as a matter of con-
venience it may be placed therein.

Whether the genus contains more than one valid species seems uncertain.
Owing to insufficient material the status of Pohlii is doubtful. Typical specimens
of the two nominal species are very different but as is so often the case, there
are certain specimens which are perfectly intermediate at least in many ways.
The red coloration of Pohlii is very noticeable and extends even to the calcareous
matter in the pedicellariae but specimens of gratiosa are sometimes quite red,
especially when young. The primary spines of graliosa are sometimes banded
and if such specimens are also more or less red, the resemblance to Pohlii is very
marked. In our specimens of Pohlii, genital 3 is excluded from contact with the
periproct, by the meeting of genitals 2 and 4, but whether this is a constant
specific character, is not known; if it is, it is an excellent one. In our specimens,
globiferous pedicellarise are extremely rare, and tridentate fairly common in
gratiosa, while in Pohlii, the globiferous are abundant while the tridentate are
very rare; but it may be doubted if this difference holds thoughout the species.

What seem to be the real specific differences may be stated as follows : —

PARASALENIA POHLII. 369

Periproct moderately large, its long diameter about | that of the long diameter of

the abactinal system; each genital plate with at least one well-developed tubercle gratiosa.

Periproct very small, its long diameter less than J that of the long diameter of the

abactinal system; genital plates without tubercles Pohlii.

Parasalenia gratiosa A. Ag.

Parasalenia gratiosa A. Agassiz, 1863. Bull. M. C. Z., I, p. 22.

The coloration of adult specimens of this species is very striking for the test
and spines are very dark, more or less nearly black, while the milled ring of each
primary spine, is pure white, in marked contrast. Half grown specimens are
much lighter, usually a sort of reddish brown. In really young specimens, the
shade is often very nearly red, and the spines are more or less banded. The only
specimen brought home by the "Albatross" has the long axis 11 mm. and the
short one, 8 mm. It has a distinctly roseate tinge and the primary spines are
banded, though not with red and white. The secondary spines, and the bases
and tips of the primaries are dull pink; the remainder of each primary is light
greenish brown with three indistinct bands of a much lighter shade; the milled
rings are pearly white. The long diameter of the abactinal system is 4 mm. and
that of the periproct is almost 2. All of the genital plates are in contact with
the periproct but not one of them has even a minute tubercle! Such a specimen
renders the distinctness of Pohlii doubtful.

Makemo, Paumotus, Oct. 21, 1899.

Parasalenia Pohlii Pfeff.

Parasalenia Pohlii Pfeffer, 1887. Verhandl. Ver. Naturw. Unterh. Hamburg, VI, p. 110.

Plate 95, figs. 1-5.

It has seemed desirable to give a few figures illustrating the pedicellarise of
this species, for the purpose of making easy, comparison with those of Echino-
metra. Only one tridentate was found (PI. 95, fig. 1} ; it had an exceedingly long
neck and slender valves about a millimeter long, and is not essentially different
from those of gratiosa and Echinometra. The globiferous are abundant, usually
with glands (PL 95, fig. 8} but occasionally without (PL 95, fig. 2) ; they vary
much in size, the valves (PL 95, fig. 4) , which lack a lateral tooth on the blade
(PL 95, fig. 5), ranging from .30 to .60 mm. The ophicephalous and triphyllous
are small but fairly common, and the former have the calcareous matter in the
valves distinctly red. In other respects, these pedicellariae cannot be distin-
guished from those of gratiosa and Echinometra.

370 HAWAIIAN AND OTHER PACIFIC ECHINI.

ECHINOMETEA.

Gray, 1825. Ann. Phil., XXVI, (n. s. X), p. 426 (4 of reprint).
Type-species, Echinus lucunter Lmn&, 1758. Syst. Nat., ed. 10, p. 665.

Although this genus is so well characterized, one occasionally meets with a
specimen, usually young, in which the long axis is so little longer than the short
one, that it is necessary to measure the specimen to determine which is the long
axis. Such specimens have very naturally a close resemblance to individuals
of Heliocidaris and Strongylocentrotus, but they can usually be distinguished
without difficulty by the flatness of the test, which is generally very marked,
and by the large size of the auricles.

The number of valid species of Echinometra is far from being satisfactorily
determined. There are few Echini so variable in form and color as E. Mathcei
and it may well be doubted whether oblonga is anything more than an extreme
form of that species, and picta is probably not specifically distinct. Both these
species are retained however, that the points in which they differ from Mathcei may
be emphasized, in the hope that some zoologist who has a favorable opportunity,
such as the Hawaiian Islands afford, may make a careful study of the diversity
shown by the Indo-Pacific Echinometras. The American species, lucunter
and Van Brunti, are quite easily distinguished from the others by the large num-
ber of pore-pairs in each arc, but they are not so easily distinguished from each
other. The difference in the auricles however is surprisingly constant and will
separate all but very young specimens. The differences in the shape of the
test, the length of the primary spines, and the character of the tridentate pedi-
cellarise seem to be less constant, though they are all useful. Ellipsechinus
macrostomus Ltk. seems to be only a large specimen of Echinometra lucunter from
the eastern Atlantic Ocean. Large specimens from the Cape Verde Islands have
8 pore-pairs in many of the arcs and answer well to Liitken's description, except
for the position of the long axis. Dr. Mortensen has however reexamined
Liitken's type and has kindly assured me that the long axis is erroneously given
in the description, and that it really coincides with that of Echinometra. The
specimens brought home by the "Albatross" from Easter Island have proven
very interesting, for their superficial appearance is so much like that of Van Brunti,
it was not until the ambulacra were examined that their nearer relationship to
Mathcei became evident. The difference from Mathcei is however so constant as
to require a new species, insularis, for this interesting form. Jackson (1911,
Mem. Boston Soc. Nat. Hist., VII, p. 147) has shown that lucunter and Van

ECHINOMETRA. 371

Brunli differ from the other members of the genus, in that ocular V is insert
in the great majority of specimens, whereas in all of the others the oculars
are usually all exsert. It is interesting to note that in Van Brunli, which is
most progressive as regards number of elements in an ambulacral plate, we find
the most progressive abactinal system, oculars V and I being insert in nearly
three quarters of the specimens. — The largest Echinometras seen are from
Bermuda and the Abrolhos Reefs, Brazil. They measure from 85-95 mm. in
length and are about 75-85 mm. wide and 42-48 mm. high. Great diversity
is shown in the genus in the relative length, breadth, and height of the test.
The breadth ranges from .75 to .96 of the length; when it is as much as .95,
small specimens of course appear to be circular in outline. The height ranges
from .40 to .63 of the length ; as a rule low specimens are wide and high speci-
mens are narrow.

In distinguishing the seven species herein recognized, the number of pore-
pairs in an arc is considered of first importance, for while there is more or less
variation from the normal number, yet in the mid-zone and above, the variation
is slight, until one approaches the ocular; of course in young specimens, the
largest number of pores will be above the-mid-zone. The form of the test, the
structure of the auricles and the character of the primary spines prove to be
more or less useful points in most of the species. The pedicellarise are of little
use, though the tridentate have a characteristic form in at least one species.
In some cases, color may be of use for distinguishing a given species but it is
very unreliable and in Mathcei at least is almost useless. The size and tubercu-
lation of the abactinal system and even the form of the abactinal secondary
spines are useful characters in recognizing some species.

The following table shows how the seven species may be distinguished: —

Pore-pairs, at or above ambitus, 4 or 5 in each arc, and not rarely 6.
Pore-pairs in arcs of 4, not rarely 5.

Color very rarely black; primary spines variable but usually pointed and

their thickness generally much less than ten per cent of their length.

Height of test usually .55-.6S of length; abactinal system .15-.25 of

test-length; usually 3 or more secondary tubercles on each genital

plate; color of primary spines very variable, often dark purplish or

reddish, frequently tipped with yellow or red . . • Mathcei.

Height of test less than half the length; abactinal system .20-.30 of
test-length ; seldom more than one secondary tubercle on each genital
pla{e; primaries usually light fawn-color or greenish, but sometimes

dark with light tips picla.

Color usually black, rarely dull purplish gray or even dull red; primary
spines short, stout, and usually blunt, their thickness .12-.22 of their own

372 HAWAIIAN AND OTHER PACIFIC ECHINI.

length; primary tubercles correspondingly big, especially noticeable abacti-

nally; height of test usually .55-.60 of its length oblonga.

Pore-pairs in arcs of 5, not rarely 6.

Primary spines deep purplish or reddish; abactinal system with numer-
ous secondary and miliary spines, many of which are low, thick

and truncate, or capitate; ambulacra petaloid actinally insularis.

Primary spines light brownish, more or less extensively green distally,
usually purple-tipped; abactinal system remarkably free from spines;

ambulacra not petaloid actinally viridis.

Pore-pairs, at or above ambitus usually 7 or 8 in each arc, rarely only 6.

Auricles stout with conspicuous supplementary "tags," arising vertically from
their united ends; few, sometimes none, of the arcs of pores with more than
6 pairs; test usually more than half as high as long with the width usually
about .85 of the length; primary spines usually much shorter than width
of test; valves of trident ate pedicellariac narrow at tip, in contact for most

of their length lucunter.

Auricles slender without conspicuous "tags"; pore-pairs usually 7 or 8; test
about half as high as long or less; width generally .90-. 95 of length; primary
spines nearly equal to short diameter of test; valves of tridentate pedicellaria;
widened at tip and in contact only distally Van Brunti.

Echinometra Mathaei Bl.

Echinus Mathai de Blainville, 1825. Diet. Sci. Nat., XXXVII, p. 94.
Echinometra Mathsei de Blainville, 1830. Diet. Sci. Nat., LX, p. 206.

The specimens of this very common and wide spread sea-urchin, which
the "Albatross" collected from various parts of the Pacific, reveal the usual
diversities of color and form. Some of them are very fine examples of
the typical form, except that there are none in which the color is at all green.
The finest specimens are from Laysan and Manga Reva. The former have the
test dark brown, with the spines light pinkish fawn-color; the largest is 64 mm.
long, 53 wide and 40 high. Those from Manga Reva are not so large but are
much more handsomely colored as the spines are light purplish gray abruptly
tipped with cream-color.

The following are the places, whence specimens were collected : —
Station 3959. Near Laysan Island, Hawaiian Islands, 10 fathoms. Wh., s., co.
Laysan Island, H. I. Bora-bora, Society Islands.

Necker Island, H. I. Fakarava, Paumotu Islands.

Kamalina Bay, Niihau, H. I. Makemo, Paumotu Islands.

Honolulu Reef, Oahu, H. I. Rangiroa, Paumotu Islands.

Waikiki Beach, Oahu, H. I. Manga Reva, Paumotu Islands.

Hilo, Hawaii, H. I.
Forty-nine specimens.

ECHINOMETRA OBLONGA. 373

Echinometra picta A. Ag. and CI.

Echinometra picta A. Agassiz and Clark, 1907. Bull M. C. Z., L, p. 241.

Plates 95, figs. 6-12; 114, figs. 5, 6.

The more the specimens on which this species is based are studied the more
doubtful it seems whether it is really distinct from Mathcei, but the figures of
typical specimens, as well as of the pedicellarise, will enable other zoologists to
recognize the form to which attention is called. In such specimens, the test is
distinctly flattened and wider than usual, the height being less than one half
the length, while the shorter diameter is about .85-.90 of the longer. The spines
seem to be less crowded and are somewhat more slender than in Mathcei, and
are much less numerous on the abactinal system. The latter is distinctly larger
than in that species, its diameter sometimes nearly equalling one third of the
test-length. The color is dark brown for the test and light fawn-color for the
spines. Such specimens as these occur only in material from the Hawaiian,
Society, and Philippine Islands. They intergrade with Mathosi in each one of
these characters taken by itself, more or less fully, but are usually easily distin-
guished by their general appearance resulting from the combination of all five.
The pedicellarise (PI. 95, figs. 6-12) show no characteristic peculiarities, but the
various forms illustrative of the genus are figured.

The "Albatross" collected picta at the following places: —

Station 3881. Napili Harbor, Maui, Hawaiian Islands.

Station 3975. Off Necker Island Shoal, H. I. 16-171 fathoms. Crs. S.,
co., sh.

Necker Island, H. I. Puako Bay, Hawaii, H. I.

Kamalina Bay, Niihau, H. I. Hilo Bay, Hawaii, H. I.

Waianae, Oahu, H. I. Fakarava, Paumotu Islands.

Honolulu Reef, H. I.

Thirty-four specimens.

Echinometra oblonga Bl.

Echinus oblongus de Blainville, 1825. Diet. Sci. Nat., XXXVII, p. 95.
Echinometra oblongus de Blainville, 1830. Diet. Sci. Nat., LX, p. 206.

Plate 114, figs. 1, 2.

As this interesting form has never been figured, so far as known, it has
seemed desirable to give an illustration showing its general appearance. The

374 HAWAIIAN AND OTHER PACIFIC ECHINI.

reasons for not accepting the new genus proposed by Doderlein, of which oblonga
is the type are given above (p. 367). The series of specimens at hand is most
perplexing, so great is the individual diversity shown. On the one hand, there
are from Clarion Island, black specimens with high, compressed test, big abacti-
nal tubercles, and small abactinal system with all the oculars completely ex-
cluded from the periproct, but with spines tapering to a point and not nearly as
thick as they should be in oblonga. And on the other hand there are specimens
from Makemo in the Paumotus, which have the characteristic spines of oblonga
but show a remarkable diversity in color; some are deep purple, some grayish
lavender, and one is dull purplish red. In the presence of these puzzling
specimens, much sympathy is felt with de Meijere's remark, apropos of
oblonga, "Dass letztere eine gute Art 1st, glaube ich kaum."

Mortensen first called attention to the curious fact that the globiferous
pedicellaria? of oblonga have a jointed stalk. Examination of a considerable
amount of material shows that this peculiarity exists in less than half the globif-
erous pedicellariae; at least, more than half of those examined did not show it.
Such a joint in the stalk of the pedicellarige is not known in any other echinoid,
although it is easy to see how it might be of real use.

The "Albatross" obtained specimens of oblonga from the following stations: —

Laysan, Hawaiian Islands. Hilo, Hawaii, H. I.

Necker Island, H. I. Puako Bay, Hawaii, H. I.

Lanai, H. I. Rangiroa, Paumotu Islands.

Kamalina Bay, Niihau, H. I. Makemo, Paumotus.

Hanalei, Kauai, H. I. Fakarava, Paumotus.

Waianae, Oahu, H. I. Clarion Island, Mexico.

Honolulu Reef, Oahu, H. I.
Sixty-seven specimens.

Echinometra insularis,1 sp. nov.

Plates 95, figs. 13-15; 114, figs. 3, 4.

On first examination the Echinometras from Easter Island showed considerable
resemblance to Van Brunti, but further study soon proved that they are more
nearly related to Mathcei. The test is rather wide and flat, the width being
.S5-.95 of the length and the height about half the length or a trifle less. The

1 Insularis = pertaining to an island.

ECHINOMETRA VAN BRUNTI. 375

abactinal system is about .20 of the length. The largest specimen is 45 mm.
long, without including the spines, which are long and slender as in Fan Brunti,
and may exceed the short diameter of the test. There are 10-14 primary inter-
ambulacral plates in each column. The pore-pairs are generally in arcs of five
at the ambitus but some of the abactinal arcs usually have six and in one or two
specimens most of the abactinal arcs have six. The ocular plates are all broadly
excluded from the periproct in the ten normal specimens examined. The small
spines abactinally, particularly on the abactinal system, are distinctly capitate
and many of them are remarkably low and stout (PI. 95, fig. 13). Actinally
the ambulacra become somewhat widened but they are not quite so distinctly
petaloid as in Van Brunti. The tridentate pedicellarise have rather narrow
valves (PI. 95, fig. 15) as in most Echinometras and they are not widened at the
tip as in Van Brunti. The pedicellarise of all kinds are rather rare and small;
the valves of the globiferous (PL 95, fig. H) are about .30 mm. or less, while
those of the tridentate and ophicephalous are only .30-.40 mm. No spicules
whatever were found in the tube-feet. The color of all the specimens is uniformly
deep purple or brownish red.

Not having seen this species previously, it was supposed to be peculiar to
Easter Island, and it was therefore rather surprising to find that the Echino-
metras which the "Albatross" collected in March, 1889, at Socorro Island, must
be referred to the same form. One would naturally have expected Mexican
species in the Revilla Gigedo Islands, and it is therefore interesting to find no
specimens of Van Brunti, but instead oblonga at Clarion Island and insularis at
Socorro.

The material consists of the following lots: —

Socorro, Revilla Gigedo Islands, Mexico, March, 1889.

Easter Island, southeastern Pacific Ocean, Dec. 1904.

Twenty-eight specimens.

Echinometra Van Brunti A. Ag.

Echinometra Van Brunti A. Agassiz, 1863. Bull. M. C. Z., I, p. 21.

There are three fine specimens of this species before us, taken by the "Alba-
tross" in May, 1888. One of them is the largest example of this species seen,
measuring 73 mm. in length by 70 mm. in width, regardless of the spines, some
of which are 40 mm. long.

Santa Margarita Island, Lower California, Mexico.

Three specimens.

376 HAWAIIAN AND OTHER PACIFIC ECHINI.

PODOPHORA.

L. Agassiz, 1840. Cat, Syst. Ectyp. Echinod., p. 19.
Type-species, Echinometra atrata de Blainville, 1830. Diet. Sci. Nat., LX, p. 206.

As this genus and the following have been so fully discussed in a recent
Monograph (A. Agassiz, 1908, Mem. M. C. Z., XXXIX), it would be quite
superfluous to go into any details here. There are two distinct species in
the genus, one of which (atrata) is wide spread in the Indo-Pacific region,
ranging from Zanzibar to Hawaii, while the other (pedifera) is confined, so far as
known, to the southeastern Pacific.

The two species are readily distinguished from each other as follows : -

Color abactinally, usually deep purple, rarely greenish; marginal primary spines with
rounded, sometimes swollen ends, not forming a close-set, even margin; actinal
petaliferous areas of ambulacra moderately developed, not abruptly constricting
the interambulacral spaces; pore-pairs at ambitus in arcs of 9 or 10 atrata.

Color, abaetinally olive-green; marginal primary spines with flattened, chisel-like tips,
forming a close-set, even margin; actinal petaliferous ambulacral areas greatly
developed, abruptly constricting the interambulacral spaces; pore-pairs at ambi-
tus in arcs of 10-12 pedifera.

Podophora atrata Agass.

Echinus atratus LinnS, 1758. Syst. Nat., ed. 10, p. 655.

Podophora atrata L. Agassiz, 1840. Cat. Syst. Ectyp. Echinod., p. 19.

"Colobocentrotus Quoyi Brandt," A Agassiz and Clark, 1907. Bull. M. C. Z., L, p. 240.

This appears to be one of the common and characteristic sea-urchins of the
Hawaiian Islands, and was found by the "Albatross" in considerable numbers
particularly at Puako Bay, Hawaii. A collector's note with a jar of specimens
from this place reads as follows: —

"Common on rocks along shore. Live in little depressions, the border spines
acting as the edge of a sucking disk. Dorsal surface very deep Indian purple
or prune-purple when in the water. By reflected light, out of water, almost
black. Underside of big spines, mauve. Circumoral membrane and membrane
between clumps of small spines and tubefeet, orange-red. Small spines and
tube-feet, light brownish, occasionally red or purple."

Necker Island, Hawaiian Islands.

Kamalino Bay, Niihau, H. I.

Napili Harbor, Maui, H. L, "Albatross" St. 3881.

Napili, Maui, H. I.

HETEROCENTROTUS. 377

Lanai, H. I.

Hilo, Hawaii, H. I.

Puako Bay, Hawaii, H. I.

One hundred and twenty-eight specimens.

Podophora pedifera Agass. and Des.

Echinus pedifer de Blainville, 1825. Diet. Sci. Nat., XXXVII, p. 97.
Podophora pedifera Agassiz and Desor, 1846. Ann. Sci. Nat., (3), VI, p. 374.

This species has been confused with the preceding but the large series collected
by the "Albatross" in the Paumotus shows clearly that it is quite distinct and
that its specific characters are well marked and constant. They need not be
discussed here however as they are fully dealt with in the monograph on the
genus, already referred to.

The specimens brought home by the "Albatross" were collected on the
reefs of

Rangiroa, Paumotu Islands.

Fakarava, Paumotu Islands.

One hundred and eleven specimens.

COLOBOCENTROTUS.

Brandt, 1835. Prodrom. desc. Anim., p. 266 (or 66).
Type-species, Colobocentrolus Meriensii Brandt, 1835, 1. c.

It is a curious fact that so far as any reliable records show this genus is con-
fined to the Bonin Islands, and few specimens have found their way into collec-
tions. There is some doubt as to whether all the known specimens should be
referred to a single species (Meriensii Brandt) or not, but there seems to be a
second (Stimpsoni A. Ag.).

The two are distinguished as follows : —

Short diameter of test more than .90 of long diameter so that ambitus is nearly circular . . Meriensii.
Short diameter of test less than .90 of long diameter so that ambitus is elliptical . . Stimpsoni.

HETEROCENTROTUS.

Brandt, 1835. Prodrom. desc. Anim., p. 265 (or 65).
Type-species, Echinus mammillatus Linn6, 1758. Syst. Nat., ed. 10, p. 667.

There seems to be general agreement that this well-known genus contains
two, and only two, species, and an examination of all the available material

378 HAWAIIAN AND OTHER PACIFIC ECHINI.

confirms the opinion. But this examination also shows that the primary spines
are of no value whatever as a means of distinguishing the species, few Echini
showing so great variability in this respect. The secondary spines are of more
use, though they cannot be absolutely relied on. The best specific character is
found in the arcs of pores, the difference in this respect being remarkably con-
stant. A rather marked difference in the tridentate pedicellarise has been noticed
by Mortensen, but I have made no attempt to determine how constant this is.
By the use of these various characters, the two species maybe distinguished
as follows: —

Pore-pairs, in mid-zone, in arcs of 11 (10-12) ; secondary spines, short, flaring, and trun-
cated at tip; no large primary tubercles in abactinal part of ambulacra; valves
of tridentate pedicellarise somewhat curved, widened at tip where alone they are
in contact mammillatus.

Pore-pairs, in mid-zone, in arcs of 15 (14-19); secondary spines, short, usually tapering
and pointed at tip; normal primary tubercles in abactinal part of ambulacra;
valves of tridentate pedicellariae bluntly pointed, in contact for most of their
length trigonarius.

Heterocentrotus mammillatus Br.

Echinus mammillatus Linne, 1758. Syst. Nat., ed. 10, p. 667.

Heterocentrotus mammillatus Brandt, 1835. Prodrom. desc. Anim., p. 266 (or 66)

Plates 115-117.

The specimens of Heterocentrotus in the "Albatross" collection are divisible,
at a glance, into two groups on account of their color. In one group, the general
coloration is brown, the secondaries lighter than the primaries and usually light
fawn- or even cream-color; these specimens are all mammillatus and are all from
the Hawaiian Islands. In the other group the general coloration is deep purple
and there is little difference in shade between primaries and secondaries; these
specimens are all trigonarius and are from the Paumotus. So far as known
trigonarius does not occur in the Hawaiian Islands, nor mammillatus in the
Paumotus. Indeed the latter seems to be confined to the North Pacific and
Indian Oceans; the M. C. Z. has good series from Mauritius, the Bonin Islands,
and the Hawaiian Islands, but only one specimen from south of the equator
in the Pacific, and that one is a young individual from Jarvis Island (which is
less than 23' S.), the identification of which admits of some doubt. So far as
our material is concerned, mammillatus is often very light colored, and never
shows any indication of purple; when dark colored, the shade is always a rich
chocolate-brown.

HETEROCENTROTUS TRIGONARIUS. 379

The "Albatross" material is from the following places: —
Laysan, Hawaiian Islands.
Honolulu Market, Oahu, H. I.
Puako Bay, Hawaii, H. I.
Twenty-six specimens.

Heterocentrotus trigonarius Br.

Echinus trigonarius Lamarck, 1816. Anim. s. Vert., Ill, p. 51.
Heterocentrotus trigonarius Brandt, 1835. Prodrom. desc. Anim., p. 266 (or 66).

Plates 118-120.

This species shows much greater diversity in every feature than mammillatus
does, but the number of pore-pairs in at least one arc appears to be always more
than 12, even in specimens, whose test does not exceed 20 mm. in length. In
specimens only half grown there is almost always at least one arc with 15 pairs,
but occasionally 14 is the maximum. The primaries range in form from long,
tapering, acute, trigonal spines to short, club-shaped spines with very thick
blunt tips. The actinal spines are of course more or less flattened, but the
amount of flattening and the number of primaries involved shows great diversity.
The secondaries show nearly as great diversity, for while they are usually more
or less acutely pointed, we have a number of specimens in which they are as thick
and truncate at tip as in any mammillatus. The color is also variable, and in
this particular there is a suggestion of local differences; specimens from Mauritius
are brown with distinct shades of green and orange-red on the primaries, while
spe'cimens from the Paumotus are deep purple, with very little variation except
in shade. Specimens from intermediate localities connect the two extremes,
which will probably at some future time be designated as subspecies.

Apparently trigonarius is more widely distributed or at least, more generally
distributed than mammillatus. Specimens are at hand from Mauritius, the
Philippines, New Guinea, the Carolines, the Gilberts, the Marshalls, the Fijis,
Baker's Island, Tongatabu, the Society Islands, and the Paumotus. As the
two species are so easily confused with each other, the published records especially
if they date back half a century or more are unreliable so that the question of the
geographical distribution of the two species of Heterocentrotus is still an open
one. The evidence however indicates that mammillatus is a northern species
ranging from the Red Sea (eastern Mediterranean?) to Hawaii, while tri-
gonarius is southern, ranging from Zanzibar to the Paumotus. The two occur

380 HAWAIIAN AND OTHER PACIFIC ECHINI.

together in abundance at Mauritius and probably in the East Indies, but just
where and to what extent their ranges overlap is as yet unknown.

The "Albatross" specimens of trigonarius come from the following place: —

Fakarava, Paumotus, Oct., 1899.

Sixteen specimens.

ADDENDA. 381

ADDENDA.

On p. 73 the genus Podocidaris is stated to be monotypic, but after that
statement was published, Dr. W. K. Fisher sent me a single specimen of a Podo-
cidaris which was found by him among the starfishes collected by the "Albatross"
at the Hawaiian Islands. As it seems to be distinct from the West Indian species,
it may be described as

•.

Podocidaris ornata,1 sp. nov.
Plate 102, fig. 1.

Test flattened, the height only 4 mm. while the horizontal diameter is 9 mm.
The abactinal system is 4.5 mm. across and the actinostome about half a milli-
meter more. The periproct is 2 mm. in diameter and is completely covered by
four equal plates. The primary spines are confined to the region below the
ambitus, as is characteristic of the genus, and are very flat, as seen from below,
but on the upper side are distinctly keeled along the median line. The longest
spines are not quite equal to half the diameter of the test, but they are very
broad, the width, half way between the middle and 'the tip, being equal to one
fourth of the length. Each spine ends in an enamel hoof, as usual in the family,
but this is scarcely noticeable from above. The "non-articulated spines"
or better, " cyclindrical tubercles," of the abactinal half of the test are slenderer
and less conspicuous than in sculpta; they also form less regular series, many of
them remaining very imperfectly developed. They are present in some numbers
on the genital and ocular plates and their arrangement there in two or three
series parallel to the distal margins of the genital plates, gives the abactinal
system a very ornamented appearance, quite different from that of sculpta.
The pedicellarise of ornata show no special peculiarities excepting that they are
more slender than in sculpta and the ophicephalous are smaller. The tridentate
have the valves about .30 mm. long but the blade where widest, near tip, is less
than .06 mm. broad; in sculpta, the blade is more than one fourth as wide as
the length of the valve. In the ophicephalous pedicellarise of ornatus the valves
are only .15-.20 mm. long. The spicules of the tube-feet are not distinguishable
from those of sculpta except in being rather more slender. The sphseridia are
like those of sculpta in form, but there seem to be two in each ambulacrum, one

1 ornatus = ornamented.

382 HAWAIIAN AND OTHER PACIFIC ECHINI.

distal to the other. The test is dull olive, brown on the abactinal system, but
distinctly greenish in the mid-zone. The spines are almost white but have more
or less dusky tips, and the bases show a trace of green or reddish brown.

In view of the occurrence of Habrocidaris argentea in the Hawaiian region,
it is most interesting to have the closely related genus Podocidaris discovered
there. It is a very peculiar fact in the geographical distribution of Echini that
these two genera of Arbaciadse, each of which was discovered in the West Indies,
should occur in the Hawaiian region, and so far as now known nowhere else in
the world, except perhaps the East Indies, where the "Siboga" collected two
specimens of what is probably a true Podocidaris. The genus Trigonocidaris
(see p. 295) shows a very similar distribution but otherwise it appears to be
unique among Echini. In these three cases, the Hawaiian species is identical
with or very closely related to the West Indian and it seems impossible to doubt
that it has sprung from that stock, and the lack of marked specific differentiation
would indicate that it was at no very distant date. Possibly the East Indies
were stocked from the Hawaiian Islands.

The Hawaiian Podocidaris was taken by the "Albatross" only at the follow-
ing station: —

Station 3919. Off Diamond Head, Oahu, Hawaiian Islands. Bott. temp.
45.6°. 220-257 fathoms. Gy. s.

As it seems most probable that the species designated as Podocidaris spJ
by de Meijere ("Siboga" Ech., p. 68) is quite different from any of its near allies,
it may be named cincta, since the banded spines are its most striking character.

The three species of Podocidaris may then be distinguished as follows: —

Spines not banded . . .

Tubercles on abactinal system few, low and not arranged according to any very definite

pattern sculpla.

Tubercles on abactinal system, rather numerous, arranged in series parallel to the distal

margins of the genital plates ornata.

Spines banded with carmine-red cincta.

YOSHIWARA'S ECHINI.

Through the kindness of Professor Seitaro Goto, the Imperial University
at Tokyo has entrusted to me the type-specimens of the ten species of Echini
described as new by Yoshiwarain 1897 (Ann. Zool. Jap., II, p. 57-61). In the
preceding pages reference has been made to four of these species, noting that
"Echinus multicolor" is a valid species of Nudechinus, that "Mespilia laevituber-

ADDENDA. 383

culatus" and "Salmacopsis pulchellimus" are synonyms of Mespilia globulus,
and that " Echinost rephus pentagonus" is a synonym of Echinostrephus molaris.
Of the other six species, three are clypeastroids and will be discussed in part 5
of the present work. The three remaining species are cidarids and a few com-
ments on them may be given here.

Cidaris (Stereocidaris) tenuispinus Yoshiwara. Examination of the type and
a cotype of this species leaves no doubt that the conclusion reached in 1907 that
it is identical with Doderlein's Dorocidaris japonica, is correct. For a full
discussion of the matter, see A. Agassiz and Clark, 1907, Bull. M. C. Z., LI, p.
112-114.

Cidaris (Stereocidaris) microtuberculatus Yoshiwara. Examination of the
type-specimen confirms the validity of this species, which has already been
diagnosed and figured (1907, Bull. M. C. Z., LI, p. 220, pis. 1 and 2).

Cidaris (Porocidaris) misakiensis Yoshiwara. The most striking feature of
the beautiful specimen on which this species is based is found in the primary
spines, which are more like those of some species of Cidaris (rugosa H. L. C. for
example) than they are like those of the other species of Porocidaris. They are
thickest not far from the base and then taper steadily to a blunt point. They
are longitudinally striated with about 28 series* of minute, sharp prickles, much
as in P. Sharreri. The longest spines are 95-100 mm. in length and 5 mm. in
diameter, 10-12 mm. above the base. The primaries are pure white, while the
secondaries, pedicellaria?, and test are white with a tinge of yellow. Yoshiwara
says the test, especially abactinally, and the collar of the primaries is "deep
brown" and the secondary spines "brownish." Evidently the specimen has
become completely bleached in alcohol. The abactinal system is .40 h. d. in
diameter. The pedicellarise are similar to those of P. variabilis and some of the
large ones have valves 5 mm. long. — There is no doubt that misakiensis is quite
distinct from elegans and it appears to be equally different from variabilis, but
it is very close to Sharreri and so far as our material shows, the form of the pri-
mary spines is the only difference worthy of mention. The specimen, described
and figured by deMeijere ("Siboga" Ech., p. 27, PI. 2, figs. 15, 16) as misakiensis,
does not agree well with the type, and it is doubtful if it is the same species.
It may be, as de Meijere suggests, only a form of elegans or it may possibly be
variabilis. The specific differences in the genus are very intangible and much
more material is necessary, from both the East and West Indies, before the
species already described can be satisfactorily known.

EXPLANATION OF THE PLATES.

Wherever the nature of the figure permits, the anterior ambulacrum (III) is placed
uppermost.

PLATE 90.

PLATE 90.

Showing the Arrangement of the Digestive and Reproductive Organs, and the Structure '
of the Lantern, and Auricles.

1^. Caenopedina hawaiiensis, nom. nov.

1. Interior view of actinal half of test with organs in place. X 1.5.

2. Interior view of almetinal half. X 1.5.

3. Exterior view of a pyramid from lantern. X 4.

4. Side view of same. X 4.

5-10. Glyptocidaris crenularis A. Ag.

5. Interior view of actinal half of test with organs in place. Slightly enlarged.

6. Interior view of abaotinal half. Slightly enlarged.

7. Side view of lantern. X 2.

8. Top view of lantern. X 2.

9. Side view of a single pyramid (tooth-supporting process broken off). X 2.

10. Part of base of corona, with auricles and apophyses. X 2.

11, 12. Stomopneustes variolaris Agass.

11. Interior view of actinal half of test with organs in place. Slightly reduced.

12. Interior view of abactinal half. Slightly reduced.

"ALBATROSSTACIFIC AND HAWAIIAN ECHINI

PLATE 90.

: ; r

•^MA m

A M.Westerfren del.

B.Meisi

PLATE 91.

PLATE 91.
1-13. Caenopedina hawaiiensis, nom. nov.

1. Large tridentate pedicellaria. X 70.

2. Side view of valve of same. X 70.

3. Interior view of same valve. X 70.

4. Slender tridentate pedicellaria. X 70.

5. Another slender tridentate pedicellaria. X 10.

6. Another slender tridentate pedieellaria. X 10.

7. Valve of ophicephalous pedieellaria. X 70.

8. Valve of triphyllons pedicellaria. X 70.

9. Peculiar, incomplete large tridentate pedicellaria. X 70.

10. Globiferous pedicellaria. X 70.

11. Side view of tip of valve of same. X 70.

12. Interior view of tip of same valve. X 70.

13. Sprueridium. X 70.

14-17. Caenopedina mirabilis Mrtsn.

14. Valve of ophicephalous pedieellaria. X 70.

1"). Valve of another ophicephalous pedicellaria. X 70.

Hi. Valve of globiferous pedicellaria. X 70.

17. Side view of tip of valve of same. X 150.

18-22. Caenopedina pulchella, comb. nov.

15. Side view of \-alve of globiferous pedicellaria. X 70.

19. Valve of tridentate pedicellaria. X 70. .

20. Valve of ophicephalous pedicellaria. X 70.

21. Calcareous plate from pedicel. X 300.

22. Sphseridium. X 70.

Ill

\

PLATE 92.

PLATE 92.
1-11. Glyptocidaris crenularis A. Ag.

1. Globiferous pedicellaria. X 30.

2. Tridentate pedicellaria. X 30.

3. Ophicephalous pedicellaria. X 30.

4. Triphyllous pedicellaria. X 30.

5. Valve of tridentate pedicellaria. X 30.
G. Valve of globiferous pedicellaria. X 70.
7. Side view of tip of same. X 70.

S. Valve of triphyllous pedicellaria. X 70.

9. Valve of ophicephalous pedicellaria. X 70.

10. Spha-ridiuin. X 70.

1 1 . Spicules from pedicels. X 70.

12-18. Pseudoboletia Indiana A. Ag., juv.

12. Large globiferous pedicellaria. X 30.

13. Small globiferous pedicellaria, with glands on the stalk. X 70.

14. Slender tridentate pedicellaria. X 70.

15. Valve of globiferous pedicellaria. X 70.

16. Side view of tip of same. X 70.

17. Valve of broad tridentate pedicellaria. X 70.

18. Sphreridium. X 70.

"ALBATROSr . MIAN ECHINI.

PLATE 92.

PLATE 93.
1. Nudechinus stictus, sp. nov.

1. Valve of globiferous pedicellaria. X 70.

2, 3. Echinus euryporus, sp. nov.

2. Valve of globiferous pedicellaria. X 70.

3. Side view of tip of same. X 70.

4, 5. Lytechinus euerces, sp. nov.

4. Valve of globiferous pedicellaria. X 70.

5. Valve of tridentate pedicellaria. X 70.

6, 7. Toxopneustes chloracanthus, sp. nov.

6. Side view of tip of valve of globiferous pedicellaria. X 70.

7. Valve of tridentate pedicellaria. X 70.

8, 9. Echinus anchistus, sp. nov.

S. Valve of globiferous pedicellaria. X 70.
9. Side view of tip of same. X 70.

10. Lytechinus dyscritus, sp. nov.

10. Valve of globiferous pedicellaria. X 70.

11-15. Echinus tylodes, sp. nov.

11. Valve of triphyllous pedicellaria. X 70.

12. Valve of ophicephalous pedicellaria. X 70.

13. Valve of globiferous pedicellaria. X 70.

14. Side view of tip of same. X 70.

15. Valve of tridentate pedicellaria. X 70.

16, 17. Genocidaris apoda A. Ag. and Cl.

16. Valve of globiferous pedicellaria. X 70.

17. Valve of ophicephalous pedicellaria. X 70.

18-21. Goniopneustes pentagonus Duncan.

IS. Valve of ophicephalous pedicellaria. X 70.

19. Valve of globiferous pedicellaria. X 70.

20. Side view of tip of same. X 70.

21. Valve of triphyllous pedicellaria. X 70.

22, 23. Gymnechinus epistichus, sp. nov.

22. Valve of globiferous pedicellaria. X 70.

23. Valve of tridentate pedicellaria. X 70.

24-26. Prionechinus sculptus A. Ag. and Cl.

24. Valve of globiferous pedicellaria. X 350.

25. Valve of tridentate pedicellaria. X 70.

26. SpliEEridium. X 70.

27-31. Orechinus monolini Dod.

27. Valve of globiferous pedicellaria. X 70.

28. Side view of tip of same. X 70.

29. Valve of tridentate pedicellaria. X 70.

30. Side view of same. X 70.

31. Sphseridium. X 70.

32. Echinus Wallisii A. Ag.

32. Globiferous pedicellaria, with entire stalk. X 30.

33. Salmacis Dussumieri Agass. and Des.

33. Base of valve of globiferous pedicellaria. X 70.

34. Salmacis erythracis, sp. nov.

34. Base of valve of globiferous pedicellaria. X 70.

35. Salmacis Belli Dod.

35. Base of valve of globiferous pedicellaria. X 70.

36. Salmacis bicolor Agass.

36. Base of valve of globiferous pedicellaria. X 70.

•

PLATE 94.
1-6. Pachycentrotus australiae, comb. nov.

1. Valve of globiferous pedicellaria. X 70.

2. Side view of tip of same. X 70.

3. Valve of tridentate pedicellaria. X 70.

4. Side view of a similar valve. X 70.

5. Valve of ophicephalous pedicellaria. X 70.
C. Valve of triphyllous pedicellaria. X 70.

7, 8. Heliocidaris crassispina, comb. nov.

7. Valve of globiferous pedicellaria from a specimen IS mm. h. d. X 70.

8. Side view of tip of same. X 70.

9. Strongylocentrotus franciscanus A. Ag.

9. Valve of globiferous pedicellaria. X 70.

10, 11. Paracentrotus Gaimardi Mrtsn.

10. Valve of globiferous pedicellaria. X 70.

11. Side view of same. X 70.

12. Paracentrotus lividus Mrtsn.

12. Side view of tip of valve of globiferous pedicellaria. X 70.

13-16. Strongylocentrotus echinoides A. Ag. and Cl.

13. Tridentate pedicellaria. X 20.

14. Valve of ophicephalous pedicellaria. X 70.

115. Base of valve of globiferous pedicellaria. X 70.
16. Side view of tip of same valve. X 70.
17-23. Strongylocentrotus nudus A. Ag.

I 17. Valve of ophicephalous pedicellaria. X 70.

15. Valve of globiferous pedicellaria. X 70.

19. Side view of tip of same. X 70.

20. Valve of triphyllous pedicellaria. X 70.

21. Valve of small tridentate pedicellaria. X 70.

22. Blade of valve of large tridentate pedicellaria, from specimen taken at Station 501S.

X 70.

23. Valve of large tridentate pedicellaria, from same specimen, with tip directed towards

observer. X 70.

24-27. Strongylocentrotus pulchellus A. Ag. and Cl.

24. Side view of valve of globiferous pedicellaria. X 70.

25. Base of same. X 70.

26. Valve of tridentate pedicellaria. X 70.

27. Sphseridium. X 70.

28.
29.

30.
31.
32.
33.

28, 29. Strongylocentrotus fragilis Jackson.

X 70.

Side view of valve of globiferous pedicellaria.
Base of same. X 70.

30-33. Strongylocentrotus polyacanthus A. Ag. and Cl.

Tridentate pedicellaria. X 70.

Side view of blade of valve of globiferous pedicellaria.

Base of a similar valve. X 70.

Valve of ophicephalous pedicellaria. X 70.

X 70.

1 1

AND HAWAII:.

PLATE 95.

PLATE 96.

1-5. Parasalenia Pohlii Pfeff.

1. Head and neck of tridentate pedicellaria. X 70.

2. Globiferous pedicellaria. X 70.

3. Globiferous pedicellaria with evident glands on the valves. X 70.

4. Valve of globiferous pedicellaria. X 70.

5. Side view of tip of same. X 70.

6-12. Echinometra picta A. Ag. and Cl.

6. Valve of globiferous pedicellaria. X 70.

7. Side view of same. X 70.

8. Valve of slender tridentate pedicellaria. X 70.

9. Side view of tip of same. X 70.

10. Blade of valve of stout tridentate pedicellaria. X 70.

11. Valve of ophicephalous pedicellaria. X 70.

12. Valve of triphyllous pedicellaria. X 70.

13-15. Echinometra insularis, sp. nov.

13. Abactinal secondary spines. X 10.

14. Side view of valve of globiferous pedicellaria. X 70.

15. Valve of tridentate pedicellaria. X 70.

16, 17. Caenocentrotus gibbosus, comb. nov.

16. Valve of globiferous pedicellaria. X 70.

17. Side view of tip of same. X 70.

18-22. Heliocidaris stenopora, nom. nov.

18. Valve of globiferous pedicellaria. X 70.

19. Side view of tip of same. X 70.

20. Valve of ophicephalous pedicellaria. X 70.

21. Valve of tridentate pedicellaria. X 70.

22. Valve of triphyllous pedicellaria. X 70.

23, 24. Echinostrephus aciculatus A. Ag.

23. Globiferous pedicellaria. X 70.

24. Side view of tip of a valve of globiferous pedicellaria with a lateral tooth on each side.

X 70.

.

PLATE 9 5.

23

PLATE 96.

PLATE 96.
1-3. Lytechinus dyscritus, sp. nov.

1. Side view. X 5.

2. Actinostome and base of corona. X 8.

3. Abactinal system. X 8.

4-6. Lytechinus callipeplus, sp. nov.

4. Side view. X 5.

5. Actinostome and base of corona. X 7.

6. Abactinal system. X 8.

:96.

-

. - •' ^ ' -"

J

3

i

:

\

PLATE 97.

PLATE 97.
1-3. Nudechinus stictus, sp. nov.

1. Side view. X 8.

2. Actinostome and base of corona. X 9.

3. Abactinal system. X 12.

4-G. Nudechinus scotiopremnus, sp. nov.

4. Side view. X 3.

5. Actinostome and base of corona. X 6.

6. Abactinal system. X 8.

m

•TV I i i

'

PLATE 98.

PLATE 98.
1, 2. Strongylocentrotus pulchellus A. Ag. and Cl.

1. Part of actinostome and base of corona. X 8.

2. Abactinal system. X 8.

3, 4. Lytechinus euerces, sp. nov.

3. Part of actinostome and base of corona. X 5.

4. Abactinal system. X 5.

5-8. Pachycentrotus australise, comb. nov.

5. Part of actinostome and base of corona. X 5.

6. Abactinal system. X 5.

7. Two ambulacra! plates from mid-zone, showing 5 elements in one. X 5.

8. The same seen from within the test. X 5.

"ALBATROSSTACIFIC AND HAWAIIAN ECHINI

PLATE 98.

^ ^ J) - 4 J

V ' -S

^1 <4

»tm

^ •• ^,

•:> m

'•^: 2^f

-; " ,

S?qi^

-,J Ojt —J *

-S-T,

-Lf

4

*

Ul ^.

«->V®^'0
y '_,,'.

-r^'^ '- . •.
M? goig

^-x J ^ -j ' - j i •

,'•

~r? - u; r- ^

/-^ " j .-j • ^J

j ' '^~^ "'

^ '., ; ='":-

1 " " ! ^f

i

PLATE 99.

PLATE 99.
1-3. Temnotrema hawaiiensis, comb. nov.

1. Side view of denuded test. X 6.

2. Part of actinostome and base of corona. X 8.

3. Abactinal system. X 10.

4, 5. Lytechinus anamesus, sp. nov.

4. Part of actinostome and base of corona. X 5.

5. Abactinal system. X 5.

6, 7. Lytechinus pictus, comb. nov.

6. Part of actinostome and base of corona. X 4.

7. Abactinal system. X 4.

"AlBATROSS'TACIFIC AND HAWAIIAN ECHffll.

PLATE 99.

^-::^,

,^.^m

- ;'->/; •: •"• "L J • iw : — -

-^^f^^.;^S:>-

-», •^•--'-J- - • ^---j; -•.•"- :y -'-
*-~ ," • -i: - -j .'?.. o-

v^

j) *-s - •. ^ ,', - .• • -

r^..' ..^•.•s

- a^a^r^

j -^ ,i a

gp |

.,. --/J . -:•:>

• J'^ Ul) -^ J^

^ '.^ ' Oo \S'«. -^

..'••••• • j- ;

'- i) '-' ' '" ^
•31

.,;-

J O) - -" '

- - • -' - - _

-., J >J i 4-' • '

, ^ c - g

a

.Ikioffln

PLATE 100.

PLATE 100.

1-3. Genocidaris apoda A. Ag. and Cl.
1. Side view. X 5.

Part of actinostome and base of corona. X 9.

3. Abactinal system. X 7.

4-6. Prionechinus ruber A. Ag. and Cl.

4. Side view. X 5.

5. Part of actinostome and base of corona. X 8.

6. Abactinal system. X 7.

"ALBATROSS'PACinC AND HAWAIIAN ECHINI

PLATE 100.

_u

& ,. ^f^*J ^

\

S

j)@ J .^J

yf S@ C®^,

^ * ^/
* j }

m

^i ^^. *n -,»^;^ e

^^-^ * is , , j

PLATE 101.

PLATE 101.
1-3. Prionechinus depressus A. Ag. and Cl.

1. Side view of denuded test. X 5.

2. Part of actinostome and base of corona. X 10.

3. Abactinal system. X 7.

4-6. Prionechinus sculptus A. Ag. and Cl.

4. Side view of denuded test. X 5.

5. Part of actinostome and base of corona. X 10.

6. Abactinal system. X 10.

"ALBATROSS'PACIFIC AND HAWAIIAN ECHINI

i

'i:

•J .i-

" j ,

1

• .

PLATE 102.

PLATE 102.
1. Podocidaris ornata, sp. nov.

1. Abactinal view. X 6.

2, 3. Gymnechinus megaloplax, sp. nov.

2. Part of actinostome and base of corona. X 5.

3. Abactinal system. X 6.

4, 5. Gymnechinus epistichus, sp. nov.

4. Part of actinostome and base of corona. X 5.

5. Abactinal system. X 5.5.

"AiBATROss'PAcmc AND HAWAIIAN ECHINI.

PLATE 102.

ssibv

*K^%2

fip^U ^ .••: '-ly/rTv/ -s3'-> 5)'*~=5>

• |^^§it^^|
?m W®lj©

J&=4>la <#**•- *- -> j^j ?yO> ?-Q
^.^;;;- ^ Y/^:

v^ - ' ^ +••—-)•

•; ^ • *;:. ;;

WJ'.--

-rsrta del

PLATE 103

PLATE 103.
1-3. Caenopedina pulchella, comb. nov.

1. Abactinal system. X 5.

2. Part of actinostome and base of corona. X 8.

3. Two ambulacral plates at mid-zone. X 8.

4, 5. Echinus anchistus, sp. nov.

4. Abactinal system. X 6.

5. Part of actinostome and base of corona. X 6.

6, 7. Salmacopsis olivacea Dod.

6. Abactinal system. X 5.

7. Part of actinostome and base of corona. X 4.

rc AND HAWAIIAN 1

E 103

•.'-. j)<

/•v-, x"^

0 |.

, ^j -- - ' .. " . • ^ •

^^'- j;-,:^V,

^

PLATE 104.

PLATE 104.
1-3. Heliocidaris stenopora, nom. nov.

1. Abactinal system. X 4.

2. Part of actinostome and base of corona. X 3.

3. Part of ambulacrum at ambitus, showing vertical arrangement of outer pore-pairs. X 5.

4. Holopneustes pycnotylus, sp. nov.

4. Part of ambulacrum and interambulacrum, showing relative width, arrangement of

pore-pairs, and tuberculation. X 5.

5. Amblypneustes triseriatus, sp. nov.

5. Part of ambulacrum and interambulacrum, showing relative width, arrangement of pore-

pairs, and tuberculation. X 9.

6. Amblypneustes pachistus, sp. nov.

6. Part of ambulacrum and interambulacrum, showing relative width, arrangement of

pore-pairs, and tuberculation. X 5.

HD HAWAIIAN ECHINI

ECHINI

':'--^ '^^'(:^'>
^::'^-^;^^v^"^:^^£

.5

PLATE 104

•*•• X • "- • :^.^J •' .- :

,'V;''"..

PLATE 105.

PLATE 105.
1-5. Caenopedina hawaiiensis, nom. nov.

1. Abactinal view.

2. Side view.

3. Abactinal view of denuded test.

4. Side view of same.

5. Actinal view of same.

G, 7. Caenopedina pulchella, comb. nov.

6. Side view of partly denuded specimen.

7. Abactinal view of same.

8. Caenopedina mirabilis Mrtsn.

8. Abactinal view.

9. Echinostrephus aciculatus A. Ag.

9. Side view of unusually fine specimen; Laysan, H. I.

10, 11. Echinostrephus molaris A. Ag.

10. Side view of denuded test; Mauritius.

11. Abactinal view of same.

All figures natural size.

"ALEATROSS'TACIFIC AND HAWAIIAN ECHINI.

PLATE 105

Heliotjpi

PLATE 106.

PLATE 106.
1, 2. Glyptocidaris crenularis A. Ag.

1. Side view of unusually fine specimen; Kinka San Light, Japan.

2. Abactinal view of same.

Both figures natural size.

-ALBATROSS" PACIFIC AND HAWAIIAN ECHINI,

E 106

,

PLATE 107.

PLATE 107.
1-3. Echinus lucidus Dod.

1. Side view of high, short-spined specimen.

2. Side view of long-spined specimen.

3. Side view of low, short-spined specimen.

4-6. Lytechinus euerces, sp. nov.

4. Abactinal view of type-specimen, partly denuded.

5. Side view of same.

6. Actinal view of same.

7-11. Lytechinus anamesus, sp. nov.

7. Abactinal view.

8. Side view.

9. Abactinal view of denuded test.

10. Actinal view of same.

11. Side view of same.

12-14. Lytechinus pictus, comb. nov.

12. Abactinal view of partly denuded specimen.

13. Side view of same.

14. Actinal view of same.

All figures natural size.

"ALBATROSS" PACIFIC AND HAWAIIAN ECHINI.

PLATE 107

"o.Boston,

PLATE 108.

PLATE 108.
1, 2. Echinus Wallisii A. Ag.

1. Side view of partly denuded, very large specimen.

2. Abactinal view of same.

Both figures natural size.

"ALB AT : I - /--CIFIC AND HA , CHINI

PLATE 108

''o. Boston.

PLATE 109.
1-3. Echinus tylodes, sp. nov.

1 . Ahactinal view of type-specimen partly denuded.

2. Side view of same.

3. Actinal view of same.

4-6. Echinus euryporus, sp. nov.

4. Abactinal view of type-specimen partly denuded.

5. Side view of same.

6. Actinal view of same.

All figures natural size.

"AU?ATROSS"PACIFIC AND HAWAIIAN ECHINI.

PLATE 109

licliot)peCoBostoi

Plate 110.

PLATE 110.
1-3. Echinus atlanticus Mrtsn.

1. Abactinal view of unusually fine specimen; Ascension Island.

2. Side view of same.

3. Actinal view of same.

4, 5. Heliocidaris stenopora, nom. nov.

4. Abactinal view of A. Agassiz's type of Toxocidaris mexicanus, partly denuded.

5. Side view of same.

All figures, natural size.

•'ALBA ': • ACIFIC AND HAWAIIAN r

E 1 1

?'

PLATE 111.

PLATE ill.
1-3. Echinus anchistus, sp. nov.

1. Abactinal view of type-specimen, partly denuded.

2. Side view of same.

3. Actinal view of same.

4-6. Salmacis erythracis, sp. nov.

4. Abactinal view of type-specimen, partly denuded.

5. Side view of same.

6. Actinal view of same.

7, 8. Nudechinus multicolor, comb. nov.

7. Side view of Yoshiwara's type-specimen, partly denuded.

8. Abactinal view of same.

9, 10. Strongylocentrotus pulchellus A. Ag. and Cl.

9. Side view of type-specimen, partly denuded.
10. Abactinal view of same.

All figures natural size.

"ALBA1'" s ."PACIFIC AND HAWAIIAN ECHINI

PLATE 111

Heliot^pt Co Boston,

PLATE 112.

PLATE 112.
1, 2. Temnotrema sculpta A. Ag.

1. Abactinal view of type-specimen, now a denuded test.

2. Side view of same.

3, 4. Salmacopsis olivacea Dod.

3. Side view of large specimen.

4. Abactinal view of same.

5. Amblypneustes triseriatus, sp. nov.

5. Side view of denuded test.

6-9. Holopneustes pycnotylus, sp. nov.

6. Side view of type-specimen, partly denuded.

7. Actinal view of denuded test, with pentagonal outline.

8. Actinal view of denuded test, with circular outline.

9. Side view of same.

10, 11. Amblypneustes pachistus, sp. nov.

10. Abactinal view of a large, denuded test.

11. Side view of same.

All figures natural size.

"ALBATROSS'TACIFIC AND HAWAIIAN ECHINI.

PLATE 112

;,

12

Jfi

;

PLATE 113.

PLATE 113.
1. Strongylocentrotus echinoides A. Ag. and Cl.

1. Side view of partly denuded specimen.

2. Strongylocentrotus polyacanthus A. Ag. and Cl.

2. Side view of type-specimen, partly denuded.

3-6. Strongylocentrotus fragilis Jackson.

3. Side view of type-specimen; test broken, showing auricles.

4. Abactinal view of same.

5. Side view of a smaller specimen.

6. Abactinal view of same.

All figures natural size.

"ALBATROSS" PACIFIC AND HAWAIIAN ECHINI.

-

HeliotyptCo.Bostai,

PLATE 114.

PLATE 114.
1, 2. Echinometra oblonga Bl.

1. Side view of an unusually fine specimen; Laysan, H. I.

2. Side view of a slender-spined specimen; Clarion Island.

3, 4. Echinometra insularis, sp. nov.

3. Side view of specimen from Easter Island.

4. Abactinal view of another Easter Island specimen.

5, C. Echinometra picta A. Ag. and Cl.

5. Side view of a small specimen.

6. Abactinal view of same.

All figures natural size.

PACIFIC ANOHAWAIIA

E 114.

PLATE 115.

PLATE 115.
Heterocentrotus mammillatus Br.

Abactinal view of specimen from Mauritius.
Natural size.

"ALBATROSS "PACIFIC AND HAWAIIAN ECHINI.

PLATE 115

HeliotypeCo.Boston,

PLATE 116.

PLATE 116.

Heterocentrotus mammillatus Br.

Abactinal view of specimen from Laysan, H. I.
Natural size.

"ALBATROSS "PACIFIC AND HAWAIIAN ECHINI.

PLATE 117.

PLATE 117.
1, 2. Heterocentrotus mammillatus Br.

1. Ahactinal view of specimen from Hawaiian Islands.

2. Actinal view of same.

Both figures natural size.

:-;c AND HAWAIIAN ECHINI

PLATE 11 7

Ifelinjpe

PLATE 118.

PLATE 118.
Heterocentrotus trigonarius Br.

Abactinal view of a specimen from Mauritius.
Natural size.

"ALBATROSS" PACIFIC AND HAWAIIAN ECHINI.

PLATE 118

Heliotype Co. Boston,

PLATE 119.

PLATE 119.
1, 2. Heterocentrotus trigonarius Br.

1. Actinal view of a specimen from Fakarava, Paumotu Islands.

2. Abactinal view of same.

Both figures natural size.

"ALBATROSS" TFfcCIFIC AND HAWAIIAN ECHINI.

PLATE119

Heliotypf

PLATE 120.

PLATE 120.
1, 2. Heterocentrotus trigonarius Br.

1. Actinal view of a specimen from Society Islands.

2. Abactinal view of same.

Both figures natural size.

••ALEATROSS';R<\CIFIC AND HAWAIIAN ECHINI.

PLATE 120

Heliotype Co Boston,

PLATE 121.

PLATE 121.
1-3. Amblypneustes pachistus, sp. nov.

1. Abactinal view of type-specimen, partly denuded.

2. Side view of same.

3. Actinal view of same.

4-6. Amblypneustes grandis, sp. nov.

4. Abactinal view of type-specimen, partly denuded.

5. Side view of same.

6. Actinal view of same.

All figures natural size.

,

PUBLICATIONS

OF THE

MUSEUM OF COMPAEATIVE ZOOLOGY

AT HARVARD COLLEGE.

There have been published of the BULLETIN Vols. I. to LIL;
of the MEMOIRS, Vols. I. to XXIV., and also Vols. XXVI. to XXIX.,
XXXI. to XXXIV., XXXVII., XXXVIII., and XLI.

Vols. LIII. to LV. of the BULLETIN,, and Vols. XXV., XXX.,
XXXV., XXXVI., XXXIX., XL., XLII. to XLVIL, of the
MEMOIRS, are now in course of publication.

A price list of the publications of the Museum will be sent on appli-
cation to the Director of the Museum of Comparative Zoology, Cambridge,
Mass,